Data-Driven Control of Chaotic Nonlinear Systems using Active Disturbance Rejection Control

This paper explores a data-driven control approach for nonlinear systems, integrating Active Disturbance Rejection Control (ADRC) with concepts from chaotic systems theory. Many nonlinear systems, particularly chaotic ones, suffer from unknown disturbances and model mismatches, making traditional control strategies ineffective. By leveraging real-time data, our method continuously identifies dynamic behaviors, designs extended state observers (ESOs), and regulates chaotic responses efficiently. The proposed framework extends ADRC's traditional strengths by incorporating chaos control principles, allowing for better adaptability and robustness against system uncertainties. Unlike conventional feedback controllers, which often rely on precise mathematical models, our approach dynamically adjusts control actions using data-driven estimations. We validate our method through numerical simulations on benchmark chaotic systems, such as the Lorenz system, showing superior disturbance rejection and stability improvements over standard techniques. Results highlight enhanced robustness, making data-driven ADRC a promising solution for a variety of complex control problems, including those in robotics, fluid dynamics, and power systems. Our findings suggest that combining ADRC with machine learning-inspired adaptive mechanisms can significantly improve control performance for nonlinear and chaotic systems. Future work will explore hardware implementation and applications in real-world engineering systems.

کلیدواژه ها:

Active Disturbance Rejection Control ، Chaotic Systems ، Data-Driven Control ، Extended State Observer ، Nonlinear Systems