| The performance of magnetic levitation is determined mainly by its magnetic levitation control system. Therefore, it has the important significance to study the suspension control method. This paper studied the control performance from two points. One was to optimize the parameters of the controllers, the other was to establish the evaluation system of the controllers. The main research contents and results are shown as follows:1、Set up two different linear systems of the maglev system.On one hand, using the first-order Taylor equation, the maglev system was turned to be the linear system 1. On the other hand, using the method of the differential geometry, the maglev system was turned into the linear system 2.Results:the linear system 1 contained two different subsystems, and the linear system 2 which had two same subsystems and had the nonlinear characteristics of the original nonlinear system.2、Do the parameters identification for the unknown parameters of the maglev system.The paper used the least square fitting in order to get the parameters of the drive coil and the interaction force between two magnets. Thus, the identification parameters were a、b、c and d.Results:in the linear system 1, the electromagnet 1 was stable and the electromagnet 2 was unstable; in the linear system 2, the electromagnet 1 and the electromagnet 2 were not stable.3、Design the state feedback suspension controller without the observer and the state feedback suspension controller with the observer, and get two optimized suspension controllers. On one hand, the damping ratio and the natural oscillation frequency were two important parameters of the state feedback suspension controllers, that is, the damping ratio affected the overshoot, and the oscillation frequency affected the steady state value. On the other hand, the initial condition was the important parameter of the state feedback suspension controller with the observer, which affected the rising time.Results:the linear system 1 was fit to use the state feedback suspension controller without the observer, the linear system 2 was suitable to use the state feedback suspension controller with the observer.4. Design the PID suspension controller based on the genetic algorithm and the PID suspension controller based on the particle swarm optimization algorithm, and obtain two optimized suspension controllers.On one hand, the paper designed the PID suspension controller based on the genetic algorithm, did the analysis of the parameters, which were the coding method, the number of the population, the crossover probability, the mutation probability and the maximum iteration, and then got the optimized suspension controller. On the other hand, the paper designed the PID suspension controller based on the particle swarm optimization algorithm, did the analysis of the parameters, which were the inertia weight method, the learning strategy, the maximum speed and the maximum iteration, then got the optimized suspension controller.Results:obtain two optimized PID suspension controllers based on the genetic algorithm and based on the particle swarm optimization algorithm respectively.5. Design the fuzzy comprehensive evaluation model based on analytic hierarchy process, assess and compare the three different controllers. Firstly, according to the carrying capacity and anti-disturbance ability of the maglev system, the linear system 1 had 2 first grade indices,7 second grade indices and 20 third grade indices, and the linear system 2 had 4 first grade indices,8 second grade indices. Secondly, using the analytic hierarchy process method, the weights of all the indices were determined. Thirdly, using three levels of evaluation, three suspension controllers for the linear system 1 was assessed, and three suspension controllers for the linear system 2 was assessed by two levels of evaluation. Finally, we got the evaluation results.Results:the linear system 1 was fit to use the PID suspension controller, and the linear system 2 was suitable to use the state feedback suspension controller. |
PDF Full Text Request