NDKF: A Neural-Enhanced Distributed Kalman Filter for Nonlinear Multi-Sensor Estimation
This work addresses distributed state estimation for nonlinear systems, offering a solution to communication and centralization bottlenecks, but it appears incremental as it builds on existing Kalman filter and neural network methods.
The paper tackles multi-sensor state estimation in nonlinear systems by proposing a neural-enhanced distributed Kalman filter that replaces analytical models with learned mappings, reducing communication overhead and avoiding central bottlenecks. Simulations on a 2D nonlinear system show it outperforms a distributed EKF baseline under model mismatch with improved accuracy and modest communication requirements.
We propose a Neural-Enhanced Distributed Kalman Filter (NDKF) for multi-sensor state estimation in nonlinear systems. Unlike traditional Kalman filters that rely on explicit analytical models and assume centralized fusion, NDKF leverages neural networks to replace analytical process and measurement models with learned mappings while each node performs local prediction and update steps and exchanges only compact posterior summaries with its neighbors. This distributed design reduces communication overhead and avoids a central fusion bottleneck. We provide sufficient mean-square stability conditions under bounded Jacobians and well-conditioned innovations, together with practically checkable proxies such as Jacobian norm control and innovation monitoring. We also discuss consistency under learned-model mismatch, including covariance inflation and covariance-intersection fusion when cross-correlations are uncertain. Simulations on a 2D nonlinear system with four partially observing nodes show that NDKF outperforms a distributed EKF baseline under model mismatch and yields improved estimation accuracy with modest communication requirements.