Study On Fuzzy Dynamic System Design And Test Of Semi-active Suspension

Suspension system is the important vehicle part to improve the ride comfort and safety of vehicle. Conventional passive suspension has great limitation in harmonizing the ride comfort and handling. Though active suspension can remedy the defects of traditional suspension system, it has not been widely applied heretofore, owing to its high cost in manufacture and operating. The semi-active suspension, which is the compromise between passive suspension and active one, overcomes the limitation to improve the performance of vehicle as well as avoids high cost and complicated control. So, study of semi-active suspension is the hot topic of the newly technology of vehicle industry. In the paper, system modeling and control of semi-active suspension based on fuzzy theory, optimization of suspension parameters, and experiment of semi-active suspension are researched further.First, Performance tests on the throttle adjustable damper have been done. The nonlinear angle-damp relation is founded, which provide a reference for the modeling and system control of semi-active suspension.Secondly, based on performance test of the damper, its F-V characteristic is carved up linearly. Using fuzzy theory, a quarter vehicle fuzzy model that can contain non-linearity of the damper is founded. To achieve matching rigidity and damp, suspension rigidity parameter is optimized with RBF neural network.Thirdly, a new design method of semi-active suspension fuzzy control system is presented, the fuzzy control system based on fuzzy model is designed and its stability is researched. Then the controller based on 80C51 of semi-active suspension is opened up. And the bench tests are done. The results of experiment and simulation are consistent by and large, which shows that the fuzzy control system of semi-active suspension is effective and stable and can improve dynamic performance of the vehicle effectively.Finally, the eight-freedom degree vehicle model is founded. Suspension parameters are optimized with neural network. The road tests are done to testify validity of the model, review influence on vehicle of parameters optimization, obtain vibration characteristic of the vehicle to actual road input.