Distributed Robust Fault Estimation Using Relative Measurements for Leader-Follower Multiagent Systems.

Area of Science:

• Control Systems Engineering
• Distributed Systems
• Fault Diagnosis

Background:

• Leader-follower multiagent systems are crucial in various applications.
• Robust fault estimation (FE) is essential for system safety and reliability.
• Existing methods may not fully address distributed robust FE using relative measurements.

Purpose of the Study:

• To develop a distributed robust fault estimation scheme for leader-follower multiagent systems.
• To enhance estimator robustness using H-infinity performance criteria.
• To propose a fully distributed fault estimation algorithm.

Main Methods:

• A distributed intermediate-based fault estimator is designed using local relative measurements and neighbor state estimations.
• Gain matrices are determined via linear matrix inequality (LMI) optimization for H-infinity performance.
• The LMI is simplified using spectral decomposition, with an equivalent condition based on eigenvalue analysis.
• A distributed eigenvalue estimation algorithm employing the power method is presented.

Main Results:

• The proposed fault estimator effectively utilizes local relative measurements and neighbor information.
• H-infinity performance guarantees robustness against uncertainties and disturbances.
• The spectral decomposition and eigenvalue-based condition simplify the design and enable full distribution.
• Numerical simulations validate the effectiveness and robustness of the distributed fault estimation scheme.

Conclusions:

• The developed distributed fault estimation scheme is effective for leader-follower multiagent systems.
• The H-infinity based LMI approach ensures robust performance.
• The proposed distributed algorithm allows for scalable and practical implementation.