Output-Feedback Robust Control of Uncertain Systems via Online Data-Driven Learning.

Area of Science:

• Control Theory
• Systems Engineering
• Applied Mathematics

Background:

• Robust control design, particularly output-feedback, remains a significant challenge in control engineering.
• Existing methods often require full state information, which is not always available.

Purpose of the Study:

• To develop a new approach for output-feedback robust control of continuous-time uncertain systems.
• To simplify the control design process and enable online learning of control gains.

Main Methods:

• Transforming the robust control problem into an optimal control problem with a constructive cost function.
• Constructing a modified algebraic Riccati equation (MARE) and reformulating the output Lyapunov function using vectorization and Kronecker's product.
• Developing a novel online data-driven learning method to solve the MARE using only measurable system inputs and outputs.

Main Results:

• The output-feedback robust control gain is obtained without requiring unknown system states.
• The stability of the control system and the convergence of the learned solution are rigorously proven.
• The method's efficacy is demonstrated through two simulation examples.

Conclusions:

• The proposed data-driven online learning method effectively addresses output-feedback robust control for uncertain systems.
• This approach simplifies control design and enhances practical applicability by relying solely on measurable data.