ABSTRACT
The COVID-19 pandemic has forced many governments around the world to implement strict lockdown measures and order citizens to stay at home, which has caused a major change in travel patterns. This study leveraged electric vehicle charging big data in Hefei, Anhui Province, China to estimate electric vehicle charging demand in the absence of the COVID-19 pandemic using multi-layer perceptron model, which quantified the impact of the COVID-19 pandemic. In addition, we employed the vector autoregressive model to investigate the dynamic relationships between the changes in charging demand and various explanatory factors. The results suggest that the daily average charging demand in Hefei decreased by 78.3% compared to the predicted value during the pandemic. Furthermore, according to the variance decomposition and impulse response function analysis, national confirmed COVID-19 cases play a dominant role in reducing charging demand. The number of daily hospitalizations and Migration Scale Index also have significant and robust effect on the decrease in charging demand. The Air Quality Index and Baidu Index are susceptible to external factors and do not have a direct impact on the change in charging demand. Findings support a better understanding of changes in travel behavior during the pandemic and provide policy makers with references to deal with similar events. © 2021 IEEE