Volume 21: Sustainable Energy Solutions for a Post-COVID Recovery towards a Better Future: Part IV

Data-driven Agent Modeling for Liquid Air Energy Storage System with Machine Learning: A Comparative Analysis Fang Yuan, Zhongxuan Liu, Yuemin Ding



With the wide adoption of renewable energy resources in the power grid, energy storage systems have drawn significant attention to improving the stability and efficiency of the power grid. Among various storage systems, Liquid Air Energy Storage (LAES) has a promising future due to its intrinsic advantages. However, the modeling of a LAES is a complex issue, and existing approaches based on principles have a heavy computational load. To facilitate modeling of LAES, this study focused on data-driven modeling with machine learning and conducted a comparative analysis for several popular methods, including K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Artificial Neural Network (ANN), and Deep Neural Networks (DNN). With LAES as the study case, data-driven models were built based on the data generated by its first-principal model developed with the Aspen HYSYS simulation software. For the selected machine learning methods, the modeling accuracy and running time were compared, showing that the DNN achieved the best performance compared to the others.

Keywords Liquid Air Energy Storage, Machine Learning, Data-driven modeling, Comparative Analysis

Copyright ©
Energy Proceedings