A vision-aided fuzzy adaptive sliding mode controller for autonomous landing of a nonlinear model helicopter on a moving marine platform

Abstract

PurposeA vision-assisted fuzzy adaptive sliding mode controller is presented in this research and implemented on a nonlinear helicopter model, which is about to land on a moving ship. Stabilization of the dynamics and tracking the landing path are required, at the same time. This study aims to take one step closer to fully autonomous landing, which is a growing trend.Design/methodology/approachAn integrated guidance and control is considered for the model helicopter. A fuzzy logic is designed to adaptively choose the best control parameters for the sliding mode controller and solve the challenge of parameter tuning. A self-organizing matrix consisting of fuzzy sliding mode parameters is formed instead of a single parameter with the goal of enhancing controller tracking capability. A simple, precise and fast image recognition system based on OpenCV is used to detect the proper point for descending without getting any special data from the ship and by only using a general “H” sign.FindingsThe problem is simulated under intense disturbances, while the approach and landing performances are acceptable. Controller performance is compared and validated. Simulation results show the robustness, agility, stability and outperformance of the proposed controller.Originality/valueThe novelty of this paper is the designed procedure for using a simple image recognition system in the process of autonomous ship-landing, which does not use any special data sent from the ship. Besides, an improved nonlinear controller is designed for integrated guidance and control in this specific application.