Nowadays, maglev suspension system (MSS), using the electromagnetic force to suspend, can effectively reduce the mechanical vibration friction and wearing loss that might cause a malfunction in operational process with the conventional contact operation. However the parameters in the mathematical model are related to the permanent magnet geometry, distance and the total mass of the platform. To achieve fast response, two fuzzy model systems and the parameter identification algorithm for maglev suspension system are proposed in this thesis. The first proposed model method based on Table Lookup Scheme (TLS) is to establish a common fuzzy system procedure. By gathering input and output data pairs, a fuzzy system is created to replace the currently used mathematical model. On the other hand, since the parameters are related to the hardware structure, the second Gradient Descent Algorithm (GDA) is proposed. By using the gathered data pairs, the parameters of membership functions are updated through the Gradient Descent Algorithm (GDA). However, the disadvantage of the two fuzzy methods is the lack of the resistance to uncertainties and external disturbance in operational time. Therefore, a main method, supervisory fuzzy model control, is proposed to cope with a shortcoming of the first proposed fuzzy model by using supervisory control. As named, the main method is the combination between the first fuzzy model and supervisory control to upgrade the system to be robust for the possible utilization in working environment and to extend the model at the practice of maglev transportation. In this thesis, the major supervisory fuzzy model systems based on Table Lookup Scheme (TLS) is applied in both simulation and experiment to ensure the effectiveness of tracking performance and handling with uncertainty problems. The comparison with other intelligent controllers are also included in this thesis to make more specific consideration in detail. Thus, the chattering phenomenon in the supervisory fuzzy model system is almost less than in the other controllers such as total sliding mode control and fuzzy total sliding mode control, and fast response can be obtained without degrading significantly the tracking performance of the system dynamic.
