Research On Shared Parking Space Allocation Considering Stochastic Arrival And Departure Times Of Parking

With the continuous development of urban economy and the continuous improvement of residents’ living standards,the number of small cars in cities continues to increase,and the problem of parking difficulties for residents’ travel gradually becomes prominent.Increasing the number of parking facilities is no longer sufficient to alleviate the problem of parking difficulties.The low utilization rate of existing parking resources in the city is the root cause of the parking problem.Shared parking can effectively alleviate such problems and improve the utilization rate of existing parking resources.However,existing research on shared parking space allocation lacks consideration of the impact of users’ arrival and departure times on parking requests,which reduces user satisfaction with the parking management platform and the stability of the platform’s parking space allocation.Therefore,this study focuses on shared parking space allocation under a random environment of user arrival and departure times,and the main work is as follows.Firstly,this article presents an analysis of the parking demand characteristics for commercial,hospital,and residential facility land,based on a summary of current research on parking and parking spaces.Secondly,the feasibility of implementing shared parking in cities is evaluated from four different perspectives,including sharing theory,application technology,sharing policy,and social user sharing willingness.Finally,the allocation mode of shared parking spaces is examined by analyzing the influencing factors of parking space sharing from three main perspectives: the shared parking space management platform,shared parking demand users,and shared parking space supply users.Secondly,the necessity of parking demand prediction in the process of shared parking space allocation was analyzed.Effective prediction of parking demand has practical significance for the rational sharing of parking spaces.Based on the characteristics and prediction principles of grey prediction models(GM(1,1),GM(1,N))and third-order exponential smoothing prediction models(TES),a joint design was made between TES and GM(1,1)/(GM(1,N))grey prediction models.Based on statistical data on parking demand in the research area,the GM(1,1)-TES-GM(1,N)joint prediction model was applied to predict parking demand,and it was found that the growth of small car ownership from 2016 to 2025 was relatively stable.The predicted parking demand for 8:00-18:00 in the research area was 2160 parking spaces per day,indicating that the results of the parking demand prediction model designed in this chapter were closer to actual parking demand and the model predictions were more reasonable.Finally,in order to improve the satisfaction of parking request users with the parking management platform,an optimization model for shared parking space allocation was proposed with the goal of maximizing system efficiency.The parking allocation model considered the characteristic of the imprecise arrival and departure times of parking request users,the potential loss cost incurred by the parking management platform rejecting request users,and the walking distance of the parking request users.A binary particle swarm optimization algorithm was designed for solving the shared parking space allocation model,which had a large solution space and 0-1type characteristics.The numerical example validated that the model had an acceptance rate of up to 70%.Finally,the parking demand prediction model based on the joint prediction method constructed in Chapter 3 was used to study the Wuyue Plaza complex in Jintan District,Changzhou City.Through simulation to generate 2160 parking requests,the target function value of the model system efficiency was 6207.326 yuan.The impact of imprecise arrival and departure times of parking request users on shared parking space allocation was also analyzed when the parking request size was increased from 1000 to 2000.Based on the research on the allocation of shared parking spaces in this article,it can be concluded that the cost of occupying buffer time for parking request users when arriving and leaving the parking space is related to the probability of users’ parking time being inaccurate and the number of users who make reservations for parking requests upon arrival and departure,considering the perspective of parking management platform system’s revenue.This provides meaningful reference for sensible rental of parking spaces by the parking management platform at different randomly sized parking times.Therefore,considering the random factors affecting user arrival and departure times,provides a reference for optimization of the parking spot matching for shared parking managers,which is more practical.