Research On Control Technology Of Vehicle Seat Suspension System Based On Magnetorheological Damper

With the progress of science and technology and the advancement of living standards,automobiles have gradually become an indispensable means of transport in people's travel.In the meantime,people have also put forward greater requirements on the ride comfort and the stability of vehicles.The seat suspension system plays a vital role in suppressing the vibration transmission.Magnetorheological damper(MRD)is novel type of vibration isolation device developed by applying the rheological mechanism of magnetorheological fluid.Compared with traditional hydraulic components,it has the advantages of not being affected by faults,low power consumption and adjustable output damping force,which makes the semi-active seat suspension of vehicles using magnetorheological damper popular among experts and scholars.It has become effective means to restrain the vibration transmission of vehicle seat suspension system through installing a magnetorheological damper device to attenuate the vibration energy transmitted to the body of driver and passenger from external excitation,and then heighten the ride comfort of the vehicle seat.However,researches on dynamic modelling of magnetorheological damper and semi-active control methods for seat suspension are not yet mature and perfect,and many related theoretical knowledge and key technologies need to be further explored.Based on this,this paper has carried out the following research work.(1)Mechanical performance test and dynamic modeling of magnetorheological damper.With reference to the requirements of relevant test standards,a fatigue tensile machine was utilized to conduct mechanical performance test on the magnetorheological damper developed by research group.A parameter identification method combining particle swarm optimization algorithm and nonlinear least square method is designed,and the unknown parameters in the modified Dahl model are identified based on the collected experimental data of the output damping force-displacement and velocity characteristics.By comparing and analyzing the experimental data and simulation data under different working conditions,the feasibility of the model to describe the mechanical properties and hysteresis characteristics of MRD is verified.Meanwhile,its high accuracy provides strong guarantee for the subsequent modeling and simulation research of the semi-active seat suspension system.(2)Modeling of vehicle semi-active seat suspension system and analysis of its dynamic characteristics.Considering the real road conditions,two types of road models,random input and bump input,are established respectively.The dynamic characteristics and advantages and disadvantages of 1/4 car,1/2 car and full car semi-active seat suspension systems are briefly described,the Newton's second law and Lagrange's equation are applied to derive the motiondifferential equations for the three models.A five degree-of-freedom(DOF)1/2 passive seat suspension system simulation model is built on the Matlab\Simulink platform,and six indexes are selected to represent its dynamic characteristics in time domain.Concurrently,the Laplace transform is used to calculate the transfer functions of seat acceleration and seat suspension dynamic displacement relative to road excitation,and the influence of four system parameters change on the dynamic performance for passive seat suspension is analyzed by the amplitude frequency characteristic curves.(3)Design and simulation verification of control strategy for vehicle semi-active seat suspension system.Combined with the complex nonlinear vibration characteristics of semiactive seat suspension system,based on the detailed introduction of the fuzzy control theory and proportional-integral-derivative(PID)control theory,a variable universe fuzzy controller based on fuzzy inference is developed to solve the problem that the fuzzy rules formulated in fuzzy control overly depend on the expert experience and lead to the low accuracy.Besides,in order to make up for defect that the three parameters of proportional,integral and derivative in PID control cannot be adjusted adaptively with system error,a fuzzy-PID controller is devised.Under this premise,the hybrid learning algorithm combining BP neural network and least square method is used to train the fuzzy rules formulated in the fuzzy-PID controller offline,and three parallel T-S type neural fuzzy network structures with two inputs and single output are constructed,namely,ANFIS-PID controller.Taking the seat acceleration and seat suspension dynamic displacement as performance evaluation indexes,under the excitation of random road surface and bump road surface,the modified Dahl model of MRD and the semiactive seat suspension system model of 5-DOF 1/2 vehicle are built to verify and analysis the vibration isolation effect for the proposed two control strategies.The simulation results show that both semi-active control strategies can effectively enhance the ride comfort of the vehicle seat suspension system,and the control effect of the ANFIS-PID strategy is better than that of the variable universe fuzzy strategy.