Research On Automatic Obstacle Avoidance Method Based On Improved Artificial Potential Field

The avoidance maneuver is based on sensors such as LIDAR,mm Wave radar,and combined inertial navigation system to perceive the surrounding environment and locate the position of the vehicle.A collision-free avoidance path is obtained through path planning algorithms,which is then output to the control module to complete path tracking.Path planning and tracking control,as one of the key components of autonomous driving technology,have become a hot research topic.Based on the national key research and development plan project "Research and Practice on Group Intelligent Consignment Vehicles for Airport Luggage Transfer"(2021YFE0203600),This article focuses on the research of automatic avoidance path planning and tracking control methods,and the specific contents are as follows:(1)Research on the traditional artificial potential field method for automatic obstacle avoidance path planning.Introduce the theory of traditional artificial potential field method,and study the causes of local optimal and target unreachable problems from the perspective of the trend of repulsion and resultant forces.(2)Research on automatic obstacle avoidance path planning using improved artificial potential field method.To address the problem of unreachable targets,the repulsive potential field function is improved to attenuate the repulsive force.Aiming at the local optimal problem,an invasive weed algorithm is introduced to guide vehicles to escape from traps by establishing an escape potential field.Propose a dynamic target point method and establish a dynamic gravitational potential field optimization path planning process.The effectiveness of the improved scheme was verified through simulation,and the results showed that the algorithm achieved the improvement goal.Establish a dynamic obstacle environment,simulate and verify the algorithm’s obstacle avoidance ability in the dynamic environment,and the results show that the planning results meet expectations.(3)Research on the improved artificial potential field method for automatic obstacle avoidance path planning.The principle of Model predictive control is introduced,and the Vehicle dynamics model is established.After linearization and Discretization,the linear timevarying model is obtained.In the process of designing the objective function,multiple constraints such as centroid sideslip angle,lateral acceleration,and tire sideslip angle are introduced.To address the issue of previous studies not paying attention to the impact of weight matrices on tracking results,a variable weight matrix is designed based on the curvature changes of the reference path.The simulation verifies the effectiveness of the controller,and the results show that the controller meets the design requirements.(4)Joint simulation and on-road experiments.A Carsim/Simulink joint simulation platform was set up,verify the automatic obstacle avoidance scheme proposed in this paper,and the results show that it meets expectations.Build a real vehicle experimental platform and verify the effectiveness of the path planning in this paper through a wire controlled car.The results show that the improved artificial potential field method performs well in real vehicle path planning.