Research On Steering And Braking Combined Collision Avoidance Control Based On MPC

According to the statistics of the white paper on China automotive aftermarket,China will become the country with the largest number of vehicles in the world in 2025.Vehicle brings great convenience for people’s travel,but the environmental and road safety problems are also increasingly serious.In order to reduce traffic accidents and improve driving safety,the active safety technology of automobile has been paid great attention and developed.As an active safety technology,vehicle collision avoidance system mainly realizes collision avoidance by braking.Studies have show that the combination of steering and braking has more advantages when the vehicle speed is high.Therefore,this paper aims to explore when the vehicle detects obstacles,by controlling the steering and braking system of the vehicle,the vehicle can smoothly avoid obstacles and ensure the stability of the vehicle,so as to improve the driving safety of the vehicle.Vehicle state parameters and road adhesion conditions have an important influence on the design of collision avoidance control system.Therefore,this paper designs an estimator to estimate the vehicle lateral velocity and road adhesion coefficient in real time with high accuracy based on the rolling horizon estimation method.By analyzing the critical collision avoidance distance and speed under different road friction coefficient and vehicle speed,the key parameters are provided for the design of collision avoidance control system.In order to solve the problem of sideslip instability of high-speed and low adhesion road vehicles using pure steering collision avoidance method,this paper designs a new linear time-varying model predictive controller to improve the lateral stability of vehicle when steering to avoid collision.This model predictive controller uses the front steering angle optimization sequence to predict the tire state stiffness and then linearizes the tire force in the predicted horizon.In addition,the MPC controller proposed in this paper is compared with traditional linear time-varying and nonlinear path tracking controllers respectively.When steering and braking are combined for vehicle collision avoidance control,hierarchical control method is mainly used for the distribution of actuator control action.It mainly includes the following two ways:(1)the upper layer calculates the front steering angle and yaw moment,and the lower layer distributes the wheel braking moment based on the required yaw moment(2)The upper layer calculates the yaw moment needed to complete collision avoidance,and the lower layer distributes the wheel angle and braking torque based on the minimum tire utilization.When the upper controller is designed by the above two methods,the vehicle dynamics constraints and the physical constraints of the lower actuator may be ignored,resulting in the occurrence of collision.In this paper,the front steering angle and the braking torque of the four wheels are taken as the optimization variables,and an integrated steering braking combined collision avoidance controller based on MPC is designed.On this basis,the integrated collision avoidance controller that combined steering and braking based on MPC is designed,and the tire force utilization ratio was also considered in the design of MPC optimization objective.The results show that the integrated controller designed in this paper has good path tracking efficiency and can great improve the vehicle driving stability obviously.