Basic Experiments of Satellite Diagnosis Using Space Robots with Acoustic Emission Sensors
Toshiaki Iwata, Yusuke Nishimura, Mikihiro Sugita, Kazuo Machida · AIAA Guidance, Navigation, and Control Conference and Exhibit · 2003
A diagnostic method and the visualization of the diagnostic information using a space maintenance vehicle with a pair of manipulators (robot satellite) in orbit are proposed. Recently, space debris has been increasing yearly. We need a system that treats satellites from the beginning of the operation to the end, and have proposed a constellation system that uses the robot satellite for accommodating the mission satellites. The robot satellite has many tasks with the diagnosis of the mission satellites being one of the most important. We propose that data mining techniques such as decision tree construction be used to diagnose mission satellites by means of an acoustic emission sensor. As an application example, a motor position is inferred from the acoustic wave distribution, and we succeed in finding the correct position with approximately 80% certainty. Moreover, as the robot satellite has to cooperate with an expert on the ground, a human-robot interface, such as visualization of the information to diagnose the mission satellite, is indispensable. We discuss the visualization methods for a human/robot interface, including the robot operation panel, original sensor data and processed sensor data, and show an example of the visualization.