Multi-body dynamics and Gyro-MPC-DOB based trajectory tracking control of live-line working robot

When the Live-line working robot carries out high-altitude operations, uncertainty factors such as external wind disturbance, end-load variations, and body vibration affect the control performance of the robotic arm, resulting in poor trajectory tracking accuracy of the robotic arm. To address the limitation where purely algorithmic control fails to eliminate low-frequency, large-amplitude disturbances due to actuator bandwidth constraints, a mechano-electrical synergistic control strategy based on gyroscopic feedforward and model predictive control-disturbance observer (MPC-DOB) feedback is proposed. First, the multi-body dynamics model of the working manipulator is established. A physical stabilization mechanism based on the gyroscope precession effect is constructed to utilize the active anti-torque generated by the gyroscope to counteract wind loads, followed by the derivation of a dynamics model incorporating the residual attitude of the base. The core innovation of this work lies in integrating the gyroscope structure physics into a unique dynamic robot platform and performing corresponding multibody dynamics modeling under complex wind disturbances, thereby bridging the gap between pure algorithm control and physical hardware limitations. Second, an attitude feedforward compensation law is designed to convert the residual inclination detected by the gyroscope into torque input, explicitly compensating for gravity coupling terms. Simultaneously, a DOB is introduced to estimate and reject high-frequency disturbances and model uncertainties in real time. On this basis, MPC integrates the feedforward signals and observed values to perform receding horizon optimization, achieving optimal control under multiple constraints. Theoretical analysis using Lyapunov stability criteria proves that introducing gyroscope feedforward is essential for the convergence of system energy. Finally, simulation and field experiments were conducted to validate the proposed method. Compared with MPC, the overall motion errors of Joint 2 and Joint 5 were reduced by 36% and 39%, respectively. These results demonstrate that the proposed strategy successfully improves trajectory tracking accuracy under strong perturbations, meeting the stringent requirements of online live-line operations.