A driver behavior decision model on the user side of electric vehicle (EV) charging platform was proposed based on the random utility maximization theory. Under the requirements of vehicle service and the constraints of charging stations, the Karush-Kuhn-Tucker (KKT) conditions, optimal quantity and charging price of different functions were derived for the objective functions of platform profit maximization and social welfare maximization, respectively. The parameter sensitivity was analyzed through the numerical experiments. The optimization results show that the appropriate charging price and the number of charging piles determine EV drivers’ willingness to different regions, the number of charging piles and the charging price have optimal solutions in multiple regions and periods. The model results show that, in spatial terms, the price and the quantity decrease with the distance increasing; in temporal terms, the pricing during the peak period is higher than that during the low period, and inter-regional pricing differences during the low period are more significant than those during the peak periods. Compared with different goals, the optimal charging price in social welfare maximization state is lower than that in profit maximization state, and the optimal number of charging piles to achieve maximum social welfare is more than that to achieve maximum profit. Results of sensitivity parameter analysis showed that, the battery capacity, the charging duration for unit battery capacity, and the sensitivity factor of multinomial Logit (MNL) model were negatively correlated with the target results, while positively correlated with users’ perceived utility.