Adaptive-Neural Finite-Time Sliding Mode Control for Quadrotor Helicopter Attitude Stabilization in Complex Environments

Abstract

Achieving attitude stabilization in quadrotor helicopters (qhs) operating in complex environments, characterized by external disturbances and model uncertainties, presents a significant challenge. This study presents an adaptive-neural finite-time sliding mode control (anft-smc) to effectively address these challenges. The proposed method integrates nonsingular fast terminal sliding mode control (nft-smc) with a radial basis function neural network (rbfnn), which is equipped with a fast auto-tuning law. Consequently, the method transcends qh model constraints and obviates the need for explicit knowledge of external disturbances and model uncertainties. The effectiveness of the proposed approach in stabilizing attitude dynamics is rigorously validated through a comprehensive Lyapunov stability analysis, scrutinizing key stability aspects. Extensive simulations conducted using matlab and Simulink, compared against a nominal nft-smc implementation based on a state observer (so) benchmark, demonstrate the superior performance and robustness of the proposed method in achieving finite-time stabilization of qh attitude dynamics.