A Combined H 2 / H ∞ Approach for Robust Joint Actuator and Sensor Fault Estimation: Application to a DC Servo-Motor...

Area of Science:

• Control Systems Engineering
• Fault Detection and Diagnosis
• Uncertainty Quantification

Background:

• Accurate fault estimation is crucial for reliable control system operation.
• Existing methods often struggle to address multiple, distinct uncertainty sources effectively.
• Uncertainty in sensor noise, external disturbances, and fault dynamics impacts estimation accuracy.

Purpose of the Study:

• To develop a robust fault estimation framework for actuators and sensors.
• To separately address different types of uncertainties including sensor noise, exogenous disturbances, and unknown fault dynamics.
• To minimize the impact of these uncertainties on fault estimation error in H2 or H-infinity norm.

Main Methods:

• A novel fault estimation framework is proposed.
• Uncertainties are categorized and treated distinctly: sensor measurement noise, process-external exogenous disturbances, and unknown fault dynamics.
• The approach minimizes fault estimation error using H2 or H-infinity optimization.

Main Results:

• A mixed performance-based actuator fault estimation framework was developed.
• Convergence conditions for the proposed framework were established.
• Performance analysis on a DC servo-motor and a three-tank system demonstrated the framework's superiority over an alternative method.

Conclusions:

• The proposed framework provides effective actuator and sensor fault estimation under various uncertainties.
• Separate treatment of uncertainty sources leads to improved estimation performance.
• The method is validated through simulations on benchmark systems, confirming its practical applicability and robustness.