Fault Estimation and Control for Unknown Discrete-Time Systems Based on Data-Driven Parameterization Approach.

Area of Science:

• Control Systems Engineering
• Systems Theory
• Data-Driven Modeling

Background:

• Unknown discrete-time systems pose challenges for fault estimation and robust control.
• Existing methods often require model-based approaches or initial stable controllers.
• Multiobjective optimization is crucial for balancing fault estimation and control performance.

Purpose of the Study:

• To develop a data-driven controller design method for fault estimation and control in unknown discrete-time systems.
• To formulate the problem as a multiobjective H_∞/H_∞ optimization.
• To optimize both fault estimation and robust control performances simultaneously.

Main Methods:

• Formulation as an H_∞/H_∞ multiobjective optimization problem.
• Development of a data-driven parameterization controller design.
• Introduction of slack variables to reduce conservatism in optimization.
• Direct parameterization of controller gain using state and input data.

Main Results:

• The data-driven method achieves performance consistent with model-based H_∞ control.
• Reduced conservatism in solving the multiobjective optimization problem.
• Elimination of the requirement for an initial stable controller.
• Demonstrated effectiveness and advantages through simulation results.

Conclusions:

• The proposed data-driven method offers an effective approach for fault estimation and control in unknown discrete-time systems.
• The method provides a robust and less conservative alternative to existing techniques.
• Direct parameterization using system data simplifies controller design and enhances applicability.