|
Average satellite lifetime in orbit has been gradually increasing since the advent of the space industry. With the increase in satellite lifespan on orbit it is becoming increasingly economically advantageous to refuel and conduct in-orbit servicing rather than launching new satellites. This presents a challenging problem of automatic docking for refueling. Traditionally, docking mechanisms have used either a physical switch, a force sensor, a torque sensor, or a combination of the three. Traditional docking verification techniques using those sensors are suboptimal for use in an orbital servicing satellite as customer’s satellite mass can vary in a wide range while docking velocity also varies. In addition, the refueling system produces a complex pattern of mechanical signals during docking, which is challenging to classify. For this reason, an SHM system with small unobtrusive piezoelectric sensors was proposed to identify and characterize satellite docking. It was decided that the passive monitoring of the docking is preferable overactive methods not to interfere with satellite dynamics and reduce power use. The mechanical waves resulted from the satellite’s docking momentum annulment, thrusters, electrical motors, and mechanical component deployment were passively monitored using an array of piezoelectric wafer sensors. Features in mechanical signals corresponding to the docking were distinguished from other mechanical events normal to satellite’s operation. An algorithm was developed that utilize features specific to docking to classify the quality of the docking engagement including potential false positives and misalignment issues. This algorithm was embedded in a real-time microprocessor which was used to capture passive ultrasonic signals and run the associated data analysis algorithm. Experiments conducted on a laboratory scale docking imitator suggested the applicability of the proposed approach and verified performance on the data acquisition and classification system on exemplary signals.
|
