Optimal Trajectory Planning under Kino-dynamics Constraints for a 6-DOF PUMA 560

Abstract

We present in this paper an effective method to deal with the problem of the trajectories planning in time-efforts quadratic optimal of the robots manipulators in the point to point tasks. The technique suggested in this work, consists at the beginning to standardizing the time scale then to break up the trajectory into two functions which are modelled by Cubic Splines (Natural and Clamped) according to the properties of each function. Finally, the problem of optimization is solved by using the Genetic Algorithms to find the optimal trajectory. An algorithm is developed which makes it possible to minimize a function objective which represents a weighting between the transfer time and the efforts of the actuators, with the satisfaction of the geometrical, kinematics and dynamic constraints. Results on a robot PUMA560 6R are presented to illustrate the effectiveness of this technique.