The novelties of this work are twofold. First, we proposed an agent-based battery electric vehicle charging demand simulation model integrating travel and charging behaviors, which was able to estimate the high-resolution spatiotemporal distribution of charging demand. Second, we constructed a novel charging behavior model for charging mode choice, which was able to capture non-linear charges in random utility, and the impact on charging choice of various factors, namely risk sensitivity, range buffer, and preference for charging rate. It focused on the modeling and forecasting of renewable energy consumption in the transportation sector, which could be directly applied in the optimal design of energy supply systems and the modeling framework allowed it to be generally adopted for broad application.