随着我国机动车保有量的不断增加，停车配套设施与汽车保有量增长不同步而导致的停车资源短缺问题日渐突出。一方面，城市有限的土地资源给停车设施的大规模扩建带来困难，另一方面，现有的停车资源在停车区域和时段等维度上存在不均衡现象。特别是居住区、办公区等停车场所虽然停车设施资源体量大，但由于存在明显的潮汐效应导致部分时段的泊位利用率不高。因此，探索共享停车的模式对解决此类问题具有一定的现实意义。由于泊位的位置及闲置时间较为分散，“先到先服务（FCFS）”这种简单的匹配方式较难满足双方的需求，因此，本文通过构建基于满意度的共享匹配模型来缓解停车资源短缺问题。本文的主要内容包括：

（1）泊位共享是共享经济的一种实践形式，因此本文首先从共享经济的概念、特点、意义及应用等几个方面对共享经济进行分析；并在此基础上介绍了泊位共享的相关概念，从理论、政策以及技术三个方面论证了泊位共享的可行性，分析了泊位共享中用户的偏好，并对泊位共享的实现途径进行探讨。

（2）本文在对不同的停车需求预测方法进行分析的基础上，根据车辆停放特性，选择时间序列模型对停车需求数量进行预测。探讨了时间序列预测模型的构建流程，运用EViews软件对实测的停车占用数据进行建模预测，并验证模型及方法的可行性。最后，在不影响原有停车需求的前提下，充分考虑共享泊位供需双方的时间匹配以及共享泊位管理成本等因素，提出了确定共享时间窗口和共享泊位数量的方法。

（3）在已知停车区域共享泊位供给量的前提下，建立考虑供需双方偏好的共享匹配模型，是提高泊位共享双方满意度和提升停车体验的必要途径。因此本文在停车需求者和泊位提供者双方停车时间窗匹配的基础上，充分考虑泊位提供者的偏好（泊位利用率最大化）和停车需求者的偏好（包括停车后的步行距离、停车的费率、安全性等因素），以最大化泊位共享供需双方的满意度为目标，构建多目标的共享泊位匹配模型，并运用加权和法结合匈牙利法的求解算法对模型进行求解。最后，将求解结果与FCFS算法计算结果进行比较，验证模型的可行性与优越性。

在上述研究的基础上，本文还提出了完善泊位共享模式的建议，希望能够对共享停车模式的发展以及城市停车资源短缺问题的缓解有所帮助。

With the popularity of automobiles in China, shortage of parking resources caused by the unsynchronized growth of parking facilities and car ownership poses a huge challenge to traffic. On the one side, the shortage of urban parking resources makes it hard to expand the parking resources on a large scale. On the other side, existing parking resources are unbalanced in dimensions such as parking area and time period. In particular, parking spaces such as residential areas and office areas have massive parking resources, but due to the apparent tidal effect, the berth utilization rate is not high in some periods. Therefore, it is of practical significance to explore the solution of such problems through berth sharing mode. Because the location and idle time of berths are relatively scattered, it is difficult to meet the needs of both parties with a simple "First Come First Served (FCFS)" matching method. Therefore, we try to build a satisfaction-based supply and demand matching model to solve this problem.

(1) Berth sharing is a practical form of sharing economy. This article conducts research on the sharing economy from the concept, characteristics, significance and application of the sharing economy. On this basis, the concept of berth sharing is introduced, and the feasibility of berth sharing is demonstrated from theoretical, policies and technical aspects. The preferences of the supplier and the demander in the process of berth sharing are analyzed. Besides, the implementation of berth sharing is discussed.

(2) This paper makes comparison of the various forecasting methods of parking demand, combining with vehicle parking characteristics. Hence, the time series model is selected to predict the number of parking needs. This paper explores the construction process of the time series prediction model. Carrying out the measured parking occupancy data on modeling and predicting by using EViews software, and verifying the feasibility of model and method. Finally, under the premise that the original parking demand is not affected, consider the time matching and sharing berth management costs of the berth sharing, and proposes the window of time sharing and the method to ensure the number of berth sharing.

(3) Under the premise that the supply of shared berths in the parking area is known, establishing a supply-demand matching model for shared berths considering the preferences of both parties is a necessary way to improve the satisfaction and parking experience of both berth sharing parties. Therefore, the scenario of sharing parking spaces is represented. On account of the time matching mechanism of the two parties in the shared parking, the preferences of the parking space provider (maximizing the efficiency of parking space) and the preferences of the parking demander are fully considered ( Such as walking distance after parking, parking rate, safety, etc.). This article takes into account the maximum satisfaction of the supplier and the demander of parking sharing simultaneously. A multi-objective optimization matching model is established and the weighted sum method with the Hungarian method is chosen. By comparison, the constructed method of this article is superior to the first-come-first-served(FCFS) algorithm.

On the basis of the above research, this article also puts forward suggestions to improve the parking sharing model, hoping to help the development of the shared parking model and the alleviation of the shortage of urban parking resources.