Robust control of electrically driven robots using adaptive uncertainty estimation

A novel adaptive uncertainty estimator using the first order Taylor series for the robust control of electrically driven robot manipulators is proposed.By using the Taylor series, the dynamics of robotic system can be estimated online and adaptively compensated using model-free control approach.The proposed controller does not require bounding functions as used for the robust estimators and to be linearized in parameters as used for the adaptive estimators. This paper presents a novel robust control for electrically driven robot manipulators by designing an adaptive uncertainty estimator based on the first order Taylor series. The estimator is simple and model-free in a decentralized structure. The uncertainty is then efficiently compensated in the control system. The controller does not require the bounding functions as an advantage over the conventional robust controller. Therefore, it is simpler, less computational, and more efficient. It is verified by stability analysis and its effectiveness is shown through comparisons with a terminal sliding mode control approach and a Legendre polynomials uncertainty bound estimator simulated on a SCARA robot driven by permanent magnet dc motors.