Citation

Li M, Dong C, Mu Y, Yu X, Xiao Q, Jia H. Patterns (N Y) 2022; 3(2): e100432.

Copyright

(Copyright © 2022, Cell Press)

DOI

10.1016/j.patter.2021.100432

PMID

35199068

PMCID

PMC8848029

Abstract

As an important type of energy storage, battery energy storage systems have been widely used. However, there are frequent cases of battery explosion due to high temperature. To address this issue, researches have been carried out either in the model-driven or the data-driven aspects to predict the temperature of the battery. In this paper, a two-node electrothermal model and a multi-scale long short-term memory network are established formulating a data-model alliance network (DMAN) for surface temperature diffusion. An improved adaptive boosting algorithm is then employed to enhance the bridge of the two models. Integrating a data-model alliance module (DMAM) and multi-step-ahead thermal warning network (MATWN), this DMAN provides an advanced online multi-step-ahead thermal warning structure to achieve early warning of temperature crossing. Experimental results verify the progressiveness of the proposed technique.

Language: en

Keywords

data-model alliance network; long short-term memory network; thermal warning