Human-in-loop Based Space Teleoperation-Oriented Research On The Operator's Hand Behavior Sensing And Recognizing Technology

Nowadays, with more and more space exploration, a large number of spacecraft on a variety of tasks are sent into space. Because of the particularity for space environment, unexpected failure happens during spacecraft orbit operations, which is inevitable. However, it is a risk to rely on astronauts to repair extravehicular in space, and autonomous space robot also cannot meet the demand, especially in unknown environment. Therefore, the human-in-the-loop based space teleoperation gradually become an important way to realize tasks of on-orbit service. By using operator's behavior to guide the actuator in the space, space operation tasks can be finished by man-machine combination. And the sensing and recognizing of the operator's hand behavior should be one of the key technologies to realize the tasks of the human-in-the-loop based space teleoperation.This dissertation aimed at the problem of the operator's hand behavior sensing and recognizing, systematically studied the extraction of the s EMG(surface electromyographic signal) and the recognition method of the hand behavior based on the s EMG. The main study is as follows:1. In view of the features of s EMG for human body, which is weak as well as susceptible to environmental interference, we set up acquisition and storage experiment system, realizing the effective collection of the s EMG.2. To determine the onset of the operator's hand behavior, we present a detection method based on TKE operator, which can pick up the effective s EMG of hand behavior accurately. The simulation results show that the using of TKE operator is an appropriate way to judge the onset of the hand behavior, which laid the method and theory foundation for real-time recognition of the gesture.3. In order to make the best of s EMG information, we propose a method of hand movement pattern recognition based on s EMG hierarchical classification. The method can reflect the characteristics of EMG from multiple sides through the introduction of s EMG hierarchical classification. So we can identify more gestures by information from less electrode electromyography. The experiments show that the eight common hand movements can be recognized effectively by two surface EMG electrodes. It performs better than the method without hierarchical classification. A new train of thought is presented for hand behavior recognition based on s EMG.4. For the problem of a different operator or the same operator in different state, the EMG level is often different. we propose a recognition algorithm of adaptive threshold, established an effective mechanism to adjust the threshold which can initiate action, according to the operator real-time electrical level. The experimental results show that on the basis of the recognition algorithm, the operator's hand clenched recognition rate can be significantly improved.5. Based on the experiment system for the gas floating platform of microgravity on the ground, we design an experiment scheme of the space robot to capture the ground target, including research on hand behavior perception and recognition. Operator's hand atcions are shown in mechanical hand repeatly in space, which puts forward to the experimental verification for methods in the dissertation.Above all, the dissertation studied the technology for operators hand behavior sensing and recognizing of space teleoperation-oriented, carried out detecting the onset of EMG based on hand behavior, hierarchical classification of hand behavior recognition, adaptive threshold algorithm research, and authentication based on ground experiment system, which is important for space robot teleoperation task in both theoretical and engineering.