Supply Forecast of Shared Parking Spaces in Social Parking Lots Around Primary Schools Based on Shared Parking Spaces

With the increasing proportion of car transportation in the school-wide mode, the traffic jam of different degrees will be formed at the entrance of every primary school in Shenzhen, this phenomenon not only wastes the time of parents and students, affects personal safety, but also causes great pressure to the traffic operation and management around. At present, there is no good solution to this problem, and there is also a lack of in-depth research. As an effective way to ease the parking problem in the city center, “Parking sharing” has attracted the attention of some scholars at home and abroad in recent years. However, most of the existing studies are biased towards the specific parking lot to do a period of time in the future parking supply forecast, and do not take into account the parking demand of primary schools, in the actual operation process, there is a common situation that the prediction precision of shared parking is not enough and can not fully meet the parking demand. Based on the theory of shared parking spaces, this paper transfers the parking demand of primary school to the surrounding social parking lots, which can alleviate the traffic disorder in front of school and ensure the safety, the utility model can also improve the utilization rate of parking spaces in social parking lots. The supply of parking spaces plays an important role in evaluating the social and economic benefits of parking spaces. The supply law of parking spaces is different in different types of allocated parking lots, the trend of free shared berth is different with time. With the aim of improving the clarity and reliability of the information on free parking spaces in car parks, this paper takes three key primary schools in Luohu District as an example, and obtains the time-varying data of historical parking spaces in various types of car parks around primary schools through the parking space sharing platform, the ARIMA and BP neural network models are used to forecast the time span of parking lots in different land use types, and the applicable prediction methods of shared parking spaces are found. The results show that the smaller the time span is, the higher the prediction accuracy is. BP neural network model is more suitable for office area and part of commercial area, while ARIMA model is more suitable for residential area. This study has great practical significance for mining the parking supply capacity, reducing the construction of parking facilities and investment, reducing the demand for parking space, optimizing the transportation environment in primary schools, and improving social and public benefits.