A PSO Algorithm for Mapping the Workspace Boundary of Parallel Manipulators

This paper presents a novel method to determine the multi-dimensional workspace boundary of parallel manipulators using a modified Particle Swarm Optimization (MPSO) framework. A generic objective (fitness) function that gives optimum value at points where a manipulator reaches its limit is formulated. By finding these points using MPSO, the workspace boundary of a parallel manipulator can be constructed. To demonstrate the method, the mapping of the boundary of a Stewart platform workspace is presented. The generic objective function can be modified easily to include any number of parameters and constraints, and this allows the method to be adapted for solving the workspace boundary problem of other parallel manipulators.