A Hybrid Model-Data-Driven Approach for Calculating the Dispatchable Potential of Electric Vehicle Cluster Energy Storage

Aiming at the current problem that it is difficult to calculate the dispatchable potential of EV cluster energy storage accurately, this paper proposes a model-data hybrid-driven method for calculating the dispatchable potential of EV cluster energy storage based on the Minkowski summation theory and the temporal pattern attention mechanism with long and short-term memory network (TPA-LSTM). Firstly, it aggregates EV clusters into generalized energy storage (GES) based on Minkowski summation theory. Then, based on TPA-LSTM, the EV cluster stopping data and the parameters of the EV self-scheduling model are obtained, and then the GES aggregation model is established to calculate the dispatch capacity and dispatch power of GES. Finally, the quantitative results of the dispatchable potential of EV cluster energy storage are obtained through simulation analysis, thus verifying the practicality and efficacy of the suggested approach.