Research On Path Tracking Control Strategy Of Vehicle Lane Change Based On Tire Cornering Stiffness Identification

As one of the core technologies of intelligent vehicle,active lane change technology has important research significance for solving road congestion and improving driving safety.The research on lane change strategy of vehicles under conventional working conditions has been relatively mature,but there are still deficiencies in the research on lane change strategy under extreme working conditions such as low adhesion road surface or high speed driving.In extreme working conditions,uncertain factors such as vehicle drive/braking capability limitation and structural parameter changes will bring hidden dangers to driving safety.Tire cornering stiffness will change with the constant change of tire sideslip angle,vertical load and current road conditions and other factors,at this time,the nonlinear tire cornering properties will change dramatically,which will affect vehicle control,and bring dangers to the vehicle driving in extreme conditions.Therefore,aiming at improving the tracking accuracy and stability in the process of active lane change of intelligent vehicles,this paper fully considers the strong nonlinear characteristics of tire cornering properties,and studies the vehicle active lane change tracking control strategy based on online identification of tire cornering stiffness.The research content of this paper is expected to improve the reliability of active lane change trajectory tracking under extreme working conditions.The main research work is as follows:Firstly,the intelligent vehicle active lane change system model was built based on the vehicle dynamics model of lateral,longitudinal and yaw three degrees of freedom.In order to accurately characterize the mathematical relationship between the wheel dynamics parameters,the tire dynamics model was established.Finally,the vehicle and road position relationship model is established to provide the basis for the follow-up trajectory planning research.Secondly,due to the drastic change of tire cornering stiffness in extreme working conditions will lead to the decrease of tracking accuracy and stability in the process of lane change,aiming at the dynamic change of tire cornering stiffness in real running conditions of real vehicles,the recursive fixed memory least square method was adopted to identify tire cornering stiffness online to replace the fixed tire cornering stiffness in the traditional control method.The traditional linear tire lateral force model will reduce the accuracy of parameter estimation due to the low matching degree because the tire lateral force will enter the tire nonlinear zone with the drastic change of the side Angle under the running condition.Therefore,this paper establishes a nonlinear tire lateral force model that can better reflect the characteristics of real tires,and adopts the lateral stiffness identification algorithm based on the recursive fixed memory least square method to overcome the problem of "data saturation" phenomenon caused by excessive data sampling in the traditional recursive least square method,which leads to the far distance of the estimated parameter from the real value.The method proposed in this paper reduces the influence of old data on the parameter fitting process by limiting the memory length,and the nonlinear function relationship between tire lateral force and side deflection Angle is fitted online,so as to identify the time-varying tire side deflection stiffness during lane changing.Finally,a model predictive trajectory tracking controller considering the uncertainty of tire cornering stiffness was established.The accuracy of the model predictive control algorithm in trajectory tracking was improved by adding the tire cornering stiffness identification model based on the recursive fixed memory least square method into the controller prediction model.Finally,an experimental platform based on Car Sim and Matlab/Simulink co-simulation was built to verify the effectiveness of the trajectory tracking control strategy proposed in this paper.A large number of experimental data show that the tire cornering stiffness identification algorithm adopted in this paper can accurately and quickly identify the real-time tire cornering stiffness online.Compared with the traditional algorithm,the recursive fixed memory least square method has higher fitting accuracy and smoother fitting process.In extreme working conditions,the model prediction trajectory tracking controller based on the identification strategy of tire cornering stiffness can effectively improve the tracking accuracy of the controller on the premise of ensuring the safety and comfort of lane changing.