Research On Kinematics And Obstacle Avoidance Path Planning For Redundant Manipulator

Redundant manipulator has the advantages of flexible operation, fault tolerance andhigh reliability. Its self-motion feature can be used to avoid obstacles, avoid singularities,avoid joint limit, optimize joint torque and other performance index. In this paper, inversekinematics for redundant manipulator, singularity analysis of redundant manipulator andobstacle avoidance path planning for redundant manipulator are researched, which isarmed at a7-DOF redundant manipulator for Chinese future space station and providestheoretical support for the development of our country’s space station manipulator. Themain contents are as follows:Firstly, gradient projection method based on singularity robust inverse (SRI) isstudied, which solve the problem that the usual inverse differential kinematics solutionsbased on Jacobian pseudoinverse cause very high joint velocities in the vicinity ofsingular configurations. The reason is analyzed with singular value decomposition. SRIis introduced to address this problem. SRI merges with gradient projection method so thatgradient projection method based on SRI is established. In order to reduce numerical error,pose error is added to inverse kinematics equation. The eventual inverse kinematicsequation for redundant manipulator can avoid high joints velocities near singularityconfiguration, has the ability to optimize performance index, and also has high numericalprecision. The effectiveness of the method was verified by simulation based onSimMechanics.Secondly, the singularity of a7-DOF redundant manipulator is analyzed. A depend-screw suppression approach is introduced and applied to the singular analysis of a7-DOFredundant manipulator. Five sets of conditions defining singular configurations areobtained. Each condition is analyzed in detail. Geometric features and illustration ofsingular configurations of the7-DOF manipulator are presented, which makes it easilydetermine whether the manipulator is singular.Thirdly, an improved bi-RRT*obstacle avoidance path planning algorithm and apath smoothing algorithm based on smoothing spline are proposed. The success rate ofRRT*algorithm for obstacle avoidance path planning is low and its convergence speedis slow. In order to solve these problems, an improved bi-RRT*algorithm with bi-directional search strategy and path optimization algorithm based on triangle inequality is proposed. To resolve the problem that planning path is not smoothing, a path smoothingalgorithm based on smoothing spline is proposed. The results of simulations showed that,both in low dimension space or in high dimension space, the improved bi-RRT*algorithmhas more high planning success rate and faster convergence speed than original RRT*algorithm. The path smoothed by path smoothing algorithm based smoothing spline issmoother, which also makes manipulator move steadily in obstacle-free space.Finally, in order to verify the validity of inverse kinematics algorithm and obstacleavoidance path planning algorithm for redundant manipulator, experiments wereconducted on space manipulator ground test system based on air-floating platform withjoint1, joint2, joint6and joint7fixed, which is equivalent with3-DOF planar redundantmanipulator.