自由流动公共自行车用户停车行为管理：基于客户回报计划的激励机制设计

The lack of parking racks is a main disadvantage of the typical Station based Bike Sharing System (SBBS). The emerging Free Floating Bike Sharing (FFBS) is expected to fix the issue by its character of the dockless installment and the mobile application, which enables the users to rent and return the bikes at any time and any place in the serving tempo-spatial region. The FFBS service is favored by the users for its free-floating style while bring amount of new user misbehaviors like illegal parking. The research is proposed to tackle the new parking related problems in FFBS operation, which could be specified into illegal parking behavior, voluntary illegal parking report behavior and the user based bike relocating behavior. On basis of the user registration data and bike use records, the users would be first segmented to groups by mining their demographical, socio-economical characteristics along with their FFBS usage and parking behavior. Following the framework of the customer reward program, various incentive based parking behavior experimental scenarios would be designed and tested through Stated Preference (SP) and field experiments on representative samples. The behavioral mechanism would be tested and the preference of heterogeneous users on the incentive would be analyzed and introduced in the optimal incentive mechanism design stage. After formulating the gaming behavior between FFBS user and operator under various parking behavior scenarios into mathematical programming problems, the existence of an equilibrium would be analyzed, following by a solving algorithm design. A practical case study would be finally conducted to test the feasibility of the proposed theoretical work. The study would contribute to the theoretical and practical research on the user behavior management in both the FFBS scenario and a more general Sharing Economy context.

新兴的“自由流动公共自行车”不设固定车桩、用户基于APP随借随用、随停随还，直击传统公共自行车高峰期“还车难”的痛点，受到城市中短途出行群体青睐，但其自由流动的特点也引发了违规停放等大量的不当用车行为，给用户、运营商和社会公众造成了显著负面影响。本研究以此为背景，面向用户停车行为管理需求，针对与车辆停放密切相关的“用户违停”、“举报并纠正他人违停”以及“基于用户的车辆调度”行为开展研究。研究首先基于用户特征及其用车与停车行为记录挖掘用户及其停车行为的异质特征，进而借鉴客户回报计划设计框架设计激励方案，通过陈述性偏好和田野实验获得行为实验数据，建立统计学模型解释用户停车决策机制及其激励偏好，在此基础上以社会总福利最大化为目标，建立用户与服务商的博弈模型，研究最优激励机制的设计与优化方法，并进行实例验证。本研究对于自由流动公共自行车及其代表的共享经济模式下的用户行为管理有重要的理论和实践意义。

本研究以自由流动公共自行车（共享单车）“停车秩序混乱”这一现实难题为研究对象，从用户行为特征分析入手，建模分析个人潜在心理特质、人口统计学和社会经济特征以及外部激励措施共同对出行者文明交通行为的影响。研究基于问卷调查获得了出行者的选择行为数据，进而融合结构方程和离散选择建模方法，提出了个体选择行为的混合选择模型，进而比较了内部特征和外部激励措施的影响。本研究发现：（1）将个人心理特质引入自由流动公共自行车的停车选择行为模型中可以显著提高模型的解释能力；（2）对于自愿协助维护停车秩序行为而言，自利倾向会显著抑制个体的行为意愿，而利他意愿则会促进行为意愿；（3）而如果采用货币激励的方式管理用户的停车行为，自利倾向较高的用户会更容易接受激励，而利他倾向较高的用户反而不愿意接受货币激励。同时，本研究还将个体的文明交通行为扩展到更普遍意义的公共交通领域，研究发现：（1）研究了地铁乘客在车厢内的不文明行为，发现社会规范（Social Norms）对于此类行为有显著的影响，但不同类型的行为的影响因素有显著区别；（2）研究了公交乘客在公共汽车车厢内的让座行为，研究发现同情心（Empathic Concern）是影响乘客向弱势群体让座的显著因素；（3）研究比较了不同方向（奖励和惩罚）、类型（社会信用、货币激励）以及激励强度（高和低）的激励措施对于个体让座意愿的影响，研究发现惩罚惩罚类型的措施比奖励措施更有效，货币惩罚比社会信用惩罚更有效，而激励强度越大，乘客的行为反应越明显。本研究中提出的建模方法同时考虑了个人的心理潜变量以及外在激励措施，在不同的个体选择行为建模场景中具有普遍的适用性，且可以显著提高模型对行为的解释能力。同时，本研究发现的个人价值观和同情心等心理特质对于个体文明交通行为的影响也意味着对于此类行为的干预可以通过宣传教育的手段加以实现。