Parameter Identification of a Robot Arm by Particle Swarm Optimization and Haar Wavelet

Abstract

Physical parameters in robot dynamics model have to be concerned in order to design and develop model-based control algorithm. This work presents a novel parameter identification technique for a 5-axis serial robot manipulator. The system dynamics is strongly nonlinear and nonlinear parameterized with 52 unknown parameters. Most conventional identification methods cannot be employed. The proposed method is based on an integration of particle swarm optimization (PSO) and Haar wavelet. The PSO is utilized to search for the best-fit parameters and the Haar wavelet is used to simplify the computation of the objective function. The experimental results show that the predicted response generated by the identified parameters matches with the measured response. The relative RMS prediction errors are less than 2%, verifying the effectiveness of the proposed identification method.