Alternating-direction-method-of-multipliers-based fast model predictive control for an aerial trees-pruning robot

Power transmission lines require efficient and reliable tree pruning to maintain their operation. This paper presents an adaptive Alternating Direction Method of Multipliers (ADMM)-based fast Model Predictive Control (MPC) for aerial tree pruning robots to address low operating efficiency and high labor costs. The proposed control strategy leverages MPC, a modern control method proven effective in complex systems, including aerial robots, to handle the challenges of attitude and position control during tree pruning operations. The adaptive ADMM algorithm is employed to solve constrained Quadratic Programming (QP) problems in real-time, enabling the robot to respond quickly to dynamic changes and maintain stability. Designed to perform real-time calculations on embedded computers with limited computing power, the control strategy is well-suited for implementation on aerial pruning robots. Improved operational capabilities, such as faster job site access, larger working space, and fossil fuel-free operation, result in increased efficiency and reduced labor costs. The paper covers the dynamic model of the pruning robot, the fast MPC control scheme, the adaptive ADMM for solving the QP problem, and the successful simulation and experimental implementation of the proposed control strategy on the aerial pruning robot.