Stochastic Electric Vehicle Charging Optimization in Distribution Network

摘要

The integration of vehicle electrification into power grid calls for the precise estimation of electricity consumption. The inherent uncertainty and variability of electric vehicle charging load are insufficiently captured in the distribution system, in particular, integrated with distributed renewable energy resources. This paper presents a stochastic approach considering the correlations among the key variables and parameters based on Gaussian copula function. By deriving probabilistic charging patterns, it aims to minimize the overall electrical energy and the cost for performing charging activities in the power system. The proposed model formulates the correlation among the related random variables. Subsequently, the simulated charging profile is added so that the load demand is computed. The optimal charging scheduling problem is modeled using bus injection model and solved with a convex optimization relaxation technique. The case study shows that an accurate estimation of the randomness intrinsic to the charging patterns is critical to evaluate the charging impact on the distribution system.