Probabilistic reliability evaluation of distribution systems considering the spatial and temporal distribution of electric vehicles (original) (raw)

Fast-growing electric vehicles load can impose various reliability concerns and peak demand increase in electric power systems, particularly distribution networks. From a power system point of view, electric vehicles can be considered as random moving loads. A new probabilistic approach is, therefore, proposed in this paper to evaluate the impact of electric vehicles on the reliability performance of power distribution systems. A two-layer stochastic electric vehicle charging demand estimation model is proposed. The model comprises of a traffic layer representing the spatial-temporal distributions of electric vehicles and an electrical network layer describing the electric vehicles charging demand. A Dynamic Hidden Markov Model is used to capture the electric vehicle movements in the traffic layer. Electric vehicle travel patterns and charging demand are simulated using a sequential Monte Carlo simulation approach considering the vehicle distance traveled, the type of charging location and the driver class. The proposed approach and the models are used to perform reliability studies on an example test system, and a series of analysis results are presented.