| Magnetic levitation is a kind of non-contact technology. It plays a very important role in extending the service life of mechanical equipments as there is no friction. This technology is the future in transportation, non-contact actuators, precision engineering, aerospace,life sciences and other fields. However, the magnetic levitation system is highly nonlinear and unstable, and therefore, designing an effective and simple controller for such a system is very challenging. The predictive function control(PFC) has many advantages, such as good real-time and robust,and it can be used to control the fast system. In this paper the PFC is applied to the magnetic levitation ball system which is used as an object of this research. The experiment research is carried out based on dSPACE.Firstly, the purpose and significance of this research is described. The mathematical model of the magnetic levitation ball system was established according to the structure and principle of the system. Then the open-loop stability and controllability are analyzed.Secondly, a real-time control experiment platform was designed based on dSPACE. It is used to analyze the dynamic characteristics of magnetic levitation ball system. With an abundant of software and hardware resource of the dSPACE and its ControlDesk software, the experiment of the control method can be easily carried out.Thirdly, the traditional predictive function control algorithm is adopted by the magnetic levitation ball system. According to the open-loop unstable characteristics of the system, a feedback stabilization control is designed using PD control theory. In the basis of the control structure, the predictive function control is used.Finally,two different improved scheme is put forward for the shortcomings of traditional PFC methods and practical application. On one hand, for the characteristics of nonlinear uncertainty, it is difficult to get the best control effect when using the traditional PFC method. A new fuzzy PFC method was presented. The discrete-time fuzzy controller was designed to improve the dynamic quality of the system by adjusting the reference trajectory time online. Besides, the parameters of predictive function controller were calculated off-line,which ensures the timeliness. On the other hand, in order to overcome the influence of modeling mismatch and other uncertain factors on the control accuracy of the magnetic levitation ball system, the 2-degree-of-freedom(2-DOF) PFC method is presented. A disturbance attenuation controller is designed to compensate the error between the actual output and the model output based on internal model control method. The simulation and experiment results indicate that the 2-DOF PFC system has high control precision. In addition, it shows excellent rapidness and remarkable robustness. |
PDF Full Text Request