Research On Model Predictive Control Of Maglev Levitation System

With the development of modern society, people constantly put up the requirements on traffic quickness, safety, comfort and environment protection. As a fast, efficient, energy-saving and environmentally protective vehicle, maglev train will have a large development space at present and in the future. As one of the key control technology of maglev train, the levitation control is the basis and precondition to realize train operation. So, how to improve the performance of maglev levitation control system is of important significance and has become the research hotspot.At present, there exist a certain number of maglev levitation control algorithms at home and abroad, but the research using model predictive control (MPC) algorithm is still relatively lacking, especially at the treatment of objective optimal control with constraints. This thesis mainly studies the MPC algorithm for maglev levitation system through taking the electromagnetic-suspension maglev train as a research object. The main research contents are as follows:Firstly, this thesis establishes the nonlinear model of maglev levitation system and analyzes the stability of the model using the nonlinear control theory. Then the state feedback stabilization control was implemented on the levitation control system by approximate linearization stabilization method and feedback linearization stabilization one, respectively. And the merits of the two methods are compared and analyzed through MATLAB simulation.Secondly, the unconstrained MPC controller for maglev levitation system is designed, based on feedback linearization stabilization. This controller improves the disturbance-resisting ability of the levitation control system through adding the simulated climbing resistance of disturbance variables, external pressure and other uncertain factors.Thirdly, in order to further improve the control performance and ensure the train operation safety, stability and ride comfort, the constrained MPC controller for maglev levitation system is designed, considering the constraints of levitation gap, velocity and acceleration. According to the requirements of real-time control, the constrained optimization algorithm based on the primal-dual interior-point (PDI) approach is designed. Further, considering the potential problem of the PDI method, the more efficient constrained optimization algorithm based on prediction-correction primal-dual interior-point (PCPDI) approach is designed.Finally, the feasibility and validity of the control algorithms proposed in this thesis are verified through MATLAB simulation.This thesis employs the MPC method to solve the maglev levitation control problem, which can conveniently handle the objective optimal control with constraints and has good control performances to resist disturbances. The constrained optimization algorithm based on PCPDI approach has higher calculation efficiency, and can ensure the real-time performance and the feasibility of the control algorithm. The simulation results show that the MPC algorithms proposed in this thesis can satisfy the requirements of multi-constraint, real-time and disturbance-resisting performance by maglev levitation system.