Multi-Sensor Consensus Estimation of State, Sensor Biases and Unknown Input.

Area of Science:

• Control Systems Engineering
• Signal Processing
• Networked Systems

Background:

• Heterogeneous sensor networks face challenges in joint state and bias estimation due to evolving biases and common unknown inputs (UI).
• Accurate estimation is crucial for reliable data fusion and decision-making in distributed systems.

Purpose of the Study:

• To propose a novel distributed filtering approach for joint estimation of system state and generalized sensor bias (GSB) under unknown inputs.
• To address bias evolution and enable robust state and bias estimation in heterogeneous sensor networks.

Main Methods:

• Derivation of an equivalent UI-free GSB dynamic model for local estimation at each sensor node.
• Utilizing neighbor information for UI estimation via least-squares and state fusion through consensus processing.
• Refining multi-sensor bias estimates based on the consensus UI estimate.

Main Results:

• The proposed method effectively performs joint estimation of system state and GSB in the presence of UI.
• Demonstrated improved accuracy in distributed multi-sensor target tracking through a numerical example.
• Validation of the filter's performance in a complex networked sensing scenario.

Conclusions:

• The developed distributed filtering strategy provides a robust solution for joint state and GSB estimation in heterogeneous sensor networks.
• The method effectively handles bias evolution and unknown inputs, enhancing tracking accuracy.
• The approach is suitable for real-world applications requiring reliable distributed sensing and estimation.