A Practical Approach for the Auto-tuning of PD Controllers for Robotic Manipulators using Particle Swarm Optimization

Abstract

An auto-tuning method of PD controllers for robotic manipulators is proposed. This method suggests a practical implementation of the particle swarm optimization technique in order to find optimal gain values achieving the best tracking of a predefined position trajectory. For this purpose, The integral of the absolute error IAE is used as a cost function for the optimization algorithm. The optimization is achieved by performing the desired movement of the robot iteratively and evaluating the cost function for every iteration. Therefor, the necessary constraints that guarantee a safe and stable movement of the robot are defined, which are: a maximum joint torque constraint, a maximum position error constraint and an oscillation constraint. A constraint handling approach is suggested for the optimization algorithm in order to adapt it to the problem in hand. Finally, the efficiency of the proposed method is verified through a practical experiment on a real robot.