Obstacle Avoidance And Motion Planning Of 7-DOF Redundant Manipulators

Along with the complexity of robotic work space and the diversification of tasks in industrial production,the orignal six-axis robots have insufficient capacity in individualized product customization and dissatisfy the adaptability of the flexible product lines.Such as part precision assembly,product sorting,robot traction teaching and human-machine collaboration.Due to its flexibility and redundancy,the seven-axis robot have more advantages in interference avoidance,circumferential welding and high-density placement.Therefore,this paper is dedicated to seven-axis redundant robot,focusing on inverse kinematics,collision detection and obstacle avoidance,which provides theoretical support for the development of seven-axis redundant robots.The main contents are as follows:Firstly,this paper focuses on LR4-R850 redundant robot and establishes MD-H model.The forward(inverse)kinematics equation is solved according to the MD-H model.The arm angle is refered to kinematics solution of SRS configuration robot.The augmented Jacobian matrix combines the arm angle Jacobian matrix and the end-effector Jacobian matrix.The joint angle of the redundant robot can be iterated through augmented Jacobian matrix.This way is applicable to the redundant robot which does not meet Piper criterion.After analyzing augmented Jacobian matrix,kinematic singularity and algorithm singularity are derived.Based on the results of kinematics,the simulation platform of redundant robot is established,and the accuracy of the kinematics algorithm is verified.Secondly,the collision detection of obstacles with seven-axis redundant robot is analyzed.The robot motion node model is created to simplify calculation process.The collision detection between redundant robot and obstacle is simplified into geometric problem,which means positional relation is discussed with line and plane in three-dimensional space.After analyzing the relationship between robot and obstacle,the OBB bounding box model is proposed.According to the geometric relationship,the collision detection model based on the cylinder envelope,cuboid envelope and spherical envelope are analyzed and compared in detail.Thirdly,the algorithm of obstacle avoidance and motion planning for seven-axis redundant robot is analyzed.Since seven-axis redundant robot has self-motion characteristics,the nullspace motion algorithm of obstacle avoidance is introduced.Besides,redundant robot is difficult to plan in high-dimensional space,then C-Space is introduced.The RRT algorithm based on rapidly-exploring random tree,the bidirectional RRTConnect algorithm and the optimal RRTStar algorithm are analyzed.Since LR4-R850 robot model is derived from SolidWorks and ROS,the obstacles are added to workspace and simulation for redundant robot is verified.According to simulation results,null-spcae motion algorithm and rapidly-exploring random tree algorithm have good feasibility and effectiveness for solving obstacle avoidance problem of seven-axis redundant robot.Finally,kinematics experiment on LR4-R850 seven-axis redundant robot and simulation of obstacle avoidance with motion planning are introduced in detail.Kinematics experiment includes forward(inverse)kinematic verification and null-space motion verification.Obstacle avoidance with motion planing includes null-space motion and rapidly-exploring random tree algorithm.The verification results of kinematics show the effectiveness of the algorithm.The null-space motion results prove the application value of self-motion characteristics in obstacle avoidance problem.In addition,the RRTConnect and RRTStar algorithms in obstacle avoidance based on rapidly-exploring random tree also have good applicability to obstacle avoidance problem of seven-axis redundant robots.