Research On Optimal Control Method Of Vehicle Semi-active Suspension

As an important part of the driving system,the suspension has a direct impact on the vehicle dynamics.Semi-active suspension ensures better human body comfort and roadholding on different road surfaces by controlling damper damping,which significantly improves vehicle driving and riding experience compared to passive suspension,which is more important for electric vehicles with greater mass in the same size and class.As new energy vehicles are accepted by the market,more and more electric vehicles are adopting semi-active suspension to ensure the stability of vehicle body and good damping effect.At present,the common semi-active suspension is mainly divided into coil spring combined with solenoid-controlled damper scheme and air spring combined with solenoidcontrolled damper scheme.Coil springs have a long metal fatigue life,occupy little space,have near linear stiffness and are not adjustable,the adjustable damper arrangement is easy to achieve without changing the vehicle suspension structure and at a lower cost.Air springs have non-linear stiffness characteristics,which improve suspension the ability of vibration isolation and enhance passenger’s comfort.However,air springs have a shorter service life,occupy more space and are more expensive than coil springs.This paper was based on a major project of the National Natural Science Foundation of China（NSFC）entitled "Integrated Cooperative Control of Vehicle Motion under Extreme Working Conditions"（Project No.61790564）.Based on the optimal control theory,this paper investigates the multi-objective optimal control problem for semi-active suspension and the nonlinear optimal control for semi-active suspension considering the nonlinear stiffness characteristics of air spring.In this paper,the main work of this paper are as follows:Firstly,random pavement and bump pavement excitation models and semi-active suspension systems were established.The dynamic characteristics of a solenoid-controlled damper of a B-class car and a one-piece air spring damper of a C-class car were obtained using the MTS shock absorber test bench.Using test data of the B-class vehicle damper to identify model parameters to built the damper model and inverse model.The stiffness of the air spring was tested on the MTS test bench,a polynomial was fitted to elastic force of the air spring.Secondly,modeling the dynamics of the linear suspension system of a quarter vehicle;the basic theory of linear-quadratic optimal control（LQR）in optimal control is introduced and completed the design of linear LQR controller for semi-active suspension.The simulation experiments of LQR multi-objective optimal control for semi-active suspension was conducted to verify the effects of different weighting parameters of LQR on suspension performance indexes by combining time domain response and frequency domain response comparison analysis.The simulation results show that semi-active suspension LQR control can achieve multi-objective optimization of the suspension,focusing on improving vehicle dynamics or synthetically improving suspension performance.In addition,the LQR control has an explicit mathematical analytical expression,which is simple and efficient,and can ease the conflict between suspension comfort and handling stability to a certain extent while also meeting the demand for real-time control of semi-active suspension.Thirdly,the basic principle of ILQR and the theoretical relationship with LQR are introduced and applied to a quarter vehicle nonlinear suspension model considering air spring nonlinear stiffness to complete the semi-active suspension ILQR controller design and algorithm programming.Simulation analysis is carried out by time domain response and frequency domain response methods,and compared with linear passive suspension,nonlinear passive suspension and LQR controlled semi-active suspension to verify the effect of ILQR optimal control on the control of semi-active suspension using air spring.By comparing the simulation results,ILQR control can further reduce body acceleration and tire dynamic deformation under random road and bump road input,especially in the most sensitive vibration frequency（4～8Hz）range vehicle body vibration can be effectively suppressed,significantly improving human comfort.Further analysis shows that the main reason for the improved ILQR control is that the algorithm itself can optimize the damping force control parameters according to the suspension motion state,in addition to the nonlinear stiffness of the air spring contributing to the improved human comfort.Finally,an equal scale quarter car suspension HIL test rig was designed and built,including mechanical system,sensor system,control system and data acquisition system,and conduct the suspension HIL test under random road and bump road excitation.The experimental results show that the LQR algorithm can effectively suppress the body and tire vibrations and improve the comprehensive performance of the suspension under random and bump road excitation.The HIL test results are close to the simulation experimental results,which can verify that the controller proposed in this paper can improve the passenger’s comfort and road-holding.