This study investigates the electric vehicle (EV) traffic equilibrium and optimal deployment of charging locations subject to range limitation. The problem is similar to a network design problem with traffic equilibrium, which is characterized by a bilevel model structure. The upper level objective is to optimally locate charging stations such that the total generalized cost of all users is minimized, where the user’s generalized cost includes two parts, travel time and energy consumption. The total generalized cost is a measure of the total societal cost. The lower level model seeks traffic equilibrium, in which travelers minimize their individual generalized cost. All the utilized paths have identical generalized cost while satisfying the range limitation constraint. In particular, we use origin-based flows to maintain the range limitation constraint at the path level without path enumeration. To obtain the global solution, the optimality condition of the lower level model is added to the upper level problem resulting in a single level model. The nonlinear travel time function is approximated by piecewise linear functions, enabling the problem to be formulated as a mixed integer linear program. We use a modest-sized network to analyze the model and illustrate that it can determine the optimal charging station locations in a planning context while factoring the EV users’ individual path choice behaviours.

本研究调查了电动汽车 (EV) 交通平衡和受范围限制的充电地点的最佳部署。该问题类似于具有流量均衡的网络设计问题，其特点是具有双层模型结构。上层目标是优化定位充电站，使得所有用户的总广义成本最小，其中用户的广义成本包括两部分，旅行时间和能源消耗。总广义成本是对总社会成本的衡量。较低级别的模型寻求交通平衡，其中旅行者最小化他们的个人广义成本。所有使用的路径在满足范围限制约束的同时具有相同的广义成本。特别是，我们使用基于原点的流来维护路径级别的范围限制约束，而无需路径枚举。为了获得全局解，将下层模型的最优性条件添加到上层问题中，从而形成单层模型。非线性旅行时间函数由分段线性函数逼近，使问题能够被表述为混合整数线性程序。我们使用一个中等规模的网络来分析模型，并说明它可以在规划环境中确定最佳充电站位置，同时考虑电动汽车用户的个人路径选择行为。非线性旅行时间函数由分段线性函数近似，使问题能够被表述为混合整数线性程序。我们使用一个中等规模的网络来分析模型，并说明它可以在规划环境中确定最佳充电站位置，同时考虑电动汽车用户的个人路径选择行为。非线性旅行时间函数由分段线性函数近似，使问题能够被表述为混合整数线性程序。我们使用一个中等规模的网络来分析模型，并说明它可以在规划环境中确定最佳充电站位置，同时考虑电动汽车用户的个人路径选择行为。