Robust ellipsoidal set-membership fault estimation for time-varying systems with uniform quantization effects over...

Area of Science:

• Control Systems Engineering
• Signal Processing
• Networked Systems

Background:

• Sensor networks are crucial for data acquisition but face challenges from quantization and system uncertainties.
• Fault estimation is vital for system reliability and safety in networked control systems.
• Uniform quantization introduces unknown-but-bounded (UBB) noises, complicating fault estimation.

Purpose of the Study:

• To design a robust set-membership fault estimator for uncertain discrete time-varying systems operating over sensor networks.
• To address the challenges posed by uniform quantization and UBB noises in fault estimation.
• To ensure that estimation errors are confined within a specified ellipsoidal region.

Main Methods:

• Utilizing mathematical induction to establish sufficient conditions for fault estimator existence at each time step.
• Formulating a recursive set-membership approach based on matrix inequalities.
• Developing an optimization problem to minimize the size of the estimation error ellipsoid.

Main Results:

• A sufficient condition for the existence of the robust fault estimator is derived.
• The proposed method effectively confines estimation errors to an ellipsoidal region.
• Numerical examples demonstrate the practical effectiveness of the fault estimation scheme.

Conclusions:

• The developed fault estimator provides a robust solution for uncertain discrete time-varying systems with quantization effects.
• The set-membership approach guarantees bounded estimation errors, enhancing system fault diagnosis.
• The optimization strategy effectively minimizes the conservatism of the fault estimation.