Towards a Probabilistic Fusion Approach for Robust Battery Prognostics

TL;DR

This work tackles the challenge of robust battery SOH/RUL prognostics by introducing a probabilistic fusion framework that ensembles Bayesian Convolutional Neural Networks (BCNNs) via stacking of predictive distributions. Offline, multiple BCNNs are trained with a leave-one-out strategy to maximize diversity; online, their predictive PDFs are combined using log-score weights to yield a calibrated one-step-ahead forecast for capacity at $t+1$. The approach improves accuracy and uncertainty quantification over a single BCNN and a point-prediction stacking baseline, as demonstrated on NASA Ames battery aging data, with notable gains in metrics such as NLL and CRPS and better calibration and sharpness. This probabilistic fusion method enhances reliability for battery health monitoring in autonomous systems and provides a foundation for scalable, uncertainty-aware prognostics in real-world deployments.

Abstract

Batteries are a key enabling technology for the decarbonization of transport and energy sectors. The safe and reliable operation of batteries is crucial for battery-powered systems. In this direction, the development of accurate and robust battery state-of-health prognostics models can unlock the potential of autonomous systems for complex, remote and reliable operations. The combination of Neural Networks, Bayesian modelling concepts and ensemble learning strategies, form a valuable prognostics framework to combine uncertainty in a robust and accurate manner. Accordingly, this paper introduces a Bayesian ensemble learning approach to predict the capacity depletion of lithium-ion batteries. The approach accurately predicts the capacity fade and quantifies the uncertainty associated with battery design and degradation processes. The proposed Bayesian ensemble methodology employs a stacking technique, integrating multiple Bayesian neural networks (BNNs) as base learners, which have been trained on data diversity. The proposed method has been validated using a battery aging dataset collected by the NASA Ames Prognostics Center of Excellence. Obtained results demonstrate the improved accuracy and robustness of the proposed probabilistic fusion approach with respect to (i) a single BNN model and (ii) a classical stacking strategy based on different BNNs.

Figures (9)

• Figure 1: High-level block diagram of the proposed approach.
• Figure 2: Block diagram of the proposed approach.
• Figure 3: Schematic of the Bayesian convolution neural network.
• Figure 4: Feature variations due to an increasing number of discharge cycles in battery #5.
• Figure 5: Capacity degradation data of Li-ion batteries.