Fuzzy Hybrid Control Of Vehicle Active Suspension Systems Based On Genetic Algorithm

Suspension system is very important to improve the safety and ride comfort of vehicles.Compared with passive suspension and semi-active suspension,active suspension can adapt to different vehicle operating conditions and take into account the handling stability and ride comfort.Active suspension is an important trend in the development of the suspension industry.As a key part of active suspension design,the quality of control strategy directly affects the suspension’s vibration attenuation effect.Applying advanced control strategies to active suspensions is one of the most potential ways to improve the performance of suspension systems.In this paper,a quarter-car active suspension system is taken as the research object,and fuzzy hybrid controllers based on genetic algorithm(GA)are designed.The main contents include:(1)According to the dynamic characteristics of the quarter car suspension system,the mathematical model is established based on the state-space description form.The C-level random road disturbance model that causes the vehicle body vibration is established.On this basis,three indicators for evaluating the ride comfort and driving safety of the suspension system are given,and the performance of the passive suspension system is simulated and analyzed.(2)Considering the characteristics of the expansion and contraction of the universe with the error in the variable universe strategy,the variable universe fuzzy PID controller of the active suspension system is designed.In addition,since the selection of parameters in the expansion factor description function will affect the vibration attenuation effect of the suspension system,GA is used to optimize these parameters.Simulation results show that,compared with PID and fuzzy PID controllers,the GAbased variable universe fuzzy controller has superiority in reducing the vertical acceleration of the vehicle body and improving the ride comfort.In addition,the proposed control method has strong robustness against uncertainties of sprung mass and vehicle speed.(3)Considering the time-frequency localization characteristics of wavelet function and the advantages of multi-population genetic algorithm(MPGA)that can avoid premature convergence,a fuzzy wavelet neural network(FWNN)control method for active suspension system is proposed based on MPGA.The wavelet function is utilized as the membership function of the fuzzy neural network.MPGA is employed to optimize the connection weight parameters of the FWNN.Simulation results show that the proposed control method can effectively reduce the vertical acceleration of the vehicle body and ensure that the closed-loop system meets the time-domain hard constraints such as suspension stroke and relative dynamic tire load.