Study On Robot Path Planning In Complex Dynamic Environment

Path planning is a very important part of the robot in the completion of the task,a good way can save a lot of task time.With the complexity of the application environment of intelligent robots,robots will face more dynamic obstacles or targets in path planning,so the path planning of robots in dynamic environment is becoming more and more important.In this paper,the obstacles in the dynamic environment are classified from geometric and kinematic characteristics,and the characteristics of the robot are described.The difficulties of robot path planning and the shortcomings of traditional methods are analyzed in two kinds of typical dynamic obstacles:uniform variable motion obstacle,random motion obstacle.Then,this paper proposes a hybrid path planning method to solve the problem of robot path planning in complex dynamic environment with even variable motion obstacle and random motion obstacle.The method consists of two parts:1)adaptive dynamic artificial potential field method;2)random movement obstacle collision avoidance rules.The adaptive artificial potential method introduces the adaptive factor in the traditional potential field function,so that the weight of the components in the potential function is adjusted reasonably when the angle of the robot is different from the relative motion vector of the obstacle and the target point.The path of the shock and the big corner.At the same time,the collision force matrix is introduced into the repulsive potential field to solve the problem that the target point of the traditional method is unreachable.The random motion obstacle avoidance rule draws on the traditional strategy of walking along the wall,first sets the virtual target point,dynamically adjusts the distance between the robot and the obstacle while ensuring the safety of the robot,and keeps the robot moving in one direction as much as possible to avoid the path redundancy.In the planning of the path,the robot first to determine the type of obstacle,the uniform movement of the movement of adaptive dynamic artificial field method to avoid obstruction,the random movement obstacle with random movement obstacle avoidance rules to avoid.Under the guidance of this method,the robot can avoid the invalid path,smooth the obstacle,and reach the target point.Finally,the complex dynamic environment is simulated in MATLAB,and the simulation experiments are carried out on the gravitational potential field,the repulsive potential field,the random motion avoidance rule and the mixed path planning method in the adaptive artificial potential field method.The results show that the total length of the path,the length of the local path and the smoothness of the path are evaluated.The experimental results are analyzed and the validity of the method is verified by comparing the results of the traditional algorithm.The main points of this paper are as follows:1)Analyze the geometry of the obstacle,the characteristics of the movement and the movement characteristics of the robot;2)Propose the hybrid path planning method,improve the traditional artificial potential field method,introduce the adaptive factor,And designed the random movement obstacle avoidance rules;3)proposed three path evaluation indicators,the path of a more comprehensive evaluation.