Sliding Mode Control Approach for Vision-Based High-Precision Unmanned Aerial Vehicle Landing System Under Disturbances

Abstract

Unmanned aerial vehicles (UAVs) face significant challenges when landing on moving targets due to disturbances, such as wind, that affect landing precision. This study develops a system that leverages global navigation satellite system (GNSS) signals and UAV visual data to enable real-time precision landings, and incorporates a sliding mode controller (SMC) to mitigate external disturbances throughout the landing process. To this end, a reference-model-based SMC is proposed, which defines reference values for each state to enhance the steadiness and safety of the velocity control system, thereby improving velocity state tracking and accuracy. The stability of the proposed controller is demonstrated using the Lyapunov method and comparing its performance against other controllers, including backstepping, linear-quadratic regulator (LQR), and proportional–integral–derivative (PID). The experimental results reveal a 75% reduction in maximum velocity tracking error and an 80% reduction in maximum landing error with the proposed controller. Finally, extensive real-flight tests confirm the stability and feasibility of the system.