Electric Vehicle Charging Demand Forecasting: Framework and Practical Exemplary Study

While widespread relevant policy or incentives play an active role in promoting massive Electric Vehicle (EV) adoption, EV charging demand forecasting becomes a concern impacting power grid management if EV charging stations are inappropriately planned spatially over a region. Forecasting EV charging demand forecasting continues to be an imperative task as the EVs deployment is evolving. To address the pressing need, this paper presents a research methodology for innovating EV charging forecasting models and streamlining them into computation algorithms to facilitate analyzing the overall EV charging demand and distribution problems over a certain area. In this study, the bottom-up heuristic approach is explored to construct the methodological framework for integrating the developed EV Activity-based Travel Demand Forecasting (EVA-TDF) models into the analysis procedure through running trip-chain simulations. The greater Cincinnati area in Ohio is running as a case study site to test the effectiveness of the proposed methodology. The testing results suggested that the heuristic EVA-TDF framework make it easier to check the accuracy at each step for the sequential flow of information with adaptability to scalable areas. Additionally, the results revealed that the accuracy of the estimated NHB trips greatly impact the accuracy of overall travel demand forecasting for the Cincinnati area, and the secondary important impact factor is HBO trips followed by HBW trips. TAZs within the CBD or with high population density in urban area may be the best candidate locations for deploying Level 1 and 2 EV charging stations.