| Active magnetic bearings,as a new type of high performance bearing,has many advantages and a promising future.The design of the controller,is directly related to the performance of the magnetic bearing system,which is the core of the whole digital control system.As a typical nonlinear and coupling system,it is difficult to obtain an accurate model,which makes it inapplicable for many control algorithms with high precision model.Therefore,it is very meaningful to adopt the algorithm which is not dependent on the system model.In this paper,a digital control platform based on TMS320F2812 is designed,and some control algorithms are carried out on the platform.The main work and research results are show as follows:(1)According to the basic principle of electromagnetism and kinematics,the modeling of the digital control system of 5Do Fs magnetic bearing is completed.(2)The basic requirements for the design of the controller are obtained by analyzing the open-loop and closed-loop stability of the single_degree_freedom magnetic bearings.According to the characteristics of the magnetic bearings system,the traditional PID algorithm is improved with dead zone and integral separation.A new model-free control algorithm is introduced for the active magnetic bearings.Then,a model-free controller with improved reference curve is designed,and successfully applied to the control of systems.The differences between the two control algorithm are compared by simulation.(3)According to the state space model of the axial 4Do Fs system,two different control strategies is studied.The first one is the centralized parameter control method based on state feedback decoupling,and the second one is the decentralized control method which ignores the coupling inside the system.The effectiveness of the two strategies is verified by simulation.(4)The design of hardware and software for digital control system is completed,and the experimental results are consistent with the simulation results. |
PDF Full Text Request