Download PDFOpen PDF in browser

Modeling Thermal Management of Battery Energy Storage System with Machine Learning

EasyChair Preprint no. 9718

9 pagesDate: February 15, 2023

Abstract

Battery energy storage systems (BESS) are nowadays essential parts of microgrids. A thermal management system (TMS) belongs to substantial control components ensuring optimal operation and long lifespan of batteries. Advanced control strategies implemented in TMS require accurate thermal models to keep battery temperature within predefined bounds while minimizing operating costs. This paper proposes machine learning-based models to predict temperature inside real industrial BESS. Challenges represent partially continuous and partially discrete input signals. Furthermore, inner fans located inside modules affect the temperature in this particular BESS. Unfortunately, the information on fans’ operations is not available. This study also provides an accuracy analysis of bagged classification and regression trees (CART), multi-layer perceptron (MLP), and averaged neural network (avNNet). The results report high prediction accuracy, over 95%, for all models, even the ones with a more straightforward structure.

Keyphrases: Battery Energy Storage System, machine learning, modeling, thermal management

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:9718,
  author = {Amir Mosavi and F. Kristina},
  title = {Modeling Thermal Management of Battery Energy Storage System with Machine Learning},
  howpublished = {EasyChair Preprint no. 9718},

  year = {EasyChair, 2023}}
Download PDFOpen PDF in browser