Genetic Optimisation and Experimentation for the PUMA 560 Manipulator

The search for minimum-time motion of an articulated mechanical arm by tessellating the joint space involves heavy computational burden. In this work, Genetic Algorithms (GA's) are used to tackle this problem while considering different optimisation criteria, namely, minimum motion time, constraints on torque commands and constraints on end-point velocities. In addition, the search algorithm considers all constraints imposed on the manipulator design, including bounds on motor torques. In addition to reporting the simulation results of a number of case studies, experiments are carried out using a transputer-based PUMA arm.