Study On The Levitation Control Of The Hybrid Maglev System

Structure of the proposed hybrid suspension system is introduced in this paper.The four-point suspension strategies are analyzed, which are mechanical decouple andelectrical decouple. Then the equivalent method is employed to deduce themathematical model of the hybrid suspension system step by step, the levitationforces of the hybrid suspension system at different air-gaps are simulated, which aremuch larger than those of the pure electrical magnetic suspension system. The hybridsystem can really spare lots of electrical excitation, and then reduce the suspensionpower loss on a large scale, and it is a very promising suspension style. Theexperimental results of the levitation forces verify the model, hereby, the relativereference dynamic model and the absolute reference dynamic model are illustrated.The main schemes for the suspension system control are reviewed. Statefeedback control, slide-mode control, optimal control and fuzzy control in thelevitation control for the suspension system are introduced, and the design method ofthe state feedback control for the proposed hybrid system is described in detail.Experimental results of the state feedback control is presented, which verify that thestate feedback control can realize stable suspension but it is not very flexile.To improve the adaptability and increase the robustness of the suspension system,expert control and fuzzy control are imported to the levitation control, andhuman-simulated hierachical intelligent control is proposed, which includes threelevels. From the top to the bottom, the control precision is increasing while theintellectual degree is reducing. Experimental results show that the proposedhuman-simulated hierachical intelligent control used for hybrid suspension system,can obtain excellent control performance at different conditions, and realize fast,stable and precise levitation.To carry out good levitation without acceleration sensor, a fuzzy controller withadjustable factors is put forward, where the factor is coded and optimized by thegenetic algorithm which can reach global optimization. Genetic algorithm is improvedwhen it is used. With the optimized fuzzy controller, the response speed of the systemis increased, which has high control precision, and the acceleration sensor is removedand suspension that only position sensor is used is realized. Although only primaryresearch without acceleration sensor is developed, but according to the experimentalresults obtained from the lab, control only with position sensor, can obtaincomparatively high precision, and satisfy the demand of the suspension system.Normal levitation experiments, suddenly loaded experiments of movinglevitation, experiments of spanning obstacle based on state feedback control,human-simulated hierachical intelligent control and improved fuzzy control arefinished. The experimental results show that, state feedback control can realizemoving levitation, but its adaptation is a little bad;the proposed human-simulatedhierachical intelligent control is very robust and can adapt to all kinds of conditions,which is very promising;Based on improved fuzzy control, moving levitation controlfor the hybrid system without acceleration sensor is implemented, which is abeneficial research, and makes theoretic and experimental preparation for the realproject to realize suspension without acceleration sensor.