Design of Active Yaw Gaze-Locking Control for Quadcopters in Tower Orbital Inspection

Close-range 3D photogrammetric inspection of tall towers with quadcopters is degraded by wind, which pushes the vehicle off its scanning orbit and changes the camera-to-structure distance. Conventional flight controllers track a time-parameterized trajectory, so wind-induced lag must be corrected in both space and time. This paper presents a cascaded geometric vector field (VF) guidance law combined with active yaw gaze-locking for helical tower scanning. The VF decouples spatial convergence from timing, an outer velocity PI loop rejects wind through its integral action by tilting the vehicle into the wind, and the yaw channel is dedicated to keeping the camera pointed at the tower axis. In 6-DoF simulations with a 4.5 m/s mean wind, gusts, and Dryden turbulence, 30 paired Monte Carlo runs with PSO-tuned gains show that the VF reduces the mean cross-track error by 70.6% relative to PID trajectory tracking and by 48.3% relative to nonlinear guidance logic (NLGL), and reduces the stand-off distance variation by 65.8% (p < 0.001). Because the gaze-locking law uses the measured position, the camera pointing error remained about 0.8° for all guidance laws, showing that pointing accuracy is set by the yaw loop and is decoupled from path following.