Output Feedback Control of Variable-loaded Four-joint Manipulators With Unknown Saturated Actuator Dynamics

This study aims to achieve precise tracking control for a four-joint manipulator with an unknown saturation nonlinearity of the actuator, unknown dynamics, and operating under varying unknown loads. To achieve this aim, first a comprehensive model of the robotic manipulator is formulated that considers various factors such as unknown dynamics due to the parameters uncertainties, joint coupling, nonlinearities, actuator dynamics, and load effects. This model forms the fundamental basis for our proposed control approach. Next, the output feedback controller is designed by combining a backstepping controller with an extended high-gain observer (EHGO). The backstepping controller is designed to achieve a desired motion profile, and the EHGO is designed to estimate the unknown dynamics and the load variation effects using the measured position at each joint. The structure of the proposed controller allows for the motion profile tracking for each joint independently of the other joints. To evaluate the efficacy of the suggested control, we conducted simulations and experiments using a QArm robotics manipulator from Quanser to achieve a tracking motion profile through the pick-and-place task. The results of these assessments demonstrate the successful accomplishment of accurate motion control under unknown load variation during the operation, and the integral absolute error (IAE) is compared with the recent published work.