Path Planning for a Planar Hyper-Redundant Manipulator with Lockable Joints Using Particle Swarm Optimization

In this paper, the problem of path planning for a special hyper-redundant manipulator with lockable joints is solved using particle swarm optimization. Two strategies have been used. In the first strategy, the path-planning problem is solved in two stages; first, the inverse kinematics of manipulator is solved with continuous PSO and then the sequence of switching is optimized by modified discrete PSO. The results of discrete PSO are compared to discrete GA to show the effectiveness of the discrete PSO. In order to implement multi-objective optimization in inverse kinematics calculation, linear fitness function and Vector Evaluated PSO (VEPSO) are used. In the second strategy, a novel approach in particle swarm optimization is proposed. In this approach, the integer and continuous value numbers are joined together in a single particle to form a hybrid particle. The results of two strategies were compared to show the convergence speed and performance of the second strategy which was proposed here.