Research On Path Planning And Control Algorithm Of Unmanned Vehicle Based On Artificial Potential Field

As China’s industrialization continues to advance,the number of car ownership in the country is also rising,and the consequent road congestion and frequent traffic accidents have become the main problems at this stage,and drivers play a leading role in these problems,so autonomous driving technology is born.The autonomous driving technology relies on accurate environmental perception,and planning a reasonable path based on environmental information is the premise of autonomous driving,and strictly following the planned path is the guarantee for the full promotion of autonomous driving.The specific research content of this paper for the path planning and control of self-driving vehicles is as follows:Firstly,for the problems that the artificial potential field(APF)method cannot reach the target point,is easy to fall into the local minimal value point and the path is long,an algorithm of artificial potential field method planning from the starting point and the target point in both directions is proposed,which can effectively avoid the problem of planning failure when the obstacle is near the target position;the quadratic A* algorithm,eliminating some redundant nodes by judging whether there are obstacles between nodes,improving the smoothness of the path,and integrating the quadratic A*algorithm with the two-way artificial potential field method,proposing strategies such as algorithm switching conditions,map conversion method,selection and optimization of connection points,which effectively solve the problem of long paths and easy to fall into the local minima of the artificial potential field method;using Bessel curves to optimize the paths with discontinuous trajectory curvature and large The Bessel curves are used to optimize the paths to ensure that the planned paths meet the requirements of vehicle driving.The improved two-way artificial potential field method is compared with the traditional artificial potential field method,and the results show that the improved algorithm has a great improvement in the ability to cope with complex obstacles compared to the traditional algorithm path,and the path length is reduced by6.42%.Second,to address the problem of horizontal and vertical control coupling in the tracking control of self-driving vehicles,a tracking control model with the error amount as the state variable is constructed to decouple the horizontal and vertical motions,and the LQR lateral tracking controller is designed based on the Linear Quadratic Regulator(LQR)algorithm;the fuzzy control principle is introduced,and the lateral position deviation and heading angle deviation as the input of the fuzzy controller,and dynamically adjust the weight matrix in the LQR controller to solve the problem of poor tracking effect and unstable driving due to the fixed weight matrix.Through the joint simulation of Carsim/Simulink,it is verified that the fuzzy control-adjusted LQR lateral tracking controller can ensure good tracking accuracy and driving stability under different speed conditions.And at high speed,the maximum value of heading angle deviation is only 0.97 deg,and the mass side deviation angle is 0.186 deg,so the vehicle driving stability and comfort is better.Finally,the experimental platform is built to verify the effectiveness of the path planning algorithm and tracking control algorithm,and various sensors are used to obtain the environment map and vehicle location information,and the effect of vehicle path planning and tracking control is observed after setting the target point.The results of the real vehicle test show that the proposed path planning algorithm can complete the path planning problem under complex obstacles,while the path meets the vehicle driving requirements;the proposed tracking control algorithm has accurate tracking for the reference path and good driving stability.