Thermal State Estimation of Energy Storage System Based on Integrated Long Short-term Memory Network

With the increasing popularity of energy storage, managing the dynamic thermal behavior of the energy storage system has become a profound yet challenging topic. To date, various energy storage system models have been proposed. Among these, the equivalent circuit model and the two-node thermal model are widely used because of their reduced computational complexity and high accuracy. The performance of the model, however, depends on an accurate estimation of model parameters, which are time-dependent and may vary with other factors such as temperature, state of charge, and aging of the energy storage system. In this paper, an Integrated Long Short-term Memory network (ILSTM) is modeled to address efficiently the challenge of estimating variable parameters. The proposed ILSTM model is proved to be able to predict the time-varying parameters of the thermal model quickly and accurately.