Trajectory planning of a redundant planar manipulator based on joint classification and particle swarm optimization algorithm

This paper presents a general method for the trajectory planning of the redundant planar manipulator. The mathematical relation between joint space and Cartesian space in a two-dimensional space is first derived. The joint classification is employed to obtain the solutions of corresponding joints. It divides joints into class I redundant joints, class II redundant joints, and nonredundant joints. The new application of knot points in the quintic B-spline curve is introduced to generate inverse solutions of class I redundant joints. Examples show that the number and distribution of knot points have a large effect on their solutions. Moreover, the particle swarm optimization algorithm is extended to generate solutions of class II redundant joints. It also optimizes the initial trajectories of joints and end-effector. Finally, solutions of nonredundant joints can be generated by the derived relation between the joint space and Cartesian space. The proposed methodology is confirmed by a case study. Under the same conditions, results show that the solution obtained by the extended method is not only better than that obtained by the particle swarm optimization algorithm but also closer to the global optimal solution.