Parameters identification of robot manipulator based on particle swarm optimization

Abstract

In this paper, we investigate identification methods for dynamic parameters of robot manipulator. The focused method is based on heuristic particle swarm optimization algorithm (PSO) with some extended features. The estimated parameters by PSO are used to predict required joint torques for high accuracy tracking control. The effectiveness of some PSO methods for tracking control problem are verified by cross-validation with data set produced by several trajectories.