Research On Vibration Control Of Active Suspension System Of EMU Based On Fuzzy Neural Network

With the increase of the speed of the emu,the phenomenon of vehicle vibration aggravation emerges one after another.The traditional passive suspension control technology has been difficult to solve the problems at present.But the active control technology can make the suspension system adjust automatically according to the current driving environment and improve the running stability of emu.This paper establishes the active suspension system model of emu considering the use of electromechanical actuator(EMA),designs the suspension system controller according to the proposed active control strategy,and verifies the superiority of the control strategy by simulation.The main research contents of this dissertation are as follows:Firstly,the operating background of emu is analyzed,the passive,semi-active,and active characteristics of three kinds of suspension system are introduced,the advantages of active suspension system and the research significance of fuzzy neural network control strategy are indicated,and the research status of active suspension application and fuzzy neural network control strategy at home and abroad are described.Secondly,the model of active suspension system of emu is established,which mainly includes the suspension system dynamic model,wheel-rail excitation model,EMA model and controller model.Among them,the suspension system model is the semi-vertical six-degree-of-freedom state space model.The irregularity of ballastless track in China’s high-speed railway is selected as the wheel pair external excitation in the wheel-rail excitation model and the time-domain simulation is carried out by using inverse Fourier transform.The actuator model uses electromechanical actuator to ensure that the suspension system can produce adjustable control force.Thirdly,a T-S fuzzy neural network controller is built.The design of the controller adopts the concept of using passivity to seek active,and explores the relationship between the dynamic deflection of two stage suspension,dynamic deflection change rate,acceleration of dynamic deflection change and the ideal force output of the actuator,and obtains the input current value of the actuator indirectly.Then the particle swarm optimization(PSO)algorithm is used to optimize the parameters of the fuzzy neural network to realize the self-adjustment of the fuzzy rules and improve the output accuracy of the suspension system.At the same time,a fuzzy controller is designed to verify the superiority of the fuzzy neural network control strategy.Finally,in the Matlab/Simulink environment,the vibration response of the emu body under passive control was firstly obtained by combining the wheel-rail excitation model.Then the response curve of the emu body under active control was obtained by co-simulation of the suspension system dynamics model,controller model and EMA model under the condition of the same wheel-rail excitation.The superiority of active control of suspension control is verified,and the control effect of the controller is evaluated from the perspectives of time domain,frequency domain and data analysis.The simulation results show that the three control strategies of fuzzy algorithm,T-S fuzzy neural network and T-S fuzzy neural network based on PSO optimization can effectively suppress the vertical vibration of emu,and the T-S fuzzy neural network controller based on PSO optimization has the best control effect.