Behaviour‐Driven Optimization of Multi‐Type Fast‐Charging Configuration in Highway Service Areas Under Bounded Rationality

This study examined the optimal configuration of direct current fast‐charging (DCFC) and high‐power charging (HPC) piles in highway service areas. We first conducted a stated‐preference survey (348 valid responses, face‐to‐face and online). Multinomial logit and mixed logit models were fitted to quantify the heterogeneous charging‐mode preferences of electric vehicle (EV) users. A threshold–greedy decision rule was introduced to reflect bounded rationality. Based on the predicted demand, we formulated an operator profit‐maximization model subject to power capacity and pile number constraints, which incorporates queuing and unmet demand penalties. Sensitivity analyses of user volume, total power capacity, waiting penalties and bounded rationality threshold revealed how optimal DCFC/HPC ratios and operator profits varied across operating scenarios. The proposed framework provides a quantitative basis for deploying EV charging infrastructure in highway service areas.