H∞ state estimation for complex networks with uncertain inner coupling and incomplete measurements

Area of Science:

• Control Theory
• Network Systems
• Signal Processing

Background:

• Complex networks are susceptible to uncertain coupling strengths.
• Incomplete measurements (saturation, quantization, missing data) degrade state estimation.
• Existing methods struggle to unify and address diverse measurement uncertainties.

Purpose of the Study:

• To develop H∞ state estimators for complex networks with uncertain coupling and random incomplete measurements.
• To characterize uncertainties in the inner coupling matrix using the interval matrix approach.
• To propose a unified measurement model for various incomplete measurement phenomena.

Main Methods:

• Utilized the interval matrix approach to characterize coupling uncertainties.
• Introduced a stochastic Kronecker delta function for a unified measurement model.
• Designed H∞ state estimators ensuring exponential mean-square stability and H∞ performance.
• Employed convex optimization and semidefinite programming for estimator gain calculation.

Main Results:

• Successfully characterized uncertainties in the inner coupling matrix.
• Developed a novel measurement model unifying random sensor saturations, quantization, and missing data.
• Designed H∞ state estimators that guarantee stability and performance under considered uncertainties.
• Demonstrated the effectiveness of the proposed approach via numerical simulations.

Conclusions:

• The proposed H∞ state estimation method effectively handles uncertain coupling and diverse random incomplete measurements in complex networks.
• The interval matrix approach and unified measurement model provide a robust framework for dealing with system uncertainties.
• The designed estimators are applicable and effective, as validated by simulation results.