Electric vehicles (EVs) have great impacts on power distribution networks due to their spatio-temporal characteristic, especially in urban areas. This paper proposes a framework for optimal planning of urban distribution network (UDN) and electric vehicle charging stations (EVCSs), in which the geographic information system (GIS) is combined to optimize investment decisions. By using hierarchical clustering and Voronoi diagram, the planning area is divided into several sub-areas. All the sub-area is converted into grid networks based on GIS, on which the EVs’ actual temporal charging demand is analyzed. The proposed mixed integer planning model utilizes the second-order cone programming for UDN and grid-based site selection of EVCS by considering the spatio-temporal characteristic of EVs. Case studies on an urban area in Shanghai, China demonstrate the effectiveness of the proposed method.
Keywords electric vehicle, urban distribution network, mixed integer planning, geographic information system