A Locating Method of EV Fast Charging Stations Based on the Conditional Generative Adversarial Networks

The planning and locating of charging stations is of great significance for the development of EV. The existed methods are often carried out by solving an optimal model, which requires a large amount of user data to simulate user behavior. Considering that a large number of planning schemes are available, this paper proposes a construction planning method of EV charging station based on the conditional Generative Adversarial Networks, aiming to find a quick and convenient planning method and thus reduce the demand for user's data under the condition of ensuring the rationality of the results. The pix2pix problem was solved by cGANs, and the network structure was adjusted reasonably to avoid over-fitting. The construction diagram of the charging station was captured for training, and the network achieved good results in both the training set and the test set.