Learning the P2D Model for Lithium-Ion Batteries with SOH Detection

2025-02-19

AI Summary

To address the challenge of real-time performance prediction and online State-of-Health (SOH) monitoring of lithium-ion batteries under stochastic driving conditions, this paper proposes an interpretable, physics-informed CNN surrogate modeling framework tailored for the pseudo-two-dimensional (P2D) electrochemical model. For the first time, a CNN is designed as a physically consistent and differentiable surrogate of the P2D model, integrating high-fidelity simulation data generated under randomized operating conditions with an SOH-aware dynamic fine-tuning mechanism to enable efficient and accurate modeling of species concentration distributions. The proposed method achieves concentration prediction errors below 1.2% while accelerating computation by over 200× compared to conventional P2D solvers. Crucially, it maintains robustness across progressive SOH degradation, significantly enhancing both real-time electrochemical simulation capability and adaptive diagnostic performance in battery management systems (BMS).

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Probabilistic Circuits and Graphical Models
• Reasoning under Uncertainty: Stochastic Optimization

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: User modeling and simulation for interactive and conversational systems

Abstract

Lithium ion batteries are widely used in many applications. Battery management systems control their optimal use and charging and predict when the battery will cease to deliver the required output on a planned duty or driving cycle. Such systems use a simulation of a mathematical model of battery performance. These models can be electrochemical or data-driven. Electrochemical models for batteries running at high currents are mathematically and computationally complex. In this work, we show that a well-regarded electrochemical model, the Pseudo Two Dimensional (P2D) model, can be replaced by a computationally efficient Convolutional Neural Network (CNN) surrogate model fit to accurately simulated data from a class of random driving cycles. We demonstrate that a CNN is an ideal choice for accurately capturing Lithium ion concentration profiles. Additionally, we show how the neural network model can be adjusted to correspond to battery changes in State of Health (SOH).

Problem

• Replacing P2D model with efficient CNN
• Accurately simulate battery performance
• Adjusting CNN for SOH detection

Innovation

• CNN replaces P2D model
• Efficient Lithium ion simulation
• SOH detection via CNN