CKFNet: Neural Network Aided Cubature Kalman filtering

TL;DR

CKFNet enhances the cubature Kalman filter by integrating RNNs to adapt to model uncertainties, improving accuracy and robustness in nonlinear estimation tasks.

Contribution

This paper introduces CKFNet, a hybrid neural network architecture that combines RNNs with CKF, maintaining interpretability while improving performance under model mismatches.

Findings

CKFNet outperforms traditional CKF in accuracy.

CKFNet demonstrates robustness against model-environment mismatches.

Numerical simulations validate the effectiveness of CKFNet.

Abstract

The cubature Kalman filter (CKF), while theoretically rigorous for nonlinear estimation, often suffers performance degradation due to model-environment mismatches in practice. To address this limitation, we propose CKFNet-a hybrid architecture that synergistically integrates recurrent neural networks (RNN) with the CKF framework while preserving its cubature principles. Unlike conventional model-driven approaches, CKFNet embeds RNN modules in the prediction phase to dynamically adapt to unmodeled uncertainties, effectively reducing cumulative error propagation through temporal noise correlation learning. Crucially, the architecture maintains CKF's analytical interpretability via constrained optimization of cubature point distributions. Numerical simulation experiments have confirmed that our proposed CKFNet exhibits superior accuracy and robustness compared to conventional model-based methods and existing KalmanNet algorithms.