| With the rapid growth of the number of fuel vehicles and electric vehicles,the convenient transportation environment provides great convenience for people ’s daily travel and cargo transportation,accelerating social development and economic cycle.However,due to the traffic flow in the old urban areas of large cities far exceeds the current carrying capacity of the original design,traffic jams during rush hours such as commuting are very common.Therefore,intelligent vehicles have received extensive attention from leaders and scholars in recent years.With the advancement of computer technology,sensor technology,communication technology and chip technology,it provides the hardware foundation for the development of intelligent vehicles.Although intelligent driving vehicles have made great progress in recent years,there are still many problems to be solved.In this paper,the problems of tracking accuracy degradation and vehicle instability caused by vehicle longitudinal and lateral coupling constraints,model uncertainty in motion control of intelligent vehicles are studied.The specific contents are as follows:(1)Estimation of vehicle key parameters: The estimation algorithms of vehicle speed,yaw rate,sideslip angle and road adhesion coefficient are designed.This paper is based on the vehicle three-degree-of-freedom dynamic model as the algorithm basis,The lateral acceleration is used as the measurement variable,and the covariance matrix of the measurement variable is adjusted online by fuzzy control.A fuzzy adaptive extended Kalman filter observer is built to estimate the vehicle speed,yaw rate and sideslip angle.The Dugoff tire model is used to calculate the normalized tire force,and the vehicle longitudinal acceleration,lateral acceleration and estimated yaw rate are used as measurement variables.A road adhesion coefficient observer based on unscented Kalman filter is built to estimate the road adhesion coefficient in real time.(2)Research on path tracking control: The hierarchical control structure is used to design the path tracking controller.According to the relationship between yaw rate and road curvature,the upper decision-making layer transforms the path tracking problem into the tracking problem of yaw rate.Taking the target vehicle speed,yaw rate and sideslip angle as the control objectives,an adaptive sliding mode controller based on RBF neural network is built.The lower controller takes the tire utilization rate as the optimization goal,and takes the estimated value of the road adhesion coefficient and the tire force boundary value calculated by the vertical load of the tire as the constraint conditions.The distribution problem of the tire force is transformed into an optimal value solution problem with nonlinear constraints.Finally,the decoupled longitudinal and lateral forces are converted into tire slip ratio and side slip angle through the Dugoff-based tire inverse model,and the vehicle longitudinal and lateral forces are decoupled to improve the vehicle ’s path tracking performance under extreme conditions.(3)Co-simulation and experimental verification:The Simulink/Carsim co-simulation platform is used to complete the simulation verification of the state estimation algorithm and the path tracking algorithm.The control performance of the controller is simulated and verified under normal conditions,ice and snow road conditions and high-speed conditions.It is proved that the path tracking control algorithm can ensure the tracking accuracy and stability of the vehicle under extreme conditions.A real-time simulation platform based on Simulink Real-Time is built.The ’ dual-machine ’ communication mode is adopted.The upper computer builds the control algorithm and compiles it into C language and transmits it to the lower computer.At the same time,the vehicle model state of the lower computer is monitored in real time.The control algorithm is simulated and verified in the real-time simulation platform.The results show that the controller has good real-time performance. |
PDF Full Text Request