Bikeshare systems have attracted increased research interest ranging from bikeshare planning analyses to operational improvement studies (e.g., rebalancing, or station optimization). However, the interaction between bikeshare station spatial distribution and actual bikeshare activities when addressing equity issues has not been thoroughly considered. Moreover, there is a paucity of research helping governments develop incentive programs for equitable bikeshare services. To fill this research gap, we develop a model to estimate the potential demand (i.e., bikeshare trip production and attraction) and its distribution, and evaluate performance over a set of objectives (e.g., maximization of annual revenue, accessibility improvements) to find the most equitable distribution of stations. We build a genetic algorithm to solve this multi-objective optimization. The study uses the Divvy bikeshare system in Chicago as a case study, and compares the solutions of the model with the system's expansion (new stations added) in 2016, which targeted disadvantaged areas. When selecting accessibility as the main objective, the results indicate the need to provide more stations in disadvantaged areas and those results overlap with the system's expansion in 2016. On the contrary, the goal of revenue maximization results in a smaller network of stations and fewer accessibility improvements, especially in disadvantaged communities. A sensitivity analysis uncovers the greatest obstacle (i.e., station cost) to adding more stations in disadvantaged areas. More importantly, a Pareto frontier of this multi-objective optimization supports several policy suggestions for incentivizing private bikeshare companies to target more disadvantaged populations. Our results show the importance of considering accessibility and other equity constraints in developing a more inclusive, equitable and sustainable transportation system, and we provide several planning suggestions.

共享单车系统吸引了越来越多的研究兴趣，从共享单车规划分析到运营改进研究（例如，重新平衡或站点优化）。然而，在解决公平问题时，公共自行车站点空间分布与实际公共自行车活动之间的相互作用尚未得到彻底考虑。此外，帮助政府制定公平共享单车服务激励计划的研究很少。为了填补这一研究空白，我们开发了一个模型来估计潜在需求（即共享单车出行的产生和吸引力）及其分布，并评估一系列目标（例如年收入最大化、可达性改进）的绩效，以找到最公平的车站分布。我们构建了一个遗传算法来解决这个多目标优化问题。该研究以芝加哥的 Divvy 共享单车系统作为案例研究，并将该模型的解决方案与该系统在 2016 年针对弱势地区的扩展（新增站点）进行了比较。When selecting accessibility as the main objective, the results indicate the need to provide more stations in disadvantaged areas and those results overlap with the system's expansion in 2016. On the contrary, the goal of revenue maximization results in a smaller network of stations and fewer accessibility改善，尤其是在弱势社区。敏感性分析揭示了在贫困地区增加更多站点的最大障碍（即站点成本）。更重要的是，这种多目标优化的帕累托前沿支持多项政策建议，以激励私人共享单车公司瞄准更多弱势群体。我们的研究结果表明，在开发更具包容性、公平性和可持续性的交通系统时考虑可达性和其他公平性限制的重要性，我们提供了一些规划建议。