Research On Intelligent Vehicle Avoidance Strategy And Constraint Control Based On Pavement Pothole Detection

With the development of China’s economy,the road mileage increases year by year.Due to the increase of road service life and the impact of the external environment,a large number of pavement potholes appear on the road,which affects the safety and comfort of driving.At present,the active avoidance of intelligent vehicles mainly focuses on the avoidance strategy of positive obstacles such as pedestrians and vehicles,and there is less research on the avoidance strategy of negative obstacles such as road potholes.Therefore,an intelligent vehicle avoidance strategy and constraint control algorithm based on pavement potholes detection was proposed to improve the driving safety and comfort of vehicles on the pothole road in this paper.The main work of this paper is as follows:(1)Aiming at the problem of ahead road pothole detection and identification,a pavement pothole detection algorithm was designed based on Retina Net framework,and the detailed design process and training process of the algorithm were given.At the same time,a pavement pothole width and position identification method based on lane line was designed based on Lan Net framework,and a pavement pothole longitudinal distance identification method was designed based on by lidar point cloud segmentation to provide avoidance basis for pavement pothole avoidance strategy.(2)A pavement pothole avoidance strategy was proposed based on the width,location,and longitudinal distance of pavement pothole.Firstly,based on the informations of pavement pothole and the safety of lane environment,the pavement pothole avoidance strategies under five working conditions were designed,and the solution process of the end coordinates of obstacle avoidance route under different avoidance strategies was given.Then,based on the end coordinates of different obstacle avoidance routes,the slope sine function method was used to generate the avoidance trajectory,which provided the tracking route for vehicle trajectory tracking.(3)Aiming at the problem of lateral stability control in avoidance trajectory tracking,a DYC constraint controller was designed.Firstly,a two degree of freedom vehicle model was established,and the additional torque was introduced as the input.Secondly,a calculation method for front wheel angle was designed based on the avoidance trajectory.Finally,a DYC constraint controller was designed based on Barrier Lyapunov Function(BLF)and verified through simulation.The simulation results show that the designed constraint controller can achieve accurate trajectory tracking and ensure that the sideslip angle is always constrained within a safe area throughout the entire tracking process,improving lateral control stability.(4)In order to verify the effectiveness of the algorithm proposed in this article,SimulinkCar Sim joint simulation verification and real vehicle testing were carried out.Firstly,a Simulink Car Sim joint simulation platform was established,and the intelligent vehicle avoidance strategy and constraint control algorithm were validated combing the pothole offline data.Then,the real vehicle testing was conducted using Jiangsu University’s intelligent vehicle as the testing platform.The results of joint simulation and real vehicle testing showed that the intelligent vehicle avoidance strategy and constraint control algorithm based on road pothole detection proposed in this paper can realized pavement pothole avoidance,and can meet system constraints during the avoidance process,improving driving safety and comfort.