Research On Nonlinear Transmission Joint’s Position Control Stategy Of Space Manipulator

Space manipulator, as a support technology for orbit maintenance and construction, has been applied more and more widely. With the complexity of spatial manipulation tasks and improvement of intelligence, how to improve the control precision then to improve the usage reliability is the key problem to be solved at this stage.The research content of this paper comes from the national key basic research development plan project (2013CB733000), mainly research on position control strategy of space manipulator nonlinear transmission joint.Based on the sufficient research on the factors of space manipulator joints on the nonlinear effect, completed a comprehensive consideration of friction and hysteresis effects of the joint dynamics modeling of mechanical arm space, being based on genetic algorithm and RBF neural network algorithm for the parameter identification model of joint friction and hysteresis, finally the position control strategy for simultaneously compensating friction and hysteresis control method is designed based on the sliding mode, effectively improve the location accuracy of joint.The main research work of this paper can be summarized as follows:Firstly, make a comprehensive for the nonlinear characteristics of integrated joint friction and hysteresis, established the dynamic model of space manipulator joint.Secondly, based on the LuGre model to identify and compensate the friction of the joint. The LuGre model can effectively describe the dynamic and static characteristic of friction and nonlinear problems near zero velocity regions, therefor this paper uses LuGre model, and use genetic algorithm to identify the friction parameters, based on these, to complete the joint friction compensation control.In addition, based on Preisach model to identify and compensate hysteresis for the joint. Preisach model is a multi-valued mapping, by solving the inverse model, can change multi valued mapping into one-one mapping. Because the neural network has strong capability of approximating one-one mapping function, so this paper uses the neural network to do the identification and compensation control work of joint hysteresis.Finally, design the global sliding mode controller that considered both friction and hysteresis. Based on the analysis of nonlinear friction and hysteresis influences on the joint, we design a global sliding mode controller, and apply it to joint control, to eliminate the influence of friction and hysteresis by feed forward compensation, so as to improve the robustness of the system.