Tracking long-distance power transmission line under uncertain wind fields by QUAV using nonlinear model predictive control

In long-distance power transmission line inspections, utilizing a quadrotor unmanned aerial vehicle (QUAV) to stably and safely track high-voltage power lines from a safe distance has been demonstrated as an effective method. However, due to the intricate 3D spatial configuration of the power lines and the challenges in accurately detecting them, maintaining a consistent and safe tracking distance between the drone and the power lines remains a significant challenge. Additionally, the variability of wind patterns in long-distance inspection areas further complicates the task of ensuring stable tracking. Therefore, considering both the kinematic constraints of the QUAV-gimbal camera system and the influence of dynamic wind fields, this study proposes a Nonlinear Model Predictive Controller (NMPC) based approach for QUAV power line tracking. Initially, the QUAV's target tracking control task is formulated as an optimization problem, incorporating stringent motion attitude constraints to establish a cost function for target tracking distance in 3D space. Subsequently, by integrating gimbal camera functionality, a field-of-view (FOV) cost model is developed to determine the QUAV's target point while accounting for wind disturbance factors. Through the implementation of the NMPC controller, continuous airborne tracking is achieved while mitigating adverse effects from time-varying wind fields. Ultimately, optimal control input commands are generated for use with both the QUAV and gimbal controllers. Simulation experiments validate that this method effectively enables the QUAV to track long-distance power lines within time-varying wind field environments, with high precision, efficiency, and robust tracking performance.