Integrated Bayesian networks with GIS for electric vehicles charging site selection

Location selection of charging stations of electric vehicles (EVs) contributes to long-term sustainable urban development. This study proposes a hybrid approach integrated with the Geographical information system (GIS) and Bayesian network (BN) to deal with the location selection problem for EVs. GIS serves for capturing spatial and geographical data, which provides dynamic and visual information for selecting charging sites. BN is employed to process various criteria and demonstrate the cause-effect relationship in alternative site selection. A BN model consisting of nine criteria from three aspects is established to determine the most suitable locations for charging stations of EVs. A total of ten alternative locations in Singapore is used to verify the applicability and effectiveness of the developed hybrid approach. Results indicate that (1) Criteria, including the number of MRT stations, household units, and charging efficiency, are identified as the most sensitive factors to the location selection; (2) The transportation efficiency has the strongest linkage with the location selection (with an average value of the strength of 0.445), revealing that the transportation efficiency is more important than the social and economic efficiency. The novelty of this research lies in the development of the hybrid GIS-based BN approach that is more accurate and stable under noise interruption compared to the traditional decision-making method (e.g., TOPSIS). The developed approach can be used as a decision tool to identify the major contributing factors and update the optimal decisions given new observation data in GIS in an automatic manner.