Assembly in Motion With a Mobile Manipulator Based on Servoing Control and Disturbance Compensation

Abstract

Assembly in motion is a key solution to enhancing system efficiency in dynamic manufacturing environments. However, kinematic uncertainties and external disturbances often hinder mobile manipulators from performing high-precision tasks. This paper addresses this challenge by proposing a robust servoing control method. Specifically, we integrate sliding mode control with neural networks to enable accurate motion tracking and assembly operations without requiring precise knowledge of the disturbance model or uncertainty parameters. To ensure the stability of the tracking system, we implement a Barrier Function-based Adaptive Sliding Mode Controller (BFASMC) as a self-tuning position controller for the mobile manipulator. This controller compensates for external bounded disturbances while avoiding overestimation of control gains. Additionally, we employ an Input-to-State Stability Long Short-Term Memory Neural Network Controller (ISS-LSTM-NNC) as an auxiliary module to handle uncertainties in target motion. As a result, the proposed framework allows the robot to safely and effectively complete dynamic tracking and assembly tasks. We validate our method through experiments on the AMB150-UR5 mobile manipulator, involving both object tracking and assembly operations. The experimental results demonstrate the effectiveness and robustness of the proposed method.