为解决共享汽车的选址难题，提出一个新的停车网点选址方案，并以武汉市洪山区为例进 行验证。首先，利用供需平衡法、出行方式比例法和万人拥有率法对2020 年武汉市共享汽车总量 进行预测，据此估算洪山区的网点需求量；然后，从个人、企业和政府三个角度考虑选址影响因 素，确定相关假设及约束条件，并以造价总成本最低、出行总距离最短、网点覆盖率最广、满足 需求量最大、政府满意度最高为目标函数，建立网点选址的数学模型；最后，依据官方数据确定 模型各参数，采用非支配排序的遗传算法求解选址模型，得出10 组可行解。以各目标函数作为评 价指标，在给定的权重下，从中筛选出最优解作为唯一的选址结果。研究结果表明，该方法对共 享汽车网点选址可提供有效的决策支持。

In order to solve the problem of the parking site selection of car-sharing, a new scheme was proposed，which was applied in Hongshan district of Wuhan city. Firstly, the total amount of car-sharing of Wuhan city in 2020 was predicted by using the balance of supply and demand method, the travel proportion method and the population ownership method separately. Based on the estimated result, the parking sites number demanded was evaluated. Secondly, taking the perspectives of individuals, enterprises and government into consideration, some related assumptions and constraints were determined. A model was set up with the object function which is based on the lowest cost, the shortest travel distance, the biggest overall coverage of the network, the highest demand and the highest degree of government satisfaction. Finally, parameters of the model were determined according to the official data, and NSGAⅡ (Nondominated Sorting Genetic Algorithms Ⅱ) was utilized to solve the location model, resulting in ten groups of feasible solutions. Taking each objective function as the evaluation index and considering the given weight, the optimal solution was selected as the unique site solution. The results show that the proposed model can provide effective decision-making support for the parking site selection of car-sharing.