This PDF file contains the front matter associated with SPIE Proceedings Volume 12922, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Electric power failure is the most common problem in power supply, and it is also one of the biggest problems for customers. From the perspective of power management, blackouts can be divided into fault blackouts and planned blackouts. There are three types of reasons for failure and power failure. One is overload of public transformer and special transformer, iron core grounding and lightning breakdown. The other is drop fuse failure and lack of equivalence for special transformer users. The last is power failure for residential users, line disconnection after meter, short circuit, leakage switch failure, etc. For planned outage, it is mainly caused by planned maintenance and repair according to the operation cycle of power line equipment. Among them, planned power outages are controllable, and the power management unit notifies the power customers in advance. However, the failure of power outages usually causes certain damage to the power customers, so it has certain harmfulness. In view of the above problems, this paper uses digital twinning technology to analyze the causes of failure and outage, and uses digital technology to carry out failure and outage prediction to reduce the losses caused by failure and outage.

Currently, there have been many studies on digital twin models and analysis algorithms for power application scenarios. These studies mainly focus on building digital twin models for static power equipment and existing historical data. With the positive development of China's economic situation, the number of electricity customers continues to grow, and the scale of corresponding power supply facilities and equipment is also expanding. Accordingly, digital twin models built for static power grid equipment and facilities and power grid structures have to be updated to meet new model application requirements. To address this issue, this paper proposes an intelligent adaptive digital twin model reconstruction method for distribution networks, which reconstructs and optimizes existing digital twin models and reduces the cost of model construction. Reconstruction of the digital twin model for distribution network equipment requires first extracting the basic data of the distribution network equipment, including equipment nameplates, geographic information, and point cloud scanning data. Then, by using the equipment as the identifier, multi-source heterogeneous data across topics are fused. Finally, based on probabilistic and statistical methods, the data structure within the model is reconstructed and optimized for grid business data requests.

The coverage area of base stations has shrunk due to the rapid growth of 5G technology in China, but the quantity of base stations needed to cover the same region has significantly grown. For a significant number of dispersed poor coverage locations to be adequately covered by existing base stations, additional base stations are needed. This study incrementally optimizes the placement and design of a new base station in order to save costs while taking into account the base station's various coverage areas. Regional clustering of weak coverage points is done to address the coverage issue with such places. The new base station should initially cover the poor coverage points with big service volumes since the base station's coverage area is circular and the service volume of the weak coverage points covered is not less than 90% of the total number. A two-dimensional array is used to describe the position of the weak coverage point, the traffic, the location of the existing base station, and the location and type of the new base station. The computation is simplified by first excluding the fence points that are too close to the current base station and then optimizing the remaining fence points to determine if the base station can be erected and what kind of base station.C language programming was used to build relevant software to deal with this problem, and the data was cleaned and the center point of the grid was divided into 60×60 and 200×200 regions. Through programming operation, the optimization setting result of the base station was as follows: Construction of 2178 microbase stations and 1 macro base station, which together cover 90.0055795% of the total traffic, at a cost of 2188 units. Through the distribution of only micro base stations, the optimization model is established. After the improvement of the program and the cleaning and screening of the data again, the optimal base station setting result is finally obtained: the construction of 2187 micro base stations costs 2187 units, and these base stations cover the total business volume of 90.0181713819829%.

With the continuous growth of global demand for renewable energy, photovoltaic power generation as a clean and sustainable energy solution has received increasing attention. However, the operation monitoring and management of photovoltaic power generation systems face challenges such as real-time monitoring and fault detection. Based on this, this article proposes the design and research of a digital photovoltaic station operation monitoring and management system based on communication technology. Firstly, design a hardware module with high adaptability, utilizing sensors and communication devices to monitor key parameters of the photovoltaic power generation system in real-time, and then obtain and process system operation data through data collection and transmission. In addition, the system has designed fault diagnosis and early warning functions, as well as remote control and scheduling management modules, so that management personnel can timely detect abnormal situations in the photovoltaic substation area and notify operation and maintenance personnel to take timely processing measures in the form of early warning. Finally, a testing experiment was designed to verify the performance of the system in real-time monitoring, fault detection, data transmission, and other aspects. The experimental results show that the system can obtain real-time operational data of photovoltaic power generation systems, accurately detect faults in the substation area, and has stable data transmission and efficient data processing capabilities. It is expected to improve the operational management efficiency of photovoltaic substations, enhance the reliability and maintenance convenience of photovoltaic substation operation, and contribute to the development of sustainable energy.

Photovoltaic power generation, as an important component of renewable energy, is gradually becoming one of the key technologies in the energy field. This study aims to explore the operation and communication mechanisms of digital photovoltaic stations, in order to improve the efficiency, stability, and reliability of photovoltaic power generation systems. By collecting photovoltaic power generation data, optimizing energy management and communication networks, and implementing fault detection and maintenance measures, intelligent operation of photovoltaic power stations can be achieved. The performance and effectiveness of digital photovoltaic stations were evaluated through simulation experiments, providing a theoretical and practical basis for the optimization of photovoltaic power generation systems.

This paper presents an efficient design technique for array signal preprocessing spatial matrix filters. Two approaches were considered when designing the filter. Firstly, the first approach minimizes the passband fidelity by controlling the attenuation level of the overall stopband approach, and secondly the second approach by inverting the objective function and constraints. The optimal solution of the optimization is derived through the theory of Lagrangian multipliers. By analyzing the relationship between the optimal solutions, and then introducing generalized singular value decomposition to simplify the design of the optimal solutions of the two methods and improve the efficiency of solving Lagrangian multipliers. Finally, the simulation results show that the method is effective for the design of space matrix filter.

According to the requirement of four-way responder, two different methods of digital circuit and PLC are designed and compared in this paper. The first method is to use the integrated chip "4511" and "555" and the corresponding peripheral circuit to achieve, according to the four-way scrambler design block diagram, successfully designed the schematic diagram and the corresponding PCB diagram, and actually made the finished product; The other method is to use programmable logic controller PLC to realize, designed the ladder diagram and the corresponding statement table of the four-way responder PLC control, and actually made the finished product. Both of these methods have been successfully implemented, have good application value in practice, and can be based on this idea, expand into more and better multi-route responder.

With the development of power grid digitalization, miniaturized, highly sensitive, low-power sensors are required to provide high-precision sensing and measurement capabilities of current signals in a wide band under complex working conditions. This paper analyzes the sensing principle of TMR magnetic sensor and the frequency band response characteristics of TMR current sensor, builds a TMR sensor bandwidth test system, carries out the bandwidth and parameter test analysis of four TMR sensors with different internal resistance, verifies the frequency response characteristics of current sensors with a bandwidth of up to 5MHz, and greatly expands the application range of TMR sensors.

Transmission interference prediction is becoming more and more important in mobile communication networks, and DM (data mining) has become an important means to realize transmission interference prediction in mobile communication networks in the future. In order to improve the success rate of mobile communication information transmission, a communication transmission prediction model based on DM algorithm is proposed. In this paper, DM technology is used to mine transmission interference data. The obtained mining sample data is stored in a distributed data storage system. Then, the DM process is studied, the dynamic termination threshold is put forward, and the transmission interference prediction model is obtained by using parallel algorithm for DM. The results show that the average transmission delay increases after using controlled data transmission, because the probability of delayed transmission increases while the average probability of interference decreases. The average prediction accuracy of information transmission interference of mobile communication system in this model is 96.477%. It can be seen that this model reduces the prediction error of information transmission in mobile communication system and can improve the information transmission effect of mobile communication system, which has obvious advantages.

The quality and safety of bridges are related to the national economy and people's livelihood, and to thousands of families. The reason for bridge collapse is often due to the lack of scientific and timely bridge disease detection. However, when the bridge is completed and put into use, it will be repeatedly crushed by external forces and affected by natural factors such as rain, snow, flood, freezing and earthquake. As one of the most important bridge diseases, cracks seriously affect the safe operation of bridges, and even lead to accidents of bridge destruction and death. In view of this situation, this paper studies the bridge crack detection based on wireless sensor network. By setting appropriate threshold, the background and target are divided and the relevant feature information of cracks is extracted. It is composed of an off-line data processing system and a data acquisition vehicle. Finally, through the evaluation and comparison, the favorable rate of this algorithm is higher than the other two algorithms, and the favorable rate will be as high as 88.42%. In the test of running time, the running time of this algorithm is shorter than that of the other two algorithms, and the average running time is only 6.02 seconds. It can be seen that this algorithm is more suitable for bridge crack detection.

This paper analyzes the 5G high-precision synchronization networking architecture for the current demand of high precision and low latency in 5G applications. By comparing the respective advantages and deployment ideas of several different deployment methods of sub-time reference sources in synchronization technology, a method of deploying sub-time reference sources in the convergence layer to improve 5G time accuracy and reduce latency is proposed. After the actual scheme verification and testing, the experimental results meet the requirements of 5G high-precision synchronization and have certain significance for future practical commercial deployment.

With the continuous development of information and communication technology, new changes have taken place in the call industry. The transformation from traditional call center to Internet call center has become an important strategy of today's power enterprises. Through transformation, enterprises can not only manage business processes more effectively, but also reduce enterprise costs and improve customer service quality. This paper firstly summarizes the development of call center. Then the whole architecture and key technologies of traditional call center and Internet call center are introduced. Finally, based on the comparative analysis of the advantages and disadvantages between the traditional call center and the Internet call center, the inevitable trend of the call center transformation under the new situation is summarized.

An AOA (Arrive of Angle) algorithm based on phase difference projection is proposed for the problem of inaccurate localization results of signal tags in indoor positioning. In terms of position localization accuracy, there is systematic error when performing AOA localization, and the area that can actually be localized with higher accuracy is small. Finally, the localization algorithm using AOA and RSSI (Received Signal Strength Indication) information fusion can effectively reduce the influence of multipath effect caused by indoor bad environment and improve the indoor localization accuracy. The proposed AOA and RSSI fusion algorithm is simulated and experimented, and the obtained experimental results show that the localization accuracy is effectively improved.

In the pipeline leak location technology by detecting Infrasound signals, time delay estimation algorithm has become a significant method, but the quality of time delay estimation calculation method determines the accuracy of location. This article proposes an improved Smooth Coherent Transform (SCOT) generalized weight function based on the quadratic correlation delay estimation algorithm, defined as SCOT-γ to suppress the false peak of the signal by adding weight to the SCOT weight function , the two cross-correlations in the quadratic cross-correlation algorithm are weighted separately with the improved weight function. In this paper, MATLAB is used for experimental simulation. First, Infrasound signals with Gaussian white noise are constructed,then the influence of the weighting factor on the delay estimation performance is verified through multiple experiments, and the optimal value of the weighting factor γ is determined at the same time. Then perform cross correlation peak detection on the SCOT primary weighting method, SCOT-γ once weighting method, and SCOT-γ double weighting method and compare the noise reduction effect; finally, compare the performance of the three algorithms by calculating the root mean square error of the delay value of the three algorithms. The results show that the improved SCOT double weighted quadratic correlation time delay estimation algorithm has better anti-noise performance and higher precision.

For the sampling points of sensors at different moments, a sequential asynchronous algorithm to improve multi-sensor fusion is proposed after using the idea of sequential discretization to achieve fusion estimation of position states. It is assumed that the system uses N sensors with different sampling rates for the observed position states. Among them, a multi-rate Kalman filter is designed to estimate the target position at multiple scales with multiple sampling rates. The equation of state between the different sampling points during each sampling moment of the fusion cycle can be obtained as well as the measurement equation at this point in the cycle. Using an optimized linear Kalman filtering algorithm, a sequential fusion process of signals observed by asynchronous sampling is investigated in one cycle to achieve a good estimation of the target position. The effectiveness of the improved algorithm is verified by a numerical example simulation, and the advantages of the algorithm in terms of less computation, good real-time performance and high estimation of fusion accuracy are evident.

Patrol inspection is an important part of the daily work of the power plant. The quality of patrol inspection is often related to the safe operation of the power plant. In the modern power system, there are many and complex parameters of the equipment. It is a time-consuming and laborious matter to judge whether the parameters exceed the standard only by manual work. Therefore, how to use the powerful computing and storage capabilities of the ironing network to reduce the analysis of manual operation while ensuring information security is a problem worth studying. The beneficial effects of the power plant patrol management system studied in this paper are: the use of data source signal acquisition, including temperature, humidity, vibration, etc., reflects the operation status of the equipment more comprehensively, and the GPS positioning and identity verification are set up at the same time. The use of cloud computing greatly reduces the calculation and storage of the local server.

With the accelerated digitization process, it is crucial to accurately identify potential 5G demand users through models to facilitate the transition from the 4G era to the 5G era in order to realize the construction of smart cities based on deep 5G applications. In order to accurately and efficiently identify 5G telecom users and analyze the business features that have important impacts on 5G users, this paper designs a 5G user prediction model combining k-means and classifier based on data mining. This model is based on k-means clustering to deal with the current dilemma of unbalanced sample size of 5G users and non-5G users, and additionally uses a bagging strategy to integrate two base classifiers, LightGBM and CatBoost, to effectively improve the model generalization capability. In comparison with the baseline model, the accuracy of the model is 87.7% and the F1-score is 70.4%. The results are satisfactory. With this model, 5G users can be identified with higher probability, which is a great help for mobile operators to identify potential users for accurate marketing. At the same time, the business characteristics highly correlated with 5G users also provide guidelines for 6G service design and package designation.

With the development of urbanization and commercialization, it is an inevitable trend to develop and apply autonomous intelligent vehicles to solve traffic congestion and improve travel efficiency. This paper mainly studies and designs the implementation method of the auto drive system of electric vehicles, focuses on the software and hardware technology scheme of the auto drive system, and studies the technical route of the development and implementation of the auto drive system. The research results can provide some reference for the development and application of the auto drive system of intelligent vehicles.

This paper uses virtual reality technology to build a set of modern electrical control technology engineering virtual reality interactive independent practice platform. The platform converts the explicit, implicit, and random measurement and control data of modern electrical control technology engineering into physical entity measurement and control equipment information and realizes the closed-loop control between "virtual body" and "entity" through the control circuit and control program. To improve the quality of students' intelligent manufacturing measurement and control engineering practice and training, reduce the cost of training, to achieve the teaching objectives of training technical skills to meet the needs of intelligent manufacturing positions.

In order to improve the security of power IoT (Internet of things) data and enhance network expansion, strengthening the construction of power wireless private network is the most effective way. In this paper, the access and management technology of massive heterogeneous devices in power IoT is studied, and the connection management technology of distributed full asynchronous architecture is proposed, which realizes the access of massive heterogeneous devices in power IoT in low power IoT scenarios. The distributed fully asynchronous architecture has high flexibility and can expand distributed management nodes according to the actual deployment of the network. Therefore, this paper proposes a network selection algorithm based on combination weight. By combining the weights of network decision indicators, the final ranking result can not only reflect the user's needs, but also reflect the objective performance of the network. The simulation results show that the network access blocking rate of this algorithm is low, and the optimal network selected by this algorithm not only reflects the network performance, but also meets the QoS requirements of users, which effectively improves the accuracy of network selection in heterogeneous wireless network environment.

In order to ensure the stable, safe and efficient operation of the power system, it is need to continuously strengthen the inspection of transmission lines in the power system. The traditional manual patrol has a large workload and low efficiency, and it is difficult to realize manual patrol when geological disasters occur. On-line detection system of transmission lines can judge a certain fault abnormality of transmission lines in real time and accurately. In this article, an intelligent pre-alarm model for online inspection system of transmission lines is proposed, which uses image recognition technology as the main judgment method to judge the fault abnormality of transmission lines in real time, accurately and reliably. This model applies Convolutional Neural Network (CNN) to the state detection and fault identification of transmission lines, thus providing technical support for the construction of on-line intelligent patrol system of substations. The simulation results show that compared with other methods, the proposed method has higher detection accuracy, faster detection speed and lower detection error, and can intuitively and accurately grasp the equipment status of the substation. The application of the intelligent pre-alarm system can effectively realize the intelligent identification of abnormal faults of transmission lines and their equipment.

Aiming at the problems of poor interpretability of traditional neural network and bloated traditional model in photovoltaic panel defect detection. Firstly, an FN-v8 model was proposed for detecting various types of defects in photovoltaic panels. The FN-v8 model reduces redundant computing and memory access, while also reducing the GFLOPs of the model. Then, Grad-CAM was used to further demonstrate the stronger interpretability of the new model through passive interpretation. Compared with other traditional models with relatively bulky structures, the proposed new model has smaller parameter quantities and stronger interpretability. Experiments were conducted on devices with the same performance and the same dataset of photovoltaic panel defects. The results showed that the proposed new model has a smaller volume, faster calculation speed, and higher accuracy than the original model. The GFLOPs of the new model have decreased by 18.25% compared to the original model, significantly reducing the computational complexity, and the visualization of the thermodynamic diagram is more accurate, further indicating an improvement in interpretability. Moreover, the new model has better detection accuracy for photovoltaic panel detection, with an average detection accuracy ( mAP@0.5) improved by 2.1% compared to the original model.

In order to reduce torque ripple and current ripple of permanent magnet synchronous motor (PMSM) in traditional model predictive control, a improved predictive flux control scheme is proposed. First,synthesize 12 virtual voltage vectors. In order to simplify the vector selection, divide each vector into 12 sectors, quickly determine the vector range by calculating the voltage vector angle, reduce the optional 12 voltage vectors to 2, Simply calculate the cost function twice to select the first required vector. Secondly,Pairwise combine the first vector selected with the remaining 11 vectors, and the unwanted voltage vectors are filtered out through the designed flux linkage cost function to Select the two vectors with the best control effect. Flexibly add a zero voltage vector according to the switching characteristics of the inverter to adjusting the magnitude of voltage and vector angle, and a fuzzy controller is designed to improve the evaluation accuracy of speed loop. Finally, use the flux difference parameter to calculate the occupancy time of each voltage. The experimental results show that the proposed scheme reduces the torque current fluctuation and improves the stability of the system.

With the development of autonomous driving technology, intelligent vehicle detection and tracking has become a key research field. The traditional method based on a single visual sensor faces challenges in complex scenes. Therefore, the method of multi-visual information fusion has become an important way to improve the performance of detection and tracking. This research aims to explore an intelligent vehicle detection and tracking method based on multi-visual information fusion function model. First, we propose a multi-visual information fusion functional model and combine them through a feature fusion module. In this way, we can comprehensively utilize the advantages of multiple vision sensors to improve the accuracy and robustness of detection and tracking. Extensive experimental evaluations are then performed on real datasets. The final experimental results show that the intelligent vehicle detection and tracking method based on the multi-visual information fusion function model has higher accuracy and robustness than the traditional method. By comprehensively utilizing information from multiple vision sensors, we can better deal with complex scenarios and challenges, and provide more reliable perception capabilities for autonomous driving systems. The results of this research are of great significance for promoting the development of intelligent vehicle detection and tracking technology, and also provide reference and reference for multi-sensor information fusion methods in other fields.

The safe operation of smart substation is an important guarantee to ensure the stability of power supply. Aiming at the problems of substation data anomaly identification caused by large amount of substation data, complex data and high dimension, a smart substation anomaly data detection method based on K-means-SVM is proposed in this paper. Firstly, the unlabeled substation data is transformed into labeled data by K-means method, and then the unsupervised classification results provided by K-means are used as the training data of SVM. After SVM training, use the trained plane to test the two types of data obtained by K-means clustering, use the accurate data predicted by SVM to retrain the SVM segmentation plane, and iteratively update the SVM segmentation plane according to this method until the error number of SVM prediction data is not changed. This method can effectively improve the accuracy and efficiency of substation abnormal data identification, so as to improve the operation and maintenance safety and equipment maintenance ability.

In this paper, the digital modeling technology based on multi-source data of power grid is studied. First, the theoretical system of power grid digital twin model is established to provide theoretical reference for the modeling and analysis of power grid digital twin and further application. Secondly, the construction of power grid digital twin model based on physics-guided deep learning framework, interface learning framework and probability graph model is studied to create the foundation and prerequisite for subsequent applications, and the foundation and core elements of power grid digital twin. Thirdly, the assembly and fusion of power grid equipment level model, power equipment cluster level model and power system level model based on multi-source heterogeneous data fusion technology are studied to realize multi-level digital twin model fusion for power grid. Finally, the paper studies the dynamic balance of update cost and accuracy to update the digital twin model of power grid and realizes the flexible and dynamic update of power grid twin model.

The China University Student Formula Autonomous Driving Competition is the most prestigious innovation competition in the series of events initiated by the Society of Automotive Engineers of China (SAE-China) Science Popularization Center. In comparison to traditional racing, the focus of the autonomous formula competition lies in the autonomous driving capabilities of the race cars. Wire-controlled steering is a critical component of the lateral control system. This paper presents a nonlinear model predictive controller designed using a dynamic hybrid particle swarm optimization algorithm. Building upon the simplified two-degree-of-freedom vehicle model, the traditional model predictive control algorithm is improved by utilizing the dynamic hybrid particle swarm optimization algorithm to solve the optimization function. Finally, a joint simulation using Simulink and CarSim is conducted to validate the enhanced model predictive controller, demonstrating its superior tracking accuracy and driving stability. The objective of this paper is to provide design guidelines for the WUTA team in the 2023 season and share the experience in designing a steering controller with other autonomous formula racing teams.

In this paper, circuit breaker is taken as the object, and typical electromechanical faults are simulated and set. The characteristic parameters of opening/closing coil current and displacement are extracted. The evolutional trends of signal characteristics under different fault types and severity levels are analyzed. The feature optimization is realized by Spearman correlation analysis and the efficiency of the simplified feature set is evaluated. Finally, the optimal lightweight pattern recognition method is determined by sample training through six classification models. Results show that signal features selected in this paper can accurately reflect the fault types and severity levels.

Time delay estimation algorithms play an important role in the localization techniques for local discharge sources. In this paper, we first briefly analyze the currently existing time delay estimation algorithms such as traditional mutual correlation, generalized mutual correlation and cubic correlation, and then propose an improved generalized cubic correlation time delay estimation algorithm. And by constructing partial discharge ultrasonic signals, the delay estimation performance of GCC-PHAT-β algorithm, quadratic correlation GCC-PHAT-β algorithm, cubic correlation GCC-PHAT-β algorithm and the improved algorithm in this paper are verified in different signal-to-noise ratio environment, the delay estimation errors of the four algorithms are compared. The simulation results show that this improved algorithm in this thesis has better delay estimation ability and better noise immunity at low signal-to-noise ratio.

The changes in the pattern of electricity energy supply and demand, as well as the trend of energy development, have led to the reform and progress of the domestic and international electricity industry market, accelerating the construction of smart grids in various countries. In recent years, the rapid growth of China's economy has driven the continuous progress of science and technology. Information Technology (IT), Big Data (BD) technology and Artificial Intelligence (AI) are widely used in all walks of life. The traditional power grid is gradually developing towards user demand side response, making system regulation more flexible and allowing users to fully participate in the operation of the power system and energy market. In order to build a power grid user side intelligent management system based on AI algorithm driven by power BD, this article studies a user Demand Side Management (DSM) method based on load transfer technology and uses a demand side user response potential assessment model based on fuzzy optimization set to evaluate its management effectiveness. Through numerical simulation, the task allocation process of load aggregators is simulated, and the rationality, effectiveness, and feasibility of this assessment method are demonstrated from aspects such as task completion and aggregator benefits. The results show that the system can effectively control and manage the operation of various devices, keep the total consumption below the threshold, and manage the load based on priority.

Smart grid scheduling automation system has a high demand for system computation capability, the existing system has no automatic computation and storage means for grid-level real-time monitoring objects, and the model construction and automatic computation of grid monitoring object computation and analysis results need to rely on a large number of heavy manual maintenance, which cannot satisfy the demand of the gradual expansion of the grid scale. In this paper, an automatic calculation method based on abstract rule syntax for grid objects is proposed. Firstly, the calculation logic and the calculation result storage model are designed according to the principle of object-oriented modeling, so that the calculation data have the basis for modeling; secondly, an abstract rule syntax is designed to support the electronic description of the calculation logic; next, a resolution algorithm for the abstract description of the object calculation rules is discussed and the common statistical algorithms for grid objects are cited to verify the calculation results of the grid objects; finally, the parsing algorithm for the abstract description of object computation rules is discussed, and the common statistical algorithms of power grid to verify the completeness and feasibility of the abstract rule grammar are cited.

Gas insulated equipment has the advantages of small size and easy maintenance, and has been widely used in China's power system. However, the internal space of the equipment is narrow and the bonding structure is compact. The internal insulation problem is easy to bring great hidden dangers to the safe operation of the equipment. In this paper, three typical defect discharges in switchgear were simulated. The whole process of internal discharge development was measured by multi-spectral sensor. The multi-spectral parameters characterizing the discharge development process were extracted. Based on the above characteristic parameters, the judgment basis for characterizing the development of discharge to the pre-breakdown stage is proposed.

In view of the long power supply radius in Guangzhou and the low voltage problem caused by the grid connection of distributed small hydropower stations, this paper studies the influence of three-phase line regulators connected in different ways on the fault zero sequence current, and takes the 10 kV Liangkou F4 line as an example to analyze the voltage distribution along the line of regulators connected in different ways under the two operation models of high- water period with light load and low-water period with heavy load. The results show that under the Y connection mode, the zero-sequence impedance of line single-phase fault is very large, and small zero sequence current value is difficult to trigger the zero-sequence protection of substation; The zero-sequence current of line fault current can flow normally under the triangle connection mode, and the zero-sequence protection operates normally. On the other hand, for the typical long power supply lines in Guangzhou, the voltage regulators of the two connection modes can effectively improve the voltage distribution of the long lines, regardless of the wet season, light load operation mode, or dry season, heavy load operation mode, and the closed triangular connection mode has a stronger voltage regulation capability of the Y connection mode. Through comprehensive comparison of the voltage regulating capacity of the two connection modes and their influence on fault zero sequence current protection, it is considered that the closed triangle connection mode is more suitable for the 10 kV long line distribution network in Guangzhou.

A fuzzy variable-gain super-spiral sliding mode observer is proposed to address the problems of jitter and low estimation accuracy caused by the use of constant sliding mode gain in the traditional Sliding Mode Observer (SMO). The fuzzy controller is introduced to replace the traditional observer with a fixed sliding mode gain by a variable sliding mode gain, and to improve the estimation accuracy and system robustness by online adjustment of the sliding mode gain. Secondly, to address the problems of large overshoot and slow response speed caused by the tacho loop PI controller in the traditional control system, a new convergence rate based tacho sliding mode controller is proposed to reduce the system overshoot and improve the response speed. The stability of the proposed sliding mode controller is proved by using Liapunov's stability theorem. Finally, Matlab/Simulink simulation is used to verify that the proposed improved control system has faster response speed, higher estimation accuracy and better system jitter suppression.

Electric vehicles play an increasingly important role in the automobile industry, and electro-hydraulic compound feedback braking is one of the contents that need to be further studied. Its purpose is to fully recover braking energy on the basis of ensuring vehicle braking stability, so as to improve vehicle energy economy. At present, no breakthrough has been made in battery energy storage technology. Automobile braking can be used as supplementary energy of batteries, and the driving range of electric vehicles is a key problem to be solved urgently. In this paper, the intelligent energy feedback system for electric vehicle braking is designed based on fuzzy control algorithm, and the electro-hydraulic composite braking force distribution fuzzy controller with feedback braking ratio as output is analyzed, and the constraint conditions and control strategy in energy feedback braking are analyzed, and the constant torque energy feedback braking control is selected as the control strategy of this paper. Through the simulation comparison of fuzzy control strategy, the model of intelligent energy feedback braking system for electric vehicle braking is built with the fuzzy controller as the control core. Through simulation, it can be seen that the proposed constant torque fuzzy control strategy can realize energy feedback braking and achieve ideal results.

The driver has to face many uncontrollable factors in the process of driving, which can lead to traffic accidents if he is not careful. LabVIEW, a graphical programming software, is used to design a distance monitoring system that monitors the distance travelled by an intelligent vehicle in real time, displays it dynamically and enables human-machine interaction. The serial communication part of the system is designed through the VISA toolkit in LabVIEW. It receives the distance measurement information from the lower computer system through the serial port, so that the upper computer system can judge the vehicles, pedestrians and obstacles around the driver in real time, monitor the traffic condition around the road in real time, display it dynamically and give safety tips when it is lower than the safe distance, thus helping the driver to make a safe distance judgment in time. This helps the driver to make timely judgements on the safe distance so that he can be prepared to brake and avoid traffci accidents.

For the detect recognition of transmission line insulators, there are some problems for the traditional fault recognition algorithms, such as false detection, omission of detection, low and slow recognition rate. Compared with the traditional convolutional neural networks, vector is used by the capsule network as the input. Each sub-structure of the capsule makes the details highly fidelity in the original graph, which can effectively identify the defective insulator image. Therefore, a detect recognition method based on improved capsule network and YOLO-V5 is proposed in this paper. Experimental results demonstrate that the algorithm performs well and meets the requirements for inspecting insulators.

With the deepening of power market reform, short-term load prediction becomes more and more important in Electrical Power System (EPS). Combining the historical data of power load with the computer model, we can infer the value or peak value of power load in the future, which is of great significance for power companies to coordinate power allocation and solve the problems of energy and supply and demand. Short-term load prediction is an important part of EPS load prediction, and accurate power load prediction plays an important role in making power generation plans, ensuring production and domestic electricity consumption, controlling economic operation of power grid, and reducing rotating reserve capacity. In this paper, an integrated power data processing architecture is constructed, and a load prediction algorithm based on Genetic Algorithm (GA) and back propagation neural network (BPNN) and an abnormal data detection method based on improved spectral clustering are proposed. The experimental results show that the system has high load prediction accuracy and anomaly detection accuracy, and the operation efficiency is high.

Substation is the core hub of power grid transmission, which provides the necessary guarantee for the safe and stable operation of substation and even the whole power grid. There are many kinds of electrical equipment in substation, and the distribution is dense. Different equipment shows obvious differences in appearance and state. When inspecting this electrical equipment, we can not only use traditional manual inspection methods, but also rely on artificial intelligence technology and big data technology to build an intelligent inspection robot and realize the recognition of image information through image recognition technology. In this article, the application of intelligent inspection robot in autonomous intelligent inspection of substation is studied, and the path planning model of inspection robot is constructed by combining image recognition and computer vision technology. The experimental results show that the path planning algorithm of inspection robot in this article has small error and excellent accuracy, which can reach more than 98% after many iterations. The algorithm realizes the stable, reliable and high-precision inspection operation of the substation inspection robot.

In response to the problem of drone avoidance problems under low energy visibility, a laser point cloud power line positioning method was proposed. This method solves ground equations by POS data instead of traditional iterative methods to improve system real time performance. Then reduce the number of clouds through sub-sampling. Then use Euclidean clustering to separate target point clouds. Because the power line point cloud is different from the targets such as common vehicles, architecture and other goals, has low density and non-rigidity characteristics, we use the point cloud registration method to obtain the power line position. For this purpose, this method has built a power line model with a range of ten meters and collects multiple power line cloud data as experimental materials. The results show that the method can obtain the target power line position quickly and accurately in the selected scenes.

In order to quickly and accurately identify the gas infrared spectral components, a relatively new research result, the extended residual network (DA-ResNet12), is introduced. The gas infrared spectral data measured using an infrared spectroelectronic meter were selected for the experiment. First, the raw data are preprocessed, and the spectral features are extracted, and the feature vectors are generated by feature engineering; then, the lightweight DA-ResNet12 model is built and the model parameters are adjusted; finally, the proposed lightweight DA-ResNet12 model is compared with the commonly used algorithms according to the evaluation indexes such as the classification accuracy, the required training time, and the size of the model parameters. The experimental results demonstrate that the newly proposed DA-ResNet12 network model outperforms other comparative algorithms in terms of the number of model parameters, accuracy and time efficiency.

Aimed at the difficulties in extracting fault features and low diagnostic accuracy of rolling bearings, an intelligent diagnosis method which combines the Variational Mode Decomposition, Envelope Spectrum Analysis and the Optimized Gradient Boosted Random Forest Model is proposed. Firstly, low-pass filtering is used to denoise the measured vibration signal of rolling bearings, and then VMD is used to decompose the processed vibration signal, and the instantaneous energy matrix of the Intrinsic Mode Functions (IMFs) obtained by VMD is calculated by ESA. Then use Principal Components Analysis (PCA) to perform feature extraction on the obtained instantaneous energy matrixes. Finally, put the extracted feature vectors into the gradient boosted random forest model optimized by sparrow search algorithm for fault diagnosis of rolling bearings and the accuracy of fault diagnosis is calculated by cross-validation. The experimental result shows that this method can accurately extract the features of rolling bearings in different fault states, which has a high accuracy in the fault diagnosis of rolling bearings.

In order to solve the problems of high cost, large volume and difficult maintenance when traditional mechanical sensors are used in permanent magnet synchronous motor, a sensorless system model of permanent magnet synchronous motor based on untracked Kalman observer is established in this paper, and the running state of permanent magnet synchronous motor is predicted by untracked Kalman algorithm to achieve closed-loop control. The simulation results show that the traceless Kalman observer has the advantages of accurate prediction, good dynamic and static performance, and strong anti-interference ability. It solves the defects of traditional permanent mechanical sensors to a large extent and has great application prospects in the future.

To address the problem of excessive common mode voltage at the output of permanent magnet synchronous motor for new energy vehicles under high voltage platform, which is likely to cause electromagnetic interference to other electronic devices, this paper proposes a method to suppress common mode voltage using model predictive current control on the inverter circuit of four legs. Firstly, the correspondence between different switching vectors and common- mode voltage is derived, and the fourth leg is decoupled from the other three legs. The control is simplified by optimizing the vector selection, while achieving the common-mode voltage suppression to essentially zero. Then MPCC is used to track current, improving the THD of the load current, and a penalty function is added to reduce the switching frequency in the cost function. Finally, through double closed-loop simulation of the electric drive circuit in MATLAB, it is confirmed that the proposed method in this paper can effectively suppress the common-mode voltage, while improving the quality of the load current, reducing the switching frequency, and effectively suppressing electromagnetic interference.

Aiming at the problem that there are many blind areas in the perimeter of articulated vehicle, this paper presents a real-time dynamic Mosaic method of multi-vehicle grouping image. Firstly, the image is obtained by fisheye camera for distortion correction, perspective transformation and image Mosaic, and the Mosaic of single car is generated. Then the image of each car is loaded to the Graphics Processing Unit (GPU) by the way of texture mapping to achieve acceleration; Finally, a multi-car image Mosaic model is established and the Angle signal at the articulation point is combined to realize the real-time dynamic Mosaic of multi-car perimeter images. The algorithm is transplanted to the embedded platform and tested on the real vehicle. The results show that the panoramic dynamic Mosaic image based on this method can clearly display the vehicle perimeter objects.

In power system operation, more and more attention is paid to the reliability and quality of power supply of the distribution network. As the last kilometer of the power network connecting the user, the reliability of power supply in remote mountainous areas becomes particularly important. This paper takes the power supply reliability improvement measures of a typical remote mountain power supply station as the research object, and improves the management on duty management, safety management and planning management. In the business, it improves the operation reliability of the distribution network through skill training, external construction damage control, tree barrier trimming, lightning protection measures, equipment quality and other methods. Meanwhile, it improves the operation reliability of the distribution network through the planning of medium voltage and low voltage. The feasibility and effectiveness of the measure have been verified through more than one year's operation.

The bearing transmits the transmission system's vibration to the box when the reducer is in operation, not only impacting the stability of the system but also creating noise pollution. Therefore, the accurate prediction of the vibration and noise of the reducer is particularly important. A presentation of a novel main reducer with double planetary gear is made, and the essential parameters influencing the reducer's vibration and radiated noise are systematically scrutinized and appraised. The stress analysis of the two-stage planetary gear system was conducted in the vibration analysis process, followed by the mesh stiffness and supporting stiffness of the shaft, the classification of the coupling system's modal and vibration mode, and the analysis of the dynamic load factor and uniform loading situation, statics, dynamics performance of the gear train of the simulation test in various working conditions. The finite element and concentrated parameter methods were utilized to determine the meshing stiffness of the gear, the supporting stiffness of the bearing, and the planetary pin in a two-stage planetary gear system. The results demonstrate that the maximum stress of each component of the gear reducer is below the yield limit of 20CrMnTi, leaving sufficient allowance. By employing the 45-order Runge-Kutta technique, the dynamic equation was solved, yielding the dynamic load factor under rated torque and peak condition. Utilizing Matlab eig function, the simplified undamped free vibration equation was then solved, yielding the mode natural frequency and mode shape. It is found that the vibration amplitude of the first planetary gear train is large under low frequency. This paper can be used to further refine the design of electric vehicle vibration noise

Aiming at the problem of assembly sequence planning of servo motor components, a method based on particle swarm algorithm to optimize the assembly sequence of servo motor components is proposed according to the characteristics of the assembly problem. This method redefines the related operations of velocity and position in particle swarm optimization. This paper evaluates the assembly sequence in terms of geometric feasibility and factors affecting the assembly process. The geometric interference information of the part is expressed by establishing the interference matrix. The objective function with the assembly tool and assembly direction as the evaluation index is established. Finally, the objective function is combined with the particle swarm algorithm, and the optimal assembly sequence of the servo motor components is obtained, which proves the effectiveness of the method.

For steam power control of steam turbine, electro-hydraulic servo valve is a necessary component. However, the electro-hydraulic servo valve of steam turbine has a high requirement for the driving ability of the drive circuit, and the ordinary drive circuit cannot meet its requirements. Therefore, this paper designs a driving circuit with greater driving capacity for electro-hydraulic servo valve. The problem of insufficient carrying capacity of OP07 is solved by combining OP07 operational amplifier and push-pull circuit to improve its driving capacity. Taking 0-2.5V input voltage as an example, the optimized driving circuit is designed and simulated. The simulation results show that in the voltage mode, the drive circuit can drive the load with smaller resistance value and output larger current than the common electro-hydraulic servo valve drive circuit, which further improves the drive ability of the electro-hydraulic servo valve drive circuit.

A new type of pneumatic magnetic separator is designed in this study. The Dry Magnetic Separation Process (DMSP) is enhanced by introducing a horizontal airflow in the opposite direction to the magnetic force, and the recovery of aim minerals in the intergrowth particles is ensured by setting three recovery zones. The optimal structural parameters are determined by simulating and calculating the magnetic field and flow field characteristics. Based on the tracking calculation of different particle motion states through multi-physical field coupling, the optimal matching scheme of the magnetic field and flow field is determined, and the separation effect is predicted. Hope this design provides a novel and feasible idea for the development of high-efficiency dry magnetic equipment.

Free metallic particle discharge seriously threatens the operation safety of the Gas Insulated Switchgear (GIS). However, limited by the detection sensitivity, anti-interference ability and intelligent diagnosis ability, the traditional electromagnetic detection and chemical detection methods face serious challenges in abnormal discharge detection. In this paper, a practical multispectral detection technology is introduced into free metallic particle discharge diagnosis. According to change the number of the free metallic particle, discharge defects with different risks are simulated. On this basis, phase-based multispectral characteristics such as, phase- resolved partial discharge pattern, pulse phase distribution, and non-phase-based multispectral characteristics such as, time series signal, maximum multispectral pulse intensity, multispectral mean pulse intensity and ternary graph distribution are analyzed to represent the degree of free metallic particle discharge development. Multispectral characteristic analysis provides an effective method to characterize the discharge defects and provides a research basis for the subsequent fine fault diagnosis combined with intelligent algorithms.

The rapid growth of communication and information industry is inseparable from electronic technology. When the semiconductor device with charge is powered on, the charge discharge path forms an instantaneous high pulse to form a conductive path, and the high-energy pulse absorbed by the device pin damages the chip. The main function of biological immune system is to detect, identify and kill non-self-substances called antigens in vivo and in vitro, and its essence is on-line autonomous fault diagnosis and health monitoring in vivo. In this article, a fault diagnosis model of semiconductor devices and integrated circuits based on Artificial Immune Algorithm (AIA) is proposed. By collecting the characteristic data of circuits in various working states, the working states of integrated circuits are analyzed and predicted, so as to realize the Electrostatic Discharge (ESD) protection of semiconductor devices and integrated circuits. The simulation results show that compared with the traditional Particle Swarm Optimization (PSO) algorithm, this method has obvious advantages in the later stage of operation, and the error is reduced by 20.85%. By introducing this model into the design of integrated circuits, the performance of circuits can be optimized and analyzed, the efficiency of circuit design can be improved, and the problem of multi-index conflict in integrated circuits can be solved well.

It is an effective way to improve the utilization rate of equipment, improve the quality and output of products, and reduce the labor intensity of workers by using programmable controller to transform the old equipment. In this paper, programmable control (PLC) technology is used to design the automatic control system of tower crane according to the principle of relay - contact controller and the principle of three phase asynchronous motor, starting from the variable amplitude action, rotary action, takeoff and landing action and the speed change in each action.

The operation of the distribution network mainly depends on the lines, grids, equipment, and customer electricity usage in the distribution network. The safe and reasonable operation of the entire process is the key to the success of the distribution network operation. Analyzing the relevant indicators of distribution network operation can provide decision- making data support for the power supply management department. The monitoring of operational indicators in the distribution network is the fundamental guarantee for the safe, orderly, healthy, and efficient coordination of power grid operations, which can significantly improve the service and management capabilities of business departments. Digital twin technology provides high-precision data base support for operational monitoring and control in the power grid and ensures the real-time and effectiveness of data models based on intelligent collection terminals. Grid operation and regulatory personnel can observe and analyze the operation of power equipment and facilities in reality through digital twin models. And use the basic algorithms of the model to deduce and predict real power data and make plans for potential faults and risks based on the results of the deduction and prediction. This article proposes a distribution network operation analysis method based on the digital twin model. The digital twin model is applied to the analysis and mining of distribution network operation data, and relevant demonstration applications are carried out. The analysis method is optimized and improved based on the application effect.

With the rapid development of computer and internet technology, the degree of power informatization is deepening, which further promotes the growing scale of power network. As a result, while bringing convenience to business, it also brings new network security risks and challenges. In this paper, a new power system security protection system combined with zero trust network is designed to improve the security protection ability of the power system as a whole. The access subject is judged through the control of dynamic access engine, and only the authentication passes can it pass, and the trusted agent encrypts the access traffic at the same time; Access rights are dynamically judged based on policies and attributes, and access control is determined from static access control to dynamic rights; The experimental results show that the transfer learning model in this paper can achieve more than 90% prediction accuracy. This model can authenticate users by analyzing their behavior patterns in real time, and deprive abnormal users of access rights, thus realizing a zero-trust network architecture.

The transfer system of Hongliu Compressor Station of the West East Gas Pipeline two is the only peak shaving hub of the West–East Gas Pipeline network in Gansu Province, which is responsible for the liaison and joint commissioning of the West East Gas Pipeline one, two, and three trunk lines. The original peak shaving high-voltage protection system was designed and produced by Yokogawa Corporation in Japan and was put into operation in 2009. As of 2022, the HIPPS system control unit is about to reach its designed service life, and the performance and reliability of various control components in the control unit have significantly decreased, unable to meet the designed SIL3 level requirements. Moreover, the control system has been discontinued, and on-site operation and maintenance are subject to dual constraints of technology and spare parts, it seriously affects the stability of the joint debugging of the first, second, and third lines, and the upgrading and transformation of the control system is imperative. We plan to use the SIL3 level safety instrument control system TCS-900 series PLC with central control technology to replace the HIPPS system domestically and improve the operational reliability of the interconnection and distribution overpressure protection system.

The plastic case circuit breaker is a large quantity and wide range of basic components in China's power system. Its life test is an important part of the product type test, and also one of the main test items of the quality inspection agency. This paper puts forward a life test device for plastic case circuit breaker, introduces its structure and use method, and provides theoretical basis for quality improvement, performance optimization and reliability research of circuit breaker products, which has important practical significance.

Energy storage plays a very important role in the whole power system. This technology can ensure stable operation of the power grid and improve power quality. In the actual development process, the application of energy storage technology in the power system has gradually become widespread. Through the relevant analysis and research on the application of energy storage in the power system, this paper hopes to provide theoretical support and practical reference for relevant personnel in the actual development process.

Traditional power transmission methods are wired, compared with wired power transmission technology, wireless power transmission technology has the advantage of eliminating the cable required for wired charging, while avoiding the electrical spark that may be generated when wired charging contact. With the development of the electronics industry, wireless charging technology has been applied in various fields. In addition to charging mobile phones, human implantable devices, electric buses, mining robots and other devices, wireless charging technology can also supply power to underwater intelligent devices such as AUV, thus improving the concealment and endurance of AUV. In this paper, the theoretical model of dual-channel UWC (Underwater Wireless Charging) system based on series-series magnetic resonance is explored, the circuit is simulated using Ansys simulation software, and the dual-channel magnetic resonance wireless charging system which can be applied to underwater is designed. The experiments show the built system can charge two loads at the same time, and the maximum DC-DC energy transfer efficiency of the whole system is 88.9%, and the total output power is 256 watts.

To ensure the operation stability of DC Molded Case Circuit Breakers (MCCB), by analyzing the structure of DC MCCB and taking the internal electric field of DC MCCB as the research object, a finite element software Ansys Workbench is used to build and analyze the temperature field simulation model of DC MCCB. According to its internal the distribution of internal temperature and air flow, it is concluded that some plastic parts can hinder the internal air flow and affect the heat dissipation of circuit breaker. The simulation results show that the temperature rise of plastic parts and the maximum temperature of contacts can be effectively reduced by shortening the connecting plate of the upper cover and punching the bracket of the moving contact, so that the internal temperature will not be too high, which enhances the operation reliability of DC MCCB.

The 10 kV distribution line in the power system has the characteristics of complex network structure and scattered load distribution. The frequency of faults in the line is relatively high, especially the single-phase ground fault, which accounts for more than 80% of the distribution network faults and has strong concealment. The most common and destructive fault in the power grid is the short-circuit fault of the transmission line. During the short-circuit, the instantaneous heat released by the arc and the enormous electric power generated by the short-circuit current can cause serious damage to the electrical equipment. In this regard, this article based on an improved decision tree algorithm to locate power system faults and designed a diagnostic system. The research results indicate that as the number of tested circuit boards increases, the prior probability of faults becomes closer to the actual distribution of faults, and the average diagnostic cost of faults is the lowest among the three algorithms. The average search length of the algorithm in this paper gradually decreases, and the downward jump of the curve indicates that the decision tree structure generated by the corresponding algorithm here has been adjusted. The power system fully and reasonably applies the temporal attributes of alarm information, effectively improving the problem of difficulty in re modeling when topology changes.

With the development and development of manufacturing industry, the precision and quality of products are also higher and higher, therefore, the role of CNC machine tools is more and more prominent. The numerical control machine tool is the important component of our manufacturing industry, it directly relates to the development of the entire manufacturing industry. At present, the development of CNC technology in China has shown a high performance, multifunctional, customized, intelligent, green and other development trends. The existing research methods of machining accuracy include error prevention and error correction. Error prevention method is to improve and optimize the structure of mechanical parts, and adjust the temperature of the surrounding environment in the manufacturing process, so as to improve the accuracy of the design and assembly of parts. While ensuring the machining accuracy, a large amount of investment is introduced to reduce the impact of errors. The error correction method is to use an efficient error detection method to measure the error, and then identify the real error value of the machine with the error, and then the same amount, reverse processing, and then use the corresponding software or equipment to correct it, so as to achieve the purpose of improving the accuracy of the machine tool. Compared with the error prevention method, it has been welcomed and popularized by the majority of researchers because of its good compensation effect, low economic cost and no need to optimize the machinery.

In order to address the low accuracy of existing transformer fault diagnosis techniques, this research suggests an unique transformer fault diagnostic model based on an improved whale algorithm optimized extreme gradient boosting (IWOA-XGBoost). First, based on Dissolved Gas Analysis (DGA) of oil, nine gas ratios were created utilizing the no coding ratio approach; Then use the hybrid strategy to improve the traditional WOA; Finally, IWOA is constructed to optimize XGBoost for transformer fault diagnosis. As compared to other algorithms, the accuracy rates of the experimental findings show that the model put forward in this study is superior, increasing by 16.89%, 11.69%, 6.5%, 3.9%, and 1.3%, respectively.

In order to realize the automatic control of electronic air pressure valve, this paper establishes an automatic control system based on STM32f103c8t6 core controller. Aiming at the problems of low precision, easy "step loss" and "stall" of open-loop motor control, this paper designs a closed-loop control algorithm based on magnetic encoder, which greatly improves the control accuracy and system stability, and effectively avoids the problem of motor stall. The system also automatically adjusts the opening and closing of the air pressure valve according to the pressure value measured by the barometric pressure sensor. The experimental results show that the control system can realize the functions of automatic adjustment of multi- pipeline pressure and closed-loop self-test.

Aiming at the conflict between thermal design and maintainability design in aircraft cockpit, a fault diagnosis method of in-cabin environmental control integrated system is proposed. The emphasis is on the electric control system and the motor system of fan compressor. The electric control box adopts independent closed design, which is installed inside the device, in the same environment as the cabin, and placed on the circulating air channel, and the bottom is provided with cooling fins, which is beneficial to the heat dissipation of electrical components. The connection is completed by shielded twisted-pair cable, etc., and the relationship between heat dissipation design and electromagnetic compatibility design must be handled well. The electric scroll compressor is adopted, and the compressor motor adopts high-efficiency permanent magnet motor with centralized winding, which weakens the mutual inductance between phases and improves the motor efficiency. The rotor magnetic steel is made of NdFeB rare earth permanent magnet material, which has the advantages of small volume, high efficiency, large output peak torque, fast response of control parameters, wide speed regulation range and accurate positioning.

Internal meshing is a commonly used mode of meshing between gears. In this paper, we focus on the parametric modeling method for spur gears in Creo 2.0 and how it can be used to design corresponding gear and gear ring systems. The Abaqus 2020 software was utilized to perform finite element simulation experiments on the internal gear-ring system under specified working conditions, allowing us to obtain detailed information on the stress and strain distribution within the system.The parametric modeling method used in this study involves the use of advanced mathematical algorithms to create a digital representation of the gear's geometry, including its teeth profile, root diameter, pitch diameter, and other critical parameters. This digital representation is then used to generate a virtual model of the gear and its associated components, such as the gear ring.By using the Abaqus 2020 software, we were able to simulate the internal meshing of the gear and its associated components under a range of different operating conditions, including high loads, high speeds, and high temperatures. These simulations allowed us to evaluate the performance of the internal gear-ring system under real-world conditions and identify any potential weaknesses or areas for improvement.

Internal meshing, an innovative mechanical transmission mode that enjoys widespread application, high transmission accuracy, and compact structure, is frequently employed within planetary transmission mechanisms. However, the assembly process of these gears inevitably introduces various types of assembled errors. In this research paper, we utilize a parametric modeling technique for gears based on the Creo software to establish three-dimensional models of gears. Subsequently, we employ the ABAQUS software to analyze and solve finite element problems related to the different forms of assembled errors. Our objective is to investigate the impact of these errors on the stress and strain of gear components, particularly the thin-walled gear rings in transmission processes. Furthermore, we also evaluate the performance of these gear rings under varying carburized layer thicknesses. Through our comprehensive analysis and examination, we aim to provide valuable insights into optimizing internal meshing systems and enhancing their structural integrity and reliability.

Based on iSIGHT optimization platform and ABAQUS, the frames of rocket liquid engine with multiple forms are optimized. Fast parametric modeling method, sensitivity analysis method and multi-objective genetic algorithm NCGA are combined to optimize the dimensions of frames. The frame modeling of different forms can be accomplished by using specified frame design parameters, thus greatly improving design efficiency. Finite element analysis and optimization were carried out on a single four-point pusher frame as an example. The influence of different section size parameters on the frame performance is analyzed. The optimized structure of frame reduces the amount of material used and increases the first order natural frequency within the constraints.

Since the advent of automobiles for more than one hundred years, especially since the mass production of automobiles and the great development of automobile industry, automobiles have exerted immeasurable influence on the development of the world economy and the modern life of mankind. With the continuous improvement of China's economic construction level, all walks of life have entered a new development stage, and the automobile design and manufacturing industry has presented new characteristics. As the key component of automobile powertrain system, the reliability and durability of transmission directly affect the overall function of automobile. In order to further improve the stability of automobile structure and give full play to the advantages of stepless transmission, it is necessary to optimize and improve the disadvantages of mechanical transmission mechanism of automobile at present. In this paper, the CAE analysis software is combined with the traditional analysis method, and the selection of transmission, the matching of speed ratio and the optimization of speed ratio are comprehensively considered by CAE analysis software from the perspective of vehicle power and economy, so that the structure of the compound inner raceway of the rotor-disk group of automobile stepless transmission is optimized.

In this paper, a manual alignment device and control system mode of field self-propelled machinery are proposed, including flexible contact induction device, information acquisition module connected with whisker sensor, microcontroller controller connected with information acquisition block, steering device connected with a single chip microcomputer, motor, remote communication and remote control device, and grating, The front end of the flexible contact induction device also contains two cross direction contact sensors. The data acquisition device collects the resistance data of the crops touched by the contact sensors and transmits it to the microcontroller controller. The single chip microcomputer can use the received data to control the steering speed of the steering gear controller and the speed of the motor in the device, and finally complete the manual alignment with the operating vehicle, manual speed regulation, and the obstacles in the front can retreat automatically, and it can remotely control the field intelligent operation vehicle platform, as well as various other external equipment.

Firstly, through the analysis and introduction of the ship electric propulsion system, the superior performance of the electric propulsion system is explained, and the problem of harmonic pollution in the electric propulsion system cannot be ignored. Secondly, a variety of harmonic control methods are studied and analyzed from the basic problem of harmonics, among which shunt active power filter is a good method. Then MATLAB/Simulink is used to build a shunt active power filter model based on the ship electric propulsion system, and its effectiveness is verified. Finally, it is proposed that the shunt active power filter can suppress the harmonic interference well and make the Marine electrical equipment run normally.

To meet the requirements of on-demand dispensing and precise application, improve the level of precise application of pesticides, and realize the protection of land, an intelligent variable application system with a small flow rate was designed, and an intelligent variable application controller was developed. The system can accurately measure the speed of operating machines and tools, with human-computer interaction function, can manually input acres spray volume, display real-time flow, can be real-time control according to the speed of operation. The mapping model of operation speed and dosage was established. The variable parameter incremental PID control algorithm was used to control the flow rate, and the first-order filtering method was used to filter the velocity sensor data. The experimental results show that the small-flow intelligent control system has better control accuracy and responsiveness under different working environments, which provides a reference for precise variable flow application control under the condition of small flow.

As an indispensable equipment intelligent manufacturing production line, industrial robots are an important symbol of realizing the intelligence of production lines and play an important supporting role in accelerating industrial automation production. In order to solve the problems of low production efficiency, low degree of automation and complex production lines of traditional production lines, taking the manipulator in the production line as the research object, based on the conceptual design and control simulation technology of mechatronics, the twin drive theory was analyzed and refined. The manipulator drive control program was written, and the virtual simulation of the manipulator driven by data was carried out through signal mapping connection. Based on virtual design and simulation technology, combined with the production line operation process and data-driven manipulator model, the logical operation relationship of the production line simulation model was established. By setting parameters such as sensors and motion control signals, the working state of the intelligent manufacturing production line was simulated, and the automatic assembly process of pen holder was realized.

For the fast-steering mirror system used for beam pointing control, in order to ensure that the fast-steering mirror can work stably under high bandwidth, this paper proposes a design method of the current loop drive system for fast steering mirror. The performance of the proposed method is verified experimentally. In the proposed design method, the current loop control is realized with hardware circuit, which ensures the stability of the fast-steering mirror system under high bandwidth conditions. Linear amplifier is used to meet the requirements of low ripple and high accuracy. In the design, both simulation and experiments are carried out. The hardware parameters are obtained through software simulation, and the parameters are adjusted through experimental test to achieve the ideal state. The experimental results show that the drive circuit system based on hardware current loop control can achieve accurate current control performance with high bandwidth and low ripple, which meet the requirements of high-speed response and stable control of the fast-steering mirror.

To support the implementation of smart grid policy, the steady-state temperature distribution of a dry-type air-core reactor is studied by a multi-physics field coupling method based on finite element method simulation software. Based on the simulation results, the one-dimensional distribution curves of the reactor encapsulation temperature along the axial and radial directions are obtained, and then the location of the temperature sensor is preferentially selected. Furthermore, based on the temperature rise test of the reactor entity, the experimental measurement results are compared and analyzed with the simulated calculation results. The results demonstrate a strong correlation b etween the temperature rise of the encapsulation obtained from the experiments and simulations. The above research results can provide valuable insights for the optimized design of dry-type air-core reactor based on temperature field.

Oil-immersed type transformer is an essential part of electric power system, and the hot-spot temperature of the winding is closely related to the oil temperature at the top of the transformer. In this paper, modeling of the thermal circuit based on the distribution of the top oil temperature of the double-wound transformer. On the basis of the consideration of nonlinear characteristics of thermal resistance of insulating oil, we also considered the influence of oil stickiness and the thermal capacitance of transformer components (iron core, insulating oil, windings) with temperature changes, and introduced the cooling ways and winding loss correction factors in the model. The Runge-Kutta method is selected to resolve the model parameters, which further simplified the calculation process. Above all, the improved model of hot- spot temperature of transformer winding is proposed. Contrast with testing transformer measured data in laboratory and calculation result of traditional hot spot analogy method, the model shows good consistency in the light load (around 20% and 40%), which can well describe the transient temperature change and propose a new method to calculate the hot- spot temperature of transformer winding.

Traditional network measurement methods can't obtain fine-grained real-time network state information, and traditional network devices are not programmable. Therefore, in order to realize network-wide detection, it is necessary to design a high-level mechanism to configure multiple INT (In-band Network Telemetry) detection paths. In order to solve the above problems, this paper studies the INT detection path configuration strategy based on adaptive algorithm and P4. In this paper, an optimal detection path configuration algorithm is proposed. In the IP layer, an adaptive algorithm is proposed to predict the traffic trend and adjust the Sel-INT (selective INT) sampling rate. At the same time, in the EON (elastic optical network) layer, the out-of-band optical link polling is decided by using INT, and P4 is used to check the consistency of data plane forwarding behavior. The research results show that, when the number of switching nodes on the forwarding path is the same, the single-path verification time of P4 in different network topologies is basically the same, and the error range is less than 3%. For a given topology diagram, the more vertices in the set. Therefore, correspondingly, the algorithm execution time has a certain decline. The simulation results verify the effectiveness of the INT detection path configuration strategy, and they all have a certain guiding role in the actual deployment.

The physical types of equipment in the distribution network are diverse and the quantity is huge. This is because the distribution network requires a large number of various types of power supply and other auxiliary equipment such as substations to provide power supply services for electricity customers, resulting in a very rich and diverse data resources in the distribution network. Moreover, with the continuous integration of new energy sources, the distribution network has become increasingly complex, not only increasing the dimensions of operational management, but also facing huge challenges in data resource management. This paper proposes a unified management method for power distribution and new energy data resources, which manages a large number of non-digital data and unstructured data such as file text in the distribution network in a unified manner, integrates heterogeneous data from different business systems, solves data quality problems caused by management, data processing and other reasons, meets the data management needs of power grid enterprises, and promotes digital transformation of enterprises. On the basis of achieving unified management of data resources in the distribution network, the deep mechanisms between data resources are explored to reveal the operational laws and characteristics of the distribution network. At the same time, data analysis algorithms are used to mine distribution network data, design and implement distribution network business analysis functional modules, and upgrade distribution network data management from digitalization to intelligence.

To cope with the problems of insufficient level of intelligence in attempting combat missions under tactical edge conditions, high command, and control delay, and difficulty in sharing data across domains, the joint intelligent tacti-cal edge cloud service model is proposed, the definition and formal description of joint intelligence are given, and the synergistic mechanism of the joint intel-ligent cloud service model is analyzed, to provide theoretical references for distributed combat at the tactical edge under intelligent conditions.

Based on the teaching method of "cloud classroom+integration of theory and practice", the curriculum group system of electrical and electronics is constructed according to the cloud platform. According to the training goal of electromechanical talents in our school, the main courses of electrical and electronics are divided based on the cloud platform, and the practice links of electrical and electronics courses group are run through with the help of "skill competition+subject competition". By using cloud classroom, mathematical knowledge and professional knowledge can be integrated into the teaching process, and the mathematical knowledge points involved in the courses of electrical engineering and electronics can be strengthened to improve the theoretical level of students. It aims to cultivate high- quality and high-level skilled talents who can understand both theory and practice and improve teachers' professional level and teaching efficiency.

The purpose of civil aircraft maintenance department is to ensure the airworthiness and safety of civil aircraft to perform various flight tasks, but the investigation of various mechanical or electronic failures of aircraft greatly limits the development of maintenance department. In recent years, civil aircraft accidents have occurred one after another, resulting in casualties and huge losses of property. A large part of the reasons for the accidents are attributed to the failure of civil aircraft. As a typical comprehensive and complex system, the complexity of its structure and working principle leads to the complexity of its fault symptoms and the relationship between faults and fault sources. In this paper, big data technology is applied to the airworthiness review model of civil aircraft to realize intelligent diagnosis of civil aircraft state. The simulation results show that the proposed civil aircraft fault location algorithm has strong information fusion performance and has significant advantages over the traditional Genetic Algorithm (GA) in terms of false alarm rate and false alarm rate. Therefore, the model can be effectively combined to analyze the aircraft state and failure reasons in the airworthiness review of civil aircraft, and the quality management and safety management strategies of civil aircraft maintenance can be put forward according to the actual situation.

In recent years, with the rapid development and expansion of China's economic construction and opening up, the aviation industry has gradually become the main mode of travel and transportation in China. Maintenance efficiency is one of the important factors that affect the maintainability of civil aircraft, and it is also an important part of the maintainability outline and maintainability design criteria of civil aircraft and has now become an important content of product maintainability design and analysis. This paper attempts to analyze the maintenance efficiency of civil aviation aircraft based on DL (Deep Learning) and constructs a risk assessment model in this paper. Human reliability analysis is to quantitatively analyze the human error probability that may exist in normal operation and accident state of the system and evaluate the risk. Finally, the experimental results show that the error rate of this algorithm in aero-engine fault data diagnosis is 89.6885%, which is better than the other two algorithms. And the model of the algorithm in this paper at the best diagnosis rate is preserved. In this paper, DNN (Deep Neural Network) theory of product maintainability under specified time and conditions is related maintenance according to agreed procedures and methods, and its purpose is to restore the related capabilities of products under specified conditions.

Budget management is the most important aspect of government financial work and a prerequisite for fiscal expenditure. Budget management is an important means of implementing national financial policies and regulations, effectively organizing fiscal revenue and expenditure, reasonably adjusting funds, and achieving national functions. The full life cycle management of financial budgets includes business processes such as budget preparation, budget execution, accounting, final accounting, and reporting. To achieve full life cycle management of fiscal budgets, it is necessary to unify the management of all data involved in fiscal budgets. However, in reality, fiscal budget data comes from a wide range of sources, making it difficult to achieve a single source of data. Therefore, it is necessary to design a unified digital model. Establish a comprehensive budget full chain data management system by connecting the data interface with multiple adaptive attributes in the model to the source system of financial data. In response to this issue, this article proposes a digital model construction method that is suitable for the full life cycle management of fiscal budgets. Through the feedback results of data extraction, it in turn promotes the continuous improvement of basic information of various fiscal expenditure units. At the same time, the construction of this model can provide effective solution support for account real matching and account card matching work in the provincial financial budget management platform.

The quality status of aircraft use involves multiple indicators and has nonlinear correlations. The commonly used traditional evaluation methods often rely heavily on expert knowledge and cannot depict the nonlinear relationship of indicators, making it difficult to accurately and effectively support mission aircraft usage decisions. This article constructs one mission aircraft usage quality status indicator system that includes annual usage plan execution rate, matching degree of aircraft lifespan usage, variability of the remaining lifespan echelon difference, rate of satisfaction with aviation materials supply, and aircraft failure free rate. Wavelet Neural Network and BP Neural Network models are respectively used to predict and evaluate the quality status. Case validation shows that Wavelet Neural Network have faster data learning speed and higher prediction accuracy compared to BP Neural Network, which are basically consistent with actual results and can provide more accurate quantitative basis for mission aircraft usage decision-making.

In the paper the Android runtime (ART) program language and its interpreter are designed based on Java. The grammar rules of ART language are designed in detail from lexical analysis, grammatical analysis and semantic analysis, and a simple but powerful scripting language is implemented, which supports object-oriented programming, functional programming, operator overloading and other functions. The ART interpreter is written based on the Java standard library. The experimental results show that it can be easily embedded into most JVM based applications including the Android platform. It improves the current problem of too few simple and easy to use scripting languages and relying on third-party library data.

Earthquakes are a major threat to the conservation of cultural relics, and how to accurately assess the impact of earthquakes on cultural relics in collections has become an important issue in the field of conservation. Traditional methods for assessing the seismic damage of cultural relics are mostly based on experimental data or structural dynamics theory, which cannot truly reflect the seismic damage of cultural relics. This paper proposes a method for analyzing the impact of seismic damage to cultural relics based on the fusion of multi-source information. Firstly, this paper is oriented to the real earthquake damage historical data of cultural relics using two-way long and short-term memory network to obtain the characteristics of the earthquake damage data and using convolutional neural network to obtain the local characteristics of the data, and then using classifiers to obtain the entities and relationships contained in the earthquake damage data. Secondly, based on the obtained dataset, through multi-source information fusion and correlation analysis among the data, the seismic impact dataset of the collection can be used for the research of seismic impact analysis of the collection. Finally, relying on the knowledge of seismic impact of cultural relics, we construct a mapping of seismic impact factors of cultural relics in the collection. This provides data support and basis for subsequent research on the quantification of the impact of seismic damage to cultural relics in collections and research on the risk assessment model for seismic damage to cultural relics in collections.

Robots need to move and operate autonomously in a constantly changing field environment, so having the ability to sense the field becomes especially critical. In this paper, we use binocular vision technology and depth cameras for autonomous localization, combined with image semantic segmentation and target prediction to predict the location of moving objects in the field and give the state of the environment. In this way, the robot is able to achieve fully autonomous path planning and operation to meet the needs of autonomous robot operation in different scenarios.

This project will study a Digital Surface elevation Model (DSM) and an image semantic segmentation method based on multi-scale weight clustering. Taking the overall partitioning metric as a constraint, the weights are classified step by step using a multi-grid algebraic algorithm. An image processing method based on a single residual method is proposed for high-resolution remote sensing images. Row Height Information Mining with Convolutional Neural Networks is proposed. The method's accuracy can reach 92.33% through the test on the existing public database. The experimental results prove that the method extracts and fuses features of multiple levels and can obtain better semantic segmentation results.

This project intends to combine the U-Net algorithm and the super pixel algorithm to study the high-resolution remote sensing image semantic segmentation algorithm based on the U-Net algorithm. A pixel-based end-to-end semantic segmentation method is proposed. By extending the original dataset, a dual classification model is built for each type of ground object. The final semantically segmented image is obtained by combining individual predicted sub-maps. The rapid identification of remote sensing images such as agricultural land, houses and rivers based on multi-band SAR images is completed using small-sample SAR images. The overall classification accuracy is 95.7%. Experimental results show that this algorithm can adapt to various environments and handle complex semantic information well.

The data science industry has experienced significant growth, highlighting the need for a comprehensive understanding of global salary dynamics. This paper aims to bridge the gap in recent research by providing valuable insights into wage forecasting in the data science industry while enhancing prediction accuracy. Utilizing the most recent dataset collated in 2023, rigorous preprocessing techniques ensure data integrity and relevance. Descriptive statistics summarize key dataset characteristics, while correlation analysis investigates variable relationships. Visualizations such as bar charts, box plots, and scatter plots effectively depict data distributions and patterns. Employing a multi-layer neural network (MLP) algorithm as the primary predictive model, the performance of neural network is compared with multiple linear regression and gradient boosting algorithms. The experimental results show that the superiority of MLP in terms of fit and effectiveness in predicting data science wages. The main contribution of this research lies in the application of deep learning methods, significantly improving prediction accuracy. This study holds significant value for data science professionals, empowering them with a valuable tool for salary negotiation and career planning.

The effectiveness of inspection tasks performed by unmanned helicopters during underground inspections is directly influenced by the performance of attitude control. Therefore, it is crucial to optimize the attitude control of unmanned helicopters. Spike neural membrane computing offers Turing computing versatility and high efficiency in various time collaborations. This paper presents a study on attitude optimization control of unmanned helicopters in underground coal mines using spike neural network computing. The approach involves constructing an extended spike neural membrane system based on the attitude dynamics model of an unmanned helicopter, with the aim of further optimizing the system and algorithm. The experimental results demonstrate that the optimized spike neural membrane system is capable of effectively achieving attitude control of the unmanned helicopter under normal conditions. Furthermore, it can successfully maintain the attitude angle within the desired range, even in the presence of wind interference. Additionally, a comparison with traditional sliding mode control and the Designed Optimization Model (OSNMS) reveals that the OSNMS proposed in this paper significantly enhances the attitude control performance of unmanned helicopters.

For many domestic supply chain information systems, most of them are imperfect in details, incomplete in functional modules, and lack of convincing solutions in system integration. This paper develops a foreign trade supply chain information system based on fuzzy clustering algorithm. Therefore, this paper improves the fuzzy clustering algorithm and applies it in supplier selection. The main idea of the improvement is divided into two steps. The first step is to cluster all objects in a small range by hierarchical clustering to generate initial clusters. The second step is to find out the cluster with a large number of objects in the initial cluster, compare the sum of distances obtained by various selection methods, and choose the one with the smallest sum of distances as the best choice. The research results prove the feasibility of the algorithm. The calculation time of the improved algorithm is not much different from that of the simple hierarchical aggregation algorithm, but the clustering effect is not as good as that of the improved algorithm. The main reason is that the hierarchical aggregation algorithm is irreversible and the operation of merging clusters cannot be corrected, which may lead to the low quality of clustering results. The improved algorithm overcomes this shortcoming, thus making the calculation result more accurate.

Intrusion detection is an advanced security technology that goes beyond traditional measures like access control, passwords, authentication, auditing, and firewalls. It safeguards against internal and external attacks, as well as misoperation. To establish a user's normal behavior, a set of behavior profiles must be created and maintained, describing the user's typical system usage. This paper presents a visualization algorithm for network intrusion scenes, employing a dual-engine structure that integrates network-based and host-based intrusion detection systems. The algorithm offers combined detection, reporting, and response capabilities. Simulation results demonstrate significant improvements in false alarm rates compared to traditional methods. The network engine implementation combines protocol analysis and pattern matching, narrowing down target matches and enhancing detection speed. The approach effectively meets the efficiency and accuracy requirements of intrusion detection systems in cloud computing environments.

Many universities adopt information management systems based on C/S architecture. Although this kind of management system can realize information processing and sharing in a certain range, it cannot share information effectively with the expansion of the scope, which has great limitations. In this paper, a mixed architecture of C/S and B/S is adopted to design and implement a college student information management system based on teaching resource sharing, which is realized based on Web technology, and an improved server flow scheduling algorithm combined with proxy cache prefix is used to maximize the utilization of system resources. In the design and implementation of this system, many efforts have been made in reusability, maintainability, upgradeability and stability. In addition, all links should get enough documents and technical information, which can make the handling of matters more convenient. The function and standard of this system can meet the reality.

With the rapid development of the Internet, the health big data industry is highly valued by society and the state, and how to effectively mine the large amount of health data generated by medical enterprises has become a current research hotspot. Neural networks, as a new network technology, can be well combined with health big data, and thus promote the deep mining of health big data. Therefore, based on the analysis of the characteristics of Convolutional Neural Networks (CNN), the study combines it with the Apriori algorithm and constructs a CNN-Apriori based enterprise big data mining model. The results show that the CNN-Apriori model takes 0.813s to converge to a steady state and has a fitting accuracy of 96.012% at a running time of 0.900s. In the mining of factors related to users' physical fitness, the users' physical fitness score improved most significantly from 90.023 to 96.321, and all the users' physical fitness scores obtained by the method improved to above 90. This shows that the proposed CNN-Apriori model can effectively mine the factors related to users' physical quality, which has good application effect in the big data mining of medical enterprises and has certain promotion value.

With the rapid development of computer Internet technology, network communication has become an important way to obtain information. However, the open Internet is also vulnerable to all kinds of malicious damage and attacks. Intrusion monitoring technology can realize the monitoring, analysis and defense of intrusion behaviors, but the current intrusion monitoring technology cannot meet the development needs of current network security. The study adopts decision tree algorithm for information classification model design and combines k proximity algorithm to construct network intrusion classification model design. The experimental results show that the decision tree-k proximity algorithm has the highest correct detection rate, with a maximum accuracy of 94.36% and strong monitoring capability; the training time of the model is short, taking only 5 s when the training is completed; at the same time, the recall rate of the decision tree-k proximity algorithm is high, and the area enclosed by the subject's characteristic curve is large, with better comprehensive performance. The data mining network intrusion monitoring model based on decision tree and k proximity algorithm can accurately monitor malicious attacks and ensure system security, which is important for maintaining computer network security.

This paper introduces the flutter test technology, flight technology, data analysis technology and some flight test results of transportation aircraft based on natural flight atmospheric environment. The flight test shows that although the amplitude of structural response signal obtained by using this kind of excitation is small, the accurate identification of dense modes of small amplitude signals is realized based on the vibration mode analysis of multi-input and multi output system. Based on Metadata Description Language (MDL), the task description model, requirement description model and equipment configuration description model of flight test are constructed to realize the standardization of task transfer and configuration implementation process, thus achieving the automation, intelligence and standardization of flight test task development process, greatly improving the test efficiency and ensuring the accuracy of test data. The flight test verification is carried out in combination with the qualified flight test of the aircraft, and the compliance verification results of the flight test of the aircraft flutter characteristics are given. The results show that this method is suitable for certified flutter flight test of transport aircraft.

Guided by the technical idea of modern data warehouse (ODS), focusing on the characteristics of unstructured big data of flight test data, facing the requirements of model engineering for unified management and efficient processing of flight test data, and aiming at the granular comprehensive management of flight test big data, it provides a basis for the establishment of on-line management and analysis system of flight test big data based on ODS and meets the requirements of model engineering for efficient analysis of flight test data. In order to test the stability of the landing gear system, the landing gear shimmy test was carried out on a real airplane for the first time in China. By installing an excitation board on the runway, the relevant modes of landing gear shimmy were effectively stimulated. The modal single peak isolation combined with the least square algorithm is used to analyze the experimental data through modal parameter identification. The results show that there is no shimmy in landing gear in flight test, and landing gear is stable in the test.

Social media is a platform for the participation of the whole people, and the "COVID-19" is a topic closely related to the whole people. Based on this, this paper conducts sentiment analysis on the hot topic of social media network public opinion "COVID-19", which can well present the public's emotional evolution process and results of this public opinion event, so as to provide certain guidance for relevant national departments to guide and control public opinion. This article chooses Twitter as the data source platform, collects tweets related to the "COVID-19" in January this year, and builds a simple public opinion analysis system using natural language processing models and different machine learning methods. The research results show that for the topic of "COVID-19", which is a long-lasting topic that is closely related to everyone, the proportion of positive emotions is relatively large, which also protects the security of the country from the level of public opinion. In other words, for a hot topic, when negative emotions prevail, relevant departments need to intervene in time to supervise, guide the direction of public opinion, and maintain social stability.

In the era of rapid development of the Internet, interactive platforms such as Weibo, Zhihu and Douban have developed very rapidly. The appearance of each hot event will bring the heated discussion of netizens, showing all kinds of netizens' emotions. By analyzing the emotional data, we can quickly obtain the positive and negative emotions of netizens, which can provide technical support for the marketing of commodities and the prevention of cyber crimes. This paper designs a netizen emotion analysis and visualization system based on social data by using platforms such as Douban and Weibo. This model takes the novel coronavirus outbreak event as the case for analysis, uses data mining, machine learning, data visualization and other methods to mine potential important information in the data, and uses the clustering method to analyze and classify emotions. Quickly obtain the positive and negative emotions of Internet users, study the emotional changes of Internet users under different states of the epidemic, and conduct correlation analysis between the epidemic events and the emotions of Internet users. It will help governments at all levels to accurately grasp the situation of online public opinion and conduct public opinion guidance and epidemic prevention and control work.

Search and Rescue (SAR) is one of the key supports for ensuring combat effectiveness, so it is of great significance to successfully conduct SAR actions to improve the operation capability. Due to the characteristics of complexity and multi-level combination of SAR, the analysis and design should pay attention to the following aspects: first, the high precision of a single capability indicator of the equipment platform is no longer the highest requirement, but the collaborative emergence of the comprehensive capability shown by the whole System of Systems (SoS) is more critical. Second, the personnel SAR in threatening environments is a large-scale, complex operation task, multi-equipment coordination required and multi-level joint SoS. Therefore, we need to take capability as a traction based on perspective of system engineering and carry out the structure modeling of personnel SAR SoS through the unified structure framework. Third, the personnel SAR SoS owns uncertainty and hierarchical emergence. Which modeling and simulation method should be selected to model and analyze the SoS from multiple perspectives, and how to combine simulation and Virtual Reality (VR) technology to play a role in assisting decision-making, so as to effectively improve the SAR capability. Therefore, this paper mainly carried out the following researches: (1) proposed the SoS design method based on EMCE (Environment, Mission, Capability and Elements) loop, and designed the personnel SAR SoS from four aspects: environment analysis, mission analysis, capability space planning and SoS elements analysis; (2) Based on DoDAF (Department of Defense Architecture Framework), the architecture of personnel SAR SoS was modeled, and the operational concept, operational nodes and information exchange, operational activities, state transition and event timing of the SAR SoS were designed in detail. (3) In view of the hierarchical emergence and simulation runtime of the SAR SoS, a composite modeling method adapted to the SAR SoS was proposed by taking advantage of the functional complementarity of the multi-Agent modeling method and DEVS (discrete event system) modeling method. (4) Using the holographic demonstration function of MR (mixed reality) equipment and the advantages of allowing users to interact with virtual objects and real objects, so as to assist commanders to better understand the virtual and real battlefield environment and make corresponding decisions.

Traffic accident not only causes casualties, but also generates traffic congestion, resulting in severe traffic delays and other indirect economic losses. Based on the accurate identification of accident road section attribute, considering the accidental traffic impact is very important to improve the comprehensiveness and accuracy of road location safety evaluation. In this paper, OBD data is used to map matching, the fuzzy clustering algorithm is used to discriminate and classify the road traffic operation status. A kind of identification algorithm of the accident road section traffic attribute impact is proposed, which is used as case analysis of Beijing Fourth Ring Road. The analysis results show that the accident occupying lane is the most important factor of traffic delay, the more lanes are occupied, the more vehicles of single lane are delayed and the delay time is longer.

The performance of most existing dehazing methods are limited by the independency of the transmission map estimation and atmospheric light. To ameliorate this, we present an algorithm based on both GANs models and atmospheric scattering. We first enhance the accuracy of transmission map estimation by constructing a multi-scale FC-DenseNet transmission map prediction network. To effectively avoids the issue of calculation errors caused by inaccurate prediction of the transmission map, a shallow residual dehazing network is also developed, which shares the learned weights of the transmission map prediction network. A multi-task loss function is constructed that dynamically optimizes the two networks simultaneously. To augment the generation of realistic images, a classical WGAN is used to fine-tune the generated transmission maps and dehazed images. The results demonstrate that our method has achieved significant improvements compared with the state-of-the-art (SOTA) methods.

Aiming at the difficulty of human material distribution under the background of normal epidemic prevention and control, a new pneumatic robot design scheme was proposed which can remote control. The robot structure was designed by using SolidWorks 3D software, and a prototype was made through Simulation analysis of key parts. The test results show that: The size of the robot model prototype is 600×600×600mm, the extension size is 1200×1200×1000mm, and the maximum weight is no more than 35KG. The test has proved that the epidemic prevention and control robot can be applied in the fields of material handling, building patrol, epidemic control and so on. This design for the robot intelligence has a certain promotion and display function.