Adaptive federated filter for multi-sensor nonlinear system with cross-correlated noises

Area of Science:

• Control Systems Engineering
• Signal Processing
• Nonlinear System Analysis

Background:

• Federated filters are crucial for multi-sensor data fusion in nonlinear systems.
• Cross-correlations between process and measurement noise degrade filter performance.
• Existing methods struggle with correlated noise in nonlinear systems.

Purpose of the Study:

• To develop an adaptive federated filter approach for nonlinear systems with cross-correlated noises.
• To introduce the adaptive Gaussian filter as a local filter within the federated filter framework.
• To propose and theoretically validate two novel adaptive federated filter algorithms.

Main Methods:

• Utilizing an adaptive Gaussian filter as the local filter to mitigate noise impact.
• Developing two adaptive federated filter variants: one with a de-correlation framework, another with an epoch-correlated Gaussian subfilter.
• Verifying the theoretical equivalence of the proposed algorithms in nonlinear fusion systems.

Main Results:

• The proposed adaptive federated filters significantly outperform traditional federated and Gaussian filters when dealing with correlated noises.
• Simulation results confirm the superior performance and robustness of the adaptive approaches.
• Theoretical equivalence between the two proposed algorithms and the high-degree cubature federated filter is demonstrated.

Conclusions:

• The novel adaptive federated filters effectively address performance degradation caused by cross-correlated noises in nonlinear systems.
• The proposed methods offer enhanced accuracy and robustness compared to existing techniques.
• The study validates the theoretical underpinnings and practical applicability of the adaptive federated filtering approach.