A novel approach to UAV control: Fuzzy PID-based ANFIS (FPIDANFIS)

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Aerospace Engineering

Background:

• Unmanned Aerial Vehicles (UAVs) require advanced control systems for precise trajectory tracking.
• Conventional Fuzzy PID (FPID) controllers lack adaptive capabilities to handle real-world uncertainties.
• Existing methods often rely on model-based approaches, limiting adaptability.

Purpose of the Study:

• To introduce a novel data-driven hybrid control strategy, FPIDANFIS, for enhanced UAV trajectory tracking.
• To develop an adaptive controller that inherits baseline FPID performance while adding real-time tuning capabilities.
• To evaluate the robustness and accuracy improvements of FPIDANFIS under various challenging conditions.

Main Methods:

• A Fuzzy PID based Adaptive Neuro-Fuzzy Inference System (FPIDANFIS) controller was designed using historical data from a baseline FPID controller.
• An Adaptive Neuro-Fuzzy Inference System (ANFIS) model was trained offline using a dataset generated via MATLAB's Neuro-Fuzzy Designer.
• The controller was evaluated in simulations involving nominal tracking, input disturbances, and parameter variations, with performance measured by ITAE.

Main Results:

• FPIDANFIS demonstrated superior performance compared to the baseline FPID controller.
• A significant reduction in Integral of Time-weighted Absolute Error (ITAE) was achieved: 42% under disturbances and 22.12% under parameter variations.
• The controller exhibited enhanced robustness and adaptability in dynamic and uncertain environments.

Conclusions:

• The proposed FPIDANFIS offers a data-driven, adaptive extension to the FPID control framework for UAVs.
• This novel approach significantly improves trajectory tracking accuracy and maneuverability for UAV applications.
• FPIDANFIS presents a promising solution for applications demanding high precision and reliability in autonomous systems.