Physics-informed mixture of experts network for interpretable battery degradation trajectory computation amid second-life complexities

2025-06-21

AI Summary

To address safety and scalability bottlenecks in second-life applications of retired EV batteries—stemming from uncertain degradation behavior and scarce historical data—this paper proposes a Physics-Informed Mixture-of-Experts (PI-MoE) neural network. The method integrates electrochemical prior knowledge with a multi-expert adaptive architecture, enabling degradation trajectory modeling from a single-cycle voltage/relaxation signal: capacity–voltage features drive degradation-mode classification and yield implicit trend representations; a scene-aware lightweight recurrent module supports long-horizon prediction without historical data. Evaluated on 207 batteries, PI-MoE achieves a mean absolute percentage error (MAPE) of 0.88% over 150 cycles, with a maximum MAPE of 6.26%. Inference time is 0.43 ms—halving both prediction error and computational cost relative to Informer and PatchTST. Moreover, the model trains and deploys effectively using only 5 MB of sample data.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems

Abstract

Retired electric vehicle batteries offer immense potential to support low-carbon energy systems, but uncertainties in their degradation behavior and data inaccessibilities under second-life use pose major barriers to safe and scalable deployment. This work proposes a Physics-Informed Mixture of Experts (PIMOE) network that computes battery degradation trajectories using partial, field-accessible signals in a single cycle. PIMOE leverages an adaptive multi-degradation prediction module to classify degradation modes using expert weight synthesis underpinned by capacity-voltage and relaxation data, producing latent degradation trend embeddings. These are input to a use-dependent recurrent network for long-term trajectory prediction. Validated on 207 batteries across 77 use conditions and 67,902 cycles, PIMOE achieves an average mean absolute percentage (MAPE) errors of 0.88% with a 0.43 ms inference time. Compared to the state-of-the-art Informer and PatchTST, it reduces computational time and MAPE by 50%, respectively. Compatible with random state of charge region sampling, PIMOE supports 150-cycle forecasts with 1.50% average and 6.26% maximum MAPE, and operates effectively even with pruned 5MB training data. Broadly, PIMOE framework offers a deployable, history-free solution for battery degradation trajectory computation, redefining how second-life energy storage systems are assessed, optimized, and integrated into the sustainable energy landscape.

Problem

• Predict battery degradation trajectories amid second-life uncertainties
• Classify degradation modes using partial field-accessible signals
• Enable long-term battery performance forecasting efficiently

Innovation

• Physics-Informed Mixture of Experts network
• Adaptive multi-degradation prediction module
• Use-dependent recurrent network for prediction