Model Predictive Control for an Aerial Tree-Pruning Robot Based on Alternating Direction Method of Multipliers

Abstract

Obstacles of some trees within the electric power transmission line channel are of great threat to the electricity supply. Nowadays, the tasks of clearing threatening tree branches are still mostly operated by hand and simple tools. In this article, an aerial tree-pruning robot with a novel structure is designed to improve the pruning operation efficiency and enhance the safety of the staff. However, the long arm of the pruning tool results in much higher rotational inertia of the robot, which brings difficulties for the robot to remain stable. Therefore, a control scheme based on model predictive control is proposed for the aerial tree-pruning robot and to deal with an uncertain system during the pruning operation period. One of the main contributions is that an ADMM (alternating direction method of multipliers) algorithm that solves the constrained QP (quadratic programming) is adopted to implement the model predictive control on embedded computers with limited computational power. The dynamic model of the pruning robot is firstly presented. Then, the control scheme of MPC for the pruning robot is presented. Moreover, the QP problem of robot control is addressed with ADMM. Finally, simulation experiments of attitude tracking as well as the antidisturbances capability verification have been conducted. Results for the system of aerial tree-pruning robot are given to demonstrate the effectiveness of the developed attitude tracking control scheme using ADMM-based MPC.