State Space Estimation Method for Robot Identification

摘要

In this paper, we study the identification of robot dynamic models. The usual technique, based on the Least-Squares method, is carefully detailed. A new procedure based on Kalman filtering and fixed interval smoothing is developed. This new technique is compared to usual one with simulated and experimental data. The obtained results show that the proposed technique is a credible alternative, especially if the system bandwidth is unknown.