Optimal Control of PEV Charging Based on Residential Base Load Prediction

Due to the characteristics of electric power generation, transmission and distribution in the U.S., experts have identified local distribution as a likely component of the chain to be adversely affected by unregulated PEV (Plug-in Hybrid Electric Vehicle/Electric Vehicle) charging. This paper studied the optimal control of PEV charging considering the loss of life of residential distribution transformer based on the prediction of residential base load. The prediction was studied using an autoregressive moving average (ARMA) model. A dynamic thermal model was used for prediction of the transformer hot-spot temperature with the information on power load and ambient temperature. The loss of life model including the dynamic thermal model was integrated into the optimization problem. The objective of the optimal control is to minimize the loss of life of transformer. Finally, a genetic algorithm (GA) was used for searching the optimal solution of PEV charging based on the predicted base load.