Electric Vehicle Charging Load Forecasting Based on Improved Gipps and Its Impact on the Distribution Network

This paper proposes an improved method based on the Gipps model for predicting electric vehicle (EV) charging load to address the spatiotemporal uncertainty and randomness of electric vehicle charging load. Firstly, using the Markov process to simulate the spatial transfer characteristics of electric vehicles, and combining it with trip chaining theory to describe the travel patterns of electric vehicles. Secondly, the Dijkstra algorithm plans vehicle travel routes to get the distance of each journey segment. The log-normal probability distribution functions are fitted to estimate the dwell time, and a spatiotemporal distribution model for daily vehicle travel is established. Then, based on the improved Gipps and taking into account the influence of road class, temperature and other factors on the electricity consumption of EVs, micro traffic analysis is carried out on the driving process of vehicles. Finally, the EV charging load is input to the corresponding 33-node distribution network, and the sequential power flow algorithm is used to evaluate the impact of EV charging on the distribution network. The results show the correctness and feasibility of the proposed model.