Research on EV Schedulable Capability Prediction Method Based on Data-model Hybrid Drive

As the number of EV continues to increase, it is increasingly important to stimulate the release of schedulable resources on the demand side, let it participate in the optimal regulation of power system. Aiming at the problem of EV cluster schedulable capability prediction, this paper proposes a data-model hybrid driven EV schedulable capability calculation method. Firstly, the EV cluster is aggregated into GES devices by Minkowski summation method. Then, Bi-LSTM is used to mine the parameters of GES equipment and predict them. Finally, the prediction results of GES equipment parameters are put into the EV schedulable capability model. The simulation results show that Bi-LSTM can better mine the rules of time series data and accurately predict the parameters of GES model. The proposed method can aggregate EV clusters into GES devices, reduce the model dimension, and realize the quantification of schedulable capability of EV clusters.