Research On Path Planning Algorithm Of Six-axis Crawling Robot

With the rapid development of science and technology,robots have become more and more automated,and are widely used in various places such as logistics,medical and industrial production to replace manual work such as grasping and handling.The application of robots not only improves productivity,but more importantly,it can replace dangerous tasks for people and ensure the safety of people’s lives.In order to make robots better serve human beings,improving the intelligence of robots has become a research focus of many scholars.The focus of this paper is to enable the robot to achieve collision-free movement in a work environment with obstacles.A path planning algorithm is proposed to enable the robot to autonomously avoid obstacles in the environment and move from the starting position to the target position.Taking the laboratory six-axis robot KUKA KR10 R1420 and supporting loading and unloading system as the research object,based on the established robot kinematics model and environment model,the APF+Q-RRT path planning algorithm is proposed.The validity and practicability of the proposed path planning algorithm are verified through theoretical simulation and experiment.The main research contents are as follows:(1)Kinematics analysis.The kinematics model of the six-axis robot was established by the DH method.Based on this,the solution of the forward and inverse kinematics of the six-axis robot was completed,and the correctness of the results of the robot’s forward and inverse kinematics was verified by simulation.Robot working space.(2)Collision detection and environment modeling.The common collision detection methods and corresponding environment model methods are analyzed.After comparative analysis of their advantages and disadvantages,the cylindrical envelope method and the encircling ball method are used to model the robot and obstacles.Based on this,a six-axis robot and space are established.Mathematical model of collision detection for obstacles set in and solved.(3)APF+Q-RRT path planning algorithm.By analyzing the characteristics of the rapid expansion random tree(RRT)algorithm,Q-learning algorithm and artificial potential field method(APF),this paper introduces the learning mechanism of the Q-learning algorithm into the RRT algorithm and proposes the RRT based on Q-learning The algorithm,Q-RRT algorithm,is used to realize the path planning of the robot so that the robot can achieve collision-free movement in the working space with obstacles.For the path planning of the robot using the Q-RRT algorithm,because the selection of path nodes has a certain randomness,the selection of the path nodes when using the APF algorithm to plan the robot is done under the guidance of the target point,based on this,APF+Q-RRT path planning algorithm is proposed.The APF+Q-RRT path planning algorithm combines the APF algorithm with the Q-RRT algorithm according to the robot’s work task requirements and its environment information when planning the path of the six-axis robot,so that the robot can safely reach the target position.In order to verify the effectiveness of the proposed path planning algorithm,the position and size of obstacles in the robot workspace are randomly set,and the focus is on the use of APF algorithm,Q-RRT algorithm and APF+Q-RRT algorithm for robot path Obstacle avoidance effect and robot working efficiency during planning.The simulation results show that the APF+Q-RRT algorithm proposed in this paper overcomes the shortcomings of the APF algorithm that is easy to fall into the local minimum point,and also overcomes the Q-RRT algorithm due to the randomness of the selection of path nodes.The problem with long paths.(4)Experiment.In order to verify the practicability of the APF+Q-RRT algorithm,this paper conducted a robot path planning verification experiment on the six-axis robot KUKA KR10 R1420 and the supporting loading and unloading system.Install claws on the six-axis robot and build an experimental platform.A three-dimensional laser scanner is used to scan the working space of the robot,and the working environment model of the robot and the relevant parameters of obstacles are obtained.First,the APF+Q-RRT algorithm is used to simulate the path planning of the robot.After the simulation result is feasible,the path nodes obtained by the planning are sent to the robot in order to control the robot from the starting position without collision with the obstacle.The movement of the target position.Both simulation and experimental results show that using the APF+Q-RRT algorithm proposed in this paper to plan the path of the robot can enable the robot to achieve collision-free movement from the starting position to the target position in the workspace with obstacles.The simulation shows that the APF+Q-RRT algorithm proposed in this paper is used to plan the path of the robot,which overcomes the shortcomings of the APF algorithm being easy to fall into the local optimum,thereby improving the obstacle avoidance ability of the six-axis robot,and also overcoming the Q-RRT algorithm There is a defect of low planning efficiency caused by the randomness of path nodes.Theoretical simulation and experimental research show that the proposed APF+Q-RRT algorithm can be used in real robots.