| With the rapid development of the automotive industry,research on intelligent vehicles has attracted the attention of numerous scholars in recent years.Among the three key core technologies of intelligent vehicles,trajectory tracking is a very important part.The tracking accuracy of intelligent vehicle trajectories is influenced by many factors,and the road adhesion coefficients can also affect the tracking accuracy of the vehicle’s trajectory.This paper proposes a model predictive control(MPC)trajectory tracking strategy based on road surface adhesion coefficients and lateral deviation angle constraints for vehicle trajectory tracking on different road surfaces.Genetic algorithm(GA)and error back propagation(BP)neural network are combined to estimate the road surface adhesion coefficient,while considering the linear constraints of tire lateral deviation angle to improve the stability of high-speed vehicle driving.The effectiveness of the proposed estimation algorithm and control strategy is verified through relevant simulations and experiments.The main research content of this paper is as follows:To meet the real-time requirements of trajectory tracking control algorithms,and fully consider the nonlinear dynamic constraints of the vehicle and the lateral stability under extreme conditions,a seven-degree-of-freedom vehicle dynamics model is established,which can reflect the vehicle’s lateral,longitudinal,yaw,and wheel motions in real time.The magic formula tire model is used to analyze the tire’s lateral characteristics and to identify the relationship between the road adhesion coefficient and the linear variation range of the tire’s lateral force-slip angle.Based on the analysis of the mechanical characteristics of intelligent vehicles and tires,this study proposes a neural network-based estimation method for the non-linear problem of road adhesion coefficient.The input parameters and network topology of the neural network estimation algorithm are determined by the vehicle dynamics model.To improve the estimation accuracy of road adhesion coefficient,the global optimization characteristic of genetic algorithm(GA)is combined with BP neural network to propose a GA-BP neural network-based road adhesion coefficient estimation method.Simulation and vehicle testing show that the GA-BP neural network-based road adhesion coefficient estimation algorithm has higher accuracy and stability compared to the BP neural network estimation algorithm.Based on the MPC theory,a variable-constrained Model Predictive Controller(MPC)considering the road adhesion coefficient is established.The control method mainly includes model linear discretization,predictive model design,and quadratic programming problem conversion.Based on the change of the road adhesion coefficient,the effect of the tire side slip characteristics on vehicle stability and trajectory tracking accuracy is considered,and a tire side slip angle constraint is set.The predictive model and constraint conditions are embedded in the quadratic programming problem,and the optimal control variables are obtained by solving it.Simulation results show that the tire side slip angle constraint designed based on the road adhesion coefficient can suppress the problem of excessive tire side force,which is beneficial to improving the trajectory tracking accuracy and driving stability of intelligent vehicles.To further investigate the impact of different road adhesion coefficients on the trajectory tracking accuracy of intelligent vehicles,an intelligent vehicle trajectory tracking control strategy is proposed that incorporates real-time estimation of road adhesion coefficient and lateral deviation angle constraints.The GA-BP neural network road adhesion coefficient estimation algorithm is combined with a variable lateral deviation constraint MPC trajectory tracking control algorithm.Using the Carsim and Simulink joint simulation platform,double lane change simulations are conducted under different road conditions.The simulation results show that compared to the traditional MPC trajectory tracking controller,the proposed intelligent vehicle trajectory tracking algorithm with consideration of road adhesion coefficient and variable lateral deviation angle constraints has higher accuracy and vehicle stability at high speeds. |
