Model-Driven Dataset Generation for Data-Driven Battery SOH Models

Khaled Sidahmed Sidahmed Alamin,Francesco Daghero,Giovanni Pollo,Daniele Jahier Pagliari,Yukai Chen,Enrico Macii,Massimo Poncino,Sara Vinco

DOI: https://doi.org/10.1109/ISLPED58423.2023.10244587

2024-01-11

Abstract:Estimating the State of Health (SOH) of batteries is crucial for ensuring the reliable operation of battery systems. Since there is no practical way to instantaneously measure it at run time, a model is required for its estimation. Recently, several data-driven SOH models have been proposed, whose accuracy heavily relies on the quality of the datasets used for their training. Since these datasets are obtained from measurements, they are limited in the variety of the charge/discharge profiles. To address this scarcity issue, we propose generating datasets by simulating a traditional battery model (e.g., a circuit-equivalent one). The primary advantage of this approach is the ability to use a simulatable battery model to evaluate a potentially infinite number of workload profiles for training the data-driven model. Furthermore, this general concept can be applied using any simulatable battery model, providing a fine spectrum of accuracy/complexity tradeoffs. Our results indicate that using simulated data achieves reasonable accuracy in SOH estimation, with a 7.2% error relative to the simulated model, in exchange for a 27X memory reduction and a =2000X speedup.

Signal Processing

What problem does this paper attempt to address?

This paper focuses on the problem of estimating the State of Health (SOH) of batteries, which is crucial for ensuring the safe and reliable operation of battery systems, especially in electric vehicles (EVs). Since the SOH cannot be directly measured in real-time, a model needs to be established to estimate it. In recent years, data-driven SOH models have become popular, relying on high-quality datasets for training. However, these datasets are limited by the diversity and quantity of measured data.

To address this issue, the paper proposes a model-driven dataset generation method, which generates a large number of different charging/discharging workload data by simulating traditional battery models (such as circuit equivalent models) to train data-driven models. The advantages of this method include the ability to generate an unlimited number of data points at a lower cost and time, provide adjustable accuracy/complexity trade-offs, and allow the combination of multiple battery models to improve the comprehensiveness of the final model. Additionally, using simulated data for SOH estimation has a relatively small error of around 7.2% compared to simulated models, while achieving a 27-fold reduction in memory usage and approximately 2000-fold speed improvement.

The paper also discusses battery aging, data-driven SOH estimation methods, and the pros and cons of different models. Finally, they propose a workflow to build a simulatable battery model from available battery data and then use this model to generate additional data points for training a higher-precision data-driven SOH model.

The experimental results show that the error in SOH estimation using the selected simulated model-generated data is about 7%, while significantly reducing memory usage and computation time. This demonstrates the effectiveness and flexibility of using simulated data for building data-driven models.