Data-driven output-feedback fault-tolerant L2 control of unknown dynamic systems

Area of Science:

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

Background:

• Unknown dynamic systems pose challenges for traditional control design.
• Active fault-tolerant control (FTC) requires robust fault detection and controller adaptation.
• L2-control aims to minimize the impact of disturbances on system performance.

Purpose of the Study:

• To develop a data-driven output-feedback fault-tolerant L2-control strategy for unknown dynamic systems.
• To integrate fault detection, controller reconfiguration, and tracking control within an active FTC framework.
• To achieve guaranteed cost control and L2-gain properties using only input-output data.

Main Methods:

• Utilizing observer-based residual generators in a data-driven framework to express system states from input-output data.
• Applying a model-free approach for L2 control of unknown linear time-invariant (LTI) discrete-time plants.
• Designing a pre-filter for tracking control and employing a data-driven time-varying value function approximation for FTC scheme development.

Main Results:

• A data-driven, model-free L2-control method for unknown LTI discrete-time systems is established.
• An active FTC scheme with guaranteed L2-gain properties is developed using input-output data.
• The proposed methodology demonstrates effectiveness through two simulation examples.

Conclusions:

• The developed data-driven FTC approach effectively addresses fault detection, controller reconfiguration, and tracking control for unknown systems.
• The model-free strategy ensures robust performance and L2-gain properties without requiring explicit system models.
• The methodology offers a practical solution for real-world applications where system dynamics are not precisely known.