Battery degradation model and multiple-indicators based lifetime estimator for energy storage system design and operation: Experimental analyses of cycling-induced aging

Batteries used in battery energy storage system (BESS) have a wide lifetime and fast aging process considering the secondary-use applications. The dispersion of the batteries rises rapidly with aging, leading to a decrease in the robustness of the lifetime estimators. In this paper, a novel multiple health indicators (MHIs) system-based battery lifetime estimator, which contains six health indicators (HIs) with different characteristics is proposed. The Back Propagation Neural Network (BPNN) is used to train the relationship between the HIs and lifetime to reduce the dispersion of different batteries. In addition, an empirical degradation model for low-capacity batteries considering different usage factors is proposed, which is significant for optimized design of BESS. Cycling-induced aging tests with different depth of discharge (DOD) and mean state-of-charge (SOC) are performed to verify the accuracy and robust of the proposed estimator. The average errors of the tested batteries are all less than 1.5%, which shows a good performance on accuracy and robustness. (c) 2021 Published by Elsevier Ltd.

Automated summary

This paper proposes a novel battery lifetime estimator based on multiple health indicators, utilizing six different characteristics to train a neural network model in order to improve accuracy and robustness of battery lifetime estimation. Additionally, an empirical degradation model for low-capacity batteries considering different usage factors is introduced, which is significant for the optimized design of BESS.