A Capacity Planning Method for EV Charging Stations Based on Fuzzy C-Means Clustering and Deep Extreme Learning Machine

Strategic Goal Based on Carbon Peak and Carbon Neutrality, the number of electric vehicles (EVs) has significantly increased and connected to the distribution network. The base of electric vehicles does not match the rapid penetration speed of electric vehicles. In response to the difficulty in determining the declared capacity of new charging stations, this paper proposes a charging station capacity planning method based on quadratic clustering theory combined with deep-limit learning machines. Firstly, a fuzzy C-means clustering is performed on the charging station dataset to obtain four sets with high similarity in daily charging loads; Then perform secondary clustering on the charging stations after completing the first clustering, so that they can form new clustering groups according to different scales; Finally, based on the new clustering group and the DELM algorithm, the daily load curve of the charging station to be built was predicted, and comparative experiments were conducted. The results showed that the method proposed in this paper can achieve accurate planning of charging station capacity.