Copula-based multivariate stochastic modeling of load demand due to plug-in electric vehicles in order to be integrated in distribution system planning

This paper develops a multivariate probabilistic framework for PEV load modelling to be embedded in system planning problems. In order to successfully integrate the uncertainty attributes of the PEVs in the probabilistic planning issues, relevant vehicular load scenarios is provided through appropriate synthetic data. A student´s t copula distribution function is utilized to capture the correlation characteristics among the included datasets namely home departure time, daily travelled distances and home arrival time of the vehicles during weekdays. Then, a Monte Carlo based stochastic simulation is provided to derive hourly load distribution functions of the PEVs. Extraction of the demand profile of the individual PEVs is fulfilled in order to estimate the demand profile of the fleet. The estimated probability distribution functions can be efficiently employed to generate load samples in probabilistic distribution system planning problems.