| The rapid increase of parking demand leads to the lack of parking facilities and unbalance distribution of parking resources,which put great pressure on the city parking management.To solve parking problems,both the development of hard wares and the improvement of management technologies are required.To be specific,a close research of daytime parking or commuting parking,when parking conflict reaches the peak,is quite important.Based the differences of commuting and non-commuting parking demand,the parking resource of single or more facilities,even the zone,can be better utilized,via the optimal distribution of parking demand.Targeting at optimizing the parking supply and improving the utilization of parking resources,based on the demand and supply characteristics of multiple land-use parking facilities,and parkers’ parking choice behavior,this paper analyses the feasibility of shared parking in combined parking facilities,especially adopting residential parking resources to meet commuting parking demand.Based on the individual researches,this paper conducted zoning researches.With spatial analysis technology,parking small zone partition method,along with judgement methods of relationship between parking supply and demand,are proposed.Aimed at balanced utilizing of parking resources and minimizing the number of needed parking spots,parking facility location choice models are built under different shared parking strategies.The details of work and the conclusions are illustrated as follows.For the study of parking demand characteristics of combined parking facilities,the gate machine data of five typical working places are collected.Besides,the commuting and non-commuting groups are identified and clustered,and the parking occupancy prediction models are built by SARIMA model.The results show commuting group can be identified by weekly parking frequency,average parking duration,and the percentage of daytime parking duration in total parking.Multilinear models can be adopted in commuting and non-commuting parking occupancy prediction.For the study of openness willingness of residential parking spots,one-year of owners’ behavior records in the application database are collected and a field survey is conducted.Two CART models are built to analyze the influential factors in owners’ two stage decisions,whether to open or how long to open,and owners’ opennesswillingness are predicted.The results show owners’ self-use behavior,spots’ physical characteristics and the last month’s rental effects all pose influence on owners’ openness willingness.The self-use parking modes deeply impacts owners whether to share decision.Once opened,the willingness of sharing duration is quite influenced by last month’s rental effects.Besides,the consideration of parking spots’ physical characteristics varies in two stages.For the study of parking choice behavior,A questionnaire survey is conducted and a PLS-SEM model is established to analyze the influential factors of respondents’ participation decisions,which concerns more about the parkers’ socioeconomic characteristics and their attitudes towards shared parking lots.Then,respondents are classified into six groups based on their participation willingness by K-Means clustering method and strategies to improve their willingness are proposed.Besides,app data are used and another PLE-SEM model is built to analyze the influential factors of parking spot choice decisions,which concern more about the physical and temporal characteristics of parking spots.The results show both parkers’ socioeconomic factors and their attitudes factors influence parkers’ willingness of adopting residential parking spots for commute parking.As for parking spot choices,the influential factors include both the physical and temporal characteristics of parking spots,the latter includes owners’ sharing willingness and habits.For the study of small-parking-zone-based parking supply and demand relationship,Voronio polygons are adopted to partition the space into small zones,and five calculation methods of parking supply and demand relationship are proposed based on Voronio polygons and Delauney triangles,covering the total quantity,the composition and the spatial distribution of parking resources.Besides,real parking resource data in study area is collected.Based on whether including residential parking facilities,two groups of small parking zones are generated and the relationship between supply and demand are calculated accordingly.The results show Voronio polygons can be used in small parking zone partition and the proposed calculation methods fit the research.From the case study,it is revealed that residential parking resource sharing can effectively improve the parking resource shortage situation and precisely location the area of unbalanced parking supply and demand.For the study of parking facility location choice,the study is conducted by targeting at minimizing the maximum parking resource shortage of small parking zones.In other words,build the medium parking zones from small parking zones byidentifying the central and hinterland small zones.Meanwhile,calculate the parking resource storage under different cluster strategies and select the best strategy based on the target function to generate the parking facilities location choice scheme.To be specific,parking flow concept is proposed and the small-parking-zone-based parking resource shortage calculation method is put forward,as well as the transferred parking demand calculation method.Based on the balanced utilization of parking resource and minimizing the needed parking facilities concepts,target function is built,and the constrains of spatial connection are proposed,as well as the heuristic algorithm AZP-CI.Besides,parking resource storage location and needed parking spots number are calculated under three different shared parking strategies.The results show the proposed method is feasible and adaptive.The case study proves the effectiveness of sharing residential parking resource during daytime in workdays along with the priority of allocating commute parking demand,which are helpful to improve the utilization of parking resources and ease the conflict of parking supply and demand relationship.The result also shows the small and dispersed residential parking supply can influence the parking condition in much bigger zone by transferring parking demand. |