No Base-Disturbance Path Planning And Neural Network Predictive Control Of Space Robotic System

This text discussed the problems of path-planning and dynamics control of free-floating and free-flying space robotic manipulator under the tiny gravity environment. First, Combination the conditions of conservations of momentum and angular momentum, the text give out the formula to compute the dynamics singularities and the way of designing of zero disturbance maps of the robotic system. Then, design neural network predictive controller for the space manipulator system using the neural network predictive controller block in the neural network toolbox and the S-function in Matlab language. After training of the neural network predictive controller, we have to reach the goal that the track of the space manipulator can converge at the track specified under the controller's output of controlling moment. Finally, we establish the space manipulator's simulation model to simulate the control process of the robotic system really.Using the neural network predictive controller and the no base disturbance path planning, we developed a way to save the expensive space fuel resources through path planning and the neural network predictive controlling. The results of the simulation show that automatically the track of the space manipulator can converge at the track specified well. The way of controlling not only has the high accuracy, but also refrains from rash action of deviation or vibration, namely the stability is also good.