数据驱动的通勤团体配对共享停车方法研究 postprint

Abstract: With the increase in the number of motor vehicles, parking has become a common problem in cities. When many people take a long time to line up for a parking space at the workplaces, there are many parking spaces are available in the residential area that around the workplaces, which cause the use of parking space ineffectively. However, the existing shared parking methods are difficult to implement because of its randomness. In order to reduce its randomness, reduce the difficulty of implementing shared parking, fianl to reduce the waste of parking space resources, and because of the complementarity of travel times between commuter groups in residential areas and office buildings where around the residential areas, we present a data-driven one-to-one pairing shared solution . To solve this problem, we analyze the entry and exit record data of vehicles , then obtain the characteristics of the idle duration of the parking space in the residential area and the duration of use of the office building , finally get the results of matching parking spaces and vehicles according to the matching method of maximizing the duration. In the experiment for selected area, the proportion of fully matched parking spaces was 37.66%, and the average utilization rate of all matching parking spaces increased by 15.24%, and the maximum increase was 57.84%. The result shows that paired shared parking is extremely feasible.