| The annual growth of vehicle population has not only brought convenience to human life,but also caused a series of social problems such as traffic congestion,environmental pollution and energy shortage.As the workload of drivers increases with the complexity of traffic environment,the frequency of traffic accidents is also increasing.Therefore,the research and development of intelligent vehicles with assistance driving and automatic driving functions has become the development demand to solve traffic congestion,road safety and other problems.Due to the great potential of intelligent vehicles in alleviating congestion,reducing accidents and saving energy,they have attracted extensive attention from academia,industry and government departments.In recent years,thanks to the rapid development of sensing technology,communication technology and chip technology,the feasibility technical reserve of automatic driving has been increasingly improved.However,in order to realize real industrialization,there are still many technical difficulties to be solved.In view of this,this paper focuses on the key problems of vehicle-road state estimation,motion control,decision-making and control architecture of intelligent vehicles.The main study contents of this dissertation are as follows:(1)Considering the longitudinal motion,lateral motion,yaw motion and wheel rotation motion of the vehicle,the vehicle system dynamics model is established,and the accuracy of the model is verified by Car Sim software.Based on the Brush tire model and the established vehicle dynamics model,the effects of tire stiffness parameters and road adhesion coefficient on tire force and vehicle dynamics response are analyzed.Based on the Square-root cubature Kalman filter(SCKF)algorithm,a state estimator which can be used to identify vehicle longitudinal velocity,lateral velocity and sideslip angle in real time is designed.(2)Aiming at the problems of model mismatch and measurement noise uncertainty in the process of road adhesion coefficient recognition,an integral estimation scheme which can avoid error accumulation is designed.By adaptively calculating the longitudinal and lateral stiffness of the tire and modifying the effective adhesion coefficient,a novel tire model with high precision is proposed based on the Brush model.Based on the maximum correntropy criterion,an improved SCKF algorithm which can avoid the interference of abnormal measurement noise is proposed.Combining the vehicle dynamics model with the proposed novel tire model and the improved SCKF algorithm,an estimation scheme for identifying the road adhesion coefficients on the left and right sides of the vehicle is designed,and the effectiveness of the framework is verified based on different types of working conditions.(3)Aiming at the uncertainty of process noise and measurement noise in intelligent vehicle target state tracking,a robust adaptive SCKF algorithm which can avoid the interference of abnormal noise is proposed.The maximum a posteriori probability criterion is used to estimate the statistical values of process noise covariance and measurement noise covariance,so as to improve the estimation accuracy of SCKF algorithm when the noise is stable.The fault detection rule is designed based on the standardized measurement innovation sequence,and the noise covariances are corrected by using the real-time measurement innovation to ensure the robustness of the algorithm to the abnormal noise in the process of state tracking.The simulation results under different noise interference conditions show that the robust adaptive SCKF algorithm combined with statistical estimation and correction principle can significantly reduce the impact of noise statistical uncertainty on vehicle target state tracking accuracy.(4)Taking the distributed intelligent electric vehicle as the research object,a trajectory tracking controller is designed based on the model predictive control theory.The cornering stiffness parameters are corrected in real time by using the tire lateral forces estimated based on SCKF,and the effectiveness of the controller is verified based on the hardware in the loop test.Considering the saturation constraint of tire force and the relative weight between yaw rate and sideslip angle,a vehicle stability controller based on sliding mode control is designed.On this basis,a weight adaptive criterion based on normalized stability index is designed,and then an adaptive cooperative strategy for vehicle trajectory tracking and stability control is proposed,and significant advantages of this method under different limit conditions are verified by comparative analysis.(5)In view of the typical driving scene of highway,an intelligent vehicle decisionmaking and control architecture considering longitudinal safety and lateral stability is proposed.Based on information such as traffic environment,relative position,road adhesion coefficient and vehicle motion status,the need for speed control and lane change operation is determined.The reference path to meet the road boundary and collision avoidance constraints is generated based on the relative position and lane line coordinates of the host vehicle and obstacle vehicles,and the optimal lane change path is selected based on the evaluation function considering safety and comfort.The designed controllers above are used for trajectory tracking and stability control,which reduces the “pressure” from the decision planning layer to the control layer,and avoids the conflict between tracking accuracy and vehicle stability. |
