Adaptive Neuro-Fuzzy Inference System-Based Sliding Mode Control in the Presence of External Disturbances and Parameter Variation for Quadcopter UAV

Quadrotor unmanned aerial vehicles (UAVs) are increasingly becoming essential tools in applications such as surveillance, military operations, crop monitoring, search and rescue, and inspection of hazardous terrain. Their control is not an easy endeavor due to the underactuated and highly coupled dynamics. Among many control methodologies, sliding mode control (SMC) has long been recognized as one that is insensitive to system nonlinearities and external disturbances. Yet, the inherent chattering effect of SMC will lead to system degradation and actuator damage. To mitigate this limitation, this study proposes an adaptive neuro-fuzzy inference system-based sliding mode control (ANFIS-SMC) method that incorporates the strength of ANFIS and the robustness of SMC to enhance quadrotor trajectory tracking with reduced chattering effects. The control system comprises position, altitude, and attitude controllers that online learn from system errors and control signals and ensure stable and precise flight under dynamic flight conditions. The performance of the ANFIS-SMC controller developed in the current study is validated using MATLAB/SIMULINK simulations and compared with a classical SMC scheme. Results confirm that a Comparison between SMC and the proposed ANFIS-SMC controller is conducted in terms of both disturbance and parameter variation, and the proposed ANFIS-SMC controller has shown better performance improvement of 58.1%. Reduces chattering and achieves improved tracking accuracy, confirming its worthiness for robust quadrotor control tasks.