Perturbation Observer-based Feedback Linearization Control of Robotic Arm Platform

Abstract

The present research practically controls the position of a robotic arm considering uncertainties in its model, together with external perturbations. An input-output linearization control is formulated based on an exact dynamic model, incorporating an extended state observer (ESO) to estimate the perturbations. By measuring the arm angular displacement, the perturbations-related term in the ESO algorithm is considered as a new state variable added to the system equations. Through state estimation using ESO, an accurate model is established to design the feedback linearization control, ensuring the arm effectively follows the desired paths. Stability analyses are provided to show the boundedness of errors in estimation and control. Ultimately, an empirical implementation on a constructed robotic arm platform demonstrates a good functionality of the observer-based control approach.