Towards robust fault detection for Takagi-Sugeno nonlinear systems

Area of Science:

• Control Systems Engineering
• Nonlinear Systems Analysis
• Fault Detection and Diagnosis

Background:

• Robust fault detection is crucial for the reliable operation of discrete-time nonlinear systems.
• Existing methods may exhibit conservatism or suboptimal performance in characterizing fault sensitivity and disturbance robustness.
• Takagi-Sugeno (T-S) fuzzy models are widely used to represent nonlinear systems.

Purpose of the Study:

• To develop a novel robust fault detection observer design for discrete-time Takagi-Sugeno nonlinear systems.
• To improve fault sensitivity and disturbance robustness using finite-frequency domain techniques.
• To reduce the conservatism associated with observer design.

Main Methods:

• Utilizing finite-frequency H_- indices to quantify fault sensitivity.
• Employing finite-frequency H_∞ indices for disturbance robustness characterization.
• Applying peak-to-peak analysis to derive a dynamic threshold for fault detection.
• Developing an iterative algorithm to minimize design conservatism.

Main Results:

• The proposed method achieves robust fault detection observer design for discrete-time T-S nonlinear systems.
• Theoretical analysis confirms that the new method's performance is comparable to or better than existing approaches.
• Simulation results validate the effectiveness and practicality of the developed fault detection strategy.

Conclusions:

• The novel finite-frequency H_-/H_∞ and peak-to-peak analysis based method offers a robust solution for fault detection in discrete-time T-S nonlinear systems.
• The iterative algorithm effectively reduces conservatism, enhancing the observer's performance.
• The proposed approach provides a viable and effective tool for ensuring system safety and reliability.