Research On Robot Arm/hand Teleoperation Toward Satellite On Orbit Servicing

Using space robots to take the place of astronauts in space operations, not only the astronauts can avoid operating in the harsh space environment, also can reduce costs, improve the efficiency of space exploration activities. Under the constraints of the key support technique of computer, control, artificial intelligence, and mechanism, completely autonomous robot working in the space environment now is hard to achieve. Therefore, the practical way is to use human intelligence to conduct high-level mission planning and command design, using ground teleoperating space robot manner to complete space operations. In connection with on orbit satellite maintenance tasks, a ground demonstration platform for teleoperation has been established, virtual reality based and Smith predictor based teleoperation of space robots have been studied.Teleoperation robot arm/hand system model has been established first in this article. Modeling is divided into two parts: the geometric model and mathematical model. In the geometry model, this paper has used Pro/E modeling, used VC++ and OpenInventor to assemble the model. For the mathematical model of the robot system, this paper has analyzed the kinematics and static. Kinematics is the basis of robot control. Forward and inverse kinematics of space robots and dexterous hand has been analyzed in detail in this paper. Static is the basis of analysis for robot interaction force. Force Jacobian by using structural differential transformation has be given in this paper.A ground teleoperation demonstration experiment platform of satellites self servicing on orbit was established. Human-machine interface based on virtual reality has been designed, Connecting a variety of teleoperation equipment into the experimental system, robot teleoperation system based on network communication was established, artificially increased the delay so that through the network, operator can simulation the ground-space large time delay teleoperation.Predict display based on virtual reality technology effectively overcomes the affect of large time delay on the teleoperation for the visual presence. A brief introduction of virtual reality technology, followed by a pin hole assembly experiments. Experiment used the robot geometry model to achieve the forecast, eliminates affect of delay for the visual presence of operator; in the virtual environment, used the feedback data of the actual robot position to rebuild the feedback robot virtual scene effectively, overcome the low bandwidth streaming video on the smooth limit, so that the operator visually to the actual state of remote sensing.Using the control structure of Smith predictor and online identification, effectively overcome the large time delay on the teleoperation of force telepresence. Firstly, a detailed analysis of the cause of instability under two-way delay teleoperation.And then establishing a common environmental model, given the parameter identification method, combined with drawing spring experiments on model identification method has been simplified. Using Matlab to simulate the stability of control structural,analysis the teleoperation system's transparency. Finally, test the tension spring experiment.Control structure based on Smith predictor and online identification effectively overcomes the affect of large time delay on the teleoperation of force telepresence. Establishing a common environmental model, giving the parameter identification method, combined with drawing spring experiments, model identification method has been simplified. Detailed analysing the cause of instability under two way delay teleoperation. And then using Matlab to simulate the stability of control structural, analysis the teleoperation system's transparency. Finally, test the drawing spring experiment.