Research On Obstacle Avoidance Trajectory Planning For Unmanned Vehicles

Faced with the safety risks and challenges brought about by the increasing number of cars,autonomous driving related technologies can provide solutions for the development of my country’s traffic safety.At present,my country’s automobile industry is developing vigorously,and the gap between the pure electric platform and the international leading technology is gradually narrowing.Among them,the intelligence and safety of automobiles are bound to become the competitiveness of the domestic automobile industry in the 21 st century,and another key to the development of the automobile industry.commanding heights.Therefore,this paper studies the obstacle avoidance trajectory planning for structured road obstacle avoidance scenarios and unstructured road scenarios respectively.The main research contents are as follows:For obstacle avoidance scenarios under structured roads,considering that the trajectory planning for driverless vehicles is mainly affected by constraints such as safety and smoothness,the local reference lines used in the trajectory following process are optimized.In order to solve the requirement of dynamically updating the target lane in the process of lane change and obstacle avoidance,a local obstacle avoidance trajectory planning method based on multi-lane lane change decision-making is proposed,and the decision-making process of multi-lane lane change and obstacle avoidance is sorted out.For the obstacle avoidance scene under unstructured roads,the local grid map is used as the reference map for trajectory planning,and the transformation between the grid map and the vehicle coordinate system is derived.In order to solve the steering angle constraints and space constraints in the maneuvering of obstacle avoidance,a local trajectory planning method based on improved RRT* is proposed,and the Dubins curve is used to complete the splicing of the local trajectory and the global path,and the splicing points are added during the splicing process.Turn to the judgment of margin,thereby improving the success rate of local planning and reducing the time required for planning.In addition,combined with the above algorithm,the obstacle avoidance trajectory planning process based on improved RRT* is sorted out.Finally,a simulation experiment is carried out to initially verify the effectiveness of the planning algorithm.Finally,simulation and real vehicle tests are carried out on the trajectory planning of obstacle avoidance under the two working conditions.Through the joint simulation of Pre Scan,Car Sim and Simulink,the simulation test of the designed urban road scene test case is carried out,so as to test the performance of obstacle avoidance trajectory planning based on multi-lane lane change decision.Using the method of real vehicle verification,based on the proposed local trajectory planning algorithm based on improved RRT*,the real vehicle verification of the unstructured road scene obstacle maneuvering of the sweeper during the cleaning operation of the grain depot is carried out.The test results show that the obstacle avoidance trajectory planning algorithm for urban roads in this paper can effectively avoid the obstacle constraints caused by surrounding vehicles and can choose the optimal lane for driving.At the same time,the performance of the algorithm in low-speed scenarios is better than that in high-speed scenarios;The obstacle avoidance trajectory planning algorithm for the roads in the grain depot park can efficiently calculate the obstacle avoidance trajectory and splicing it with the global path.The spliced trajectory has better trackability,which verifies that the algorithm proposed in this paper is more suitable for the obstacle avoidance in this scene.Trajectory planning.