Simulation Platform Built-up And Robust Two Dof Control For High-dynamic Servo System

This thesis analyzes control problems in high-dynamic performance servo systems.The nominal model of a high-dynamic servo system is built up, and from theperspective of robust control design, a robust control method is proposed by combiningtwo degree of freedom (DOF) control and μ synthesis. The two DOF control structure isbuilt up and the controller design is finished. A simulation platform based on practicalsystem experiments is established. The robust control design result is verified on thesimulation platform.Firstly, based on the principles of permanent magnet synchronous motor (PMSM)systems, the basic mechanism model of the position servo system is built up. Then,since high-frequency resonance problems exist in dynamic position servo system, amathematical model of the resonant modes is built up. On this basis, the nominal modelof position servo system combining resonance mode is set up and the model error isanalyzed. After that, High-order nominal model identification method and the upperbound of uncertainty identification method are given.Secondly, a position servo system simulation platform is built up based on thepractical system experiment. In this part, the nonlinear mechanism models of thepractical system for simulation platform are established. Then, the experiment designand the data processing method are presented and the model identification results aregiven based on practical experiment.Thirdly, the control problems of high dynamic performance position servo systemare discussed. For this specific goal, the robust two DOF control combining μ synthesismethod is proposed. The description of model uncertainty is presented. The structureof two DOF control is built up. And the principles of selection the weighting functionsare presented. In addition, the method of solving the control problem, especially theselection of scaling matrix in the D-K iteration, is discussed.Finally, based on the practical position servo system characteristics, high-ordernominal plant model and the uncertainty bound are obtained by identification.According to the principle of selecting the weighting function, the weighting functionsare determined. On this basis, controller design result via μ synthesis is provided, andthe control method is verified on the simulation platform..