A Spatio-temporal Law Mining Method Of Commuting Based On Clustering

Conducting research on the relation between job-housing relationship help deepen the understanding of the law of urban development,based on the activities of the differences between groups of different commuting for analysis,help further understanding of the relation of commuter groups for the post of living space of reaction and the affected degree,operation management and urban planning to urban traffic is of great significance.However,in the construction of existing commuting recognition rules,the experiencebased threshold is unstable,and it is difficult to quickly determine the appropriate indicators such as commuting time period and commuting days,and the classification of commuting groups seldom considers the time continuity characteristics of commuting activities,which makes it difficult to dig out more potential commuting rules.Therefore,this paper focuses on the construction of multi-period commuting recognition rules and the study of commuter crowd classification methods.Taking the subway card swiping data of Shanghai as an example,this paper identifies the subway commuter crowd in Shanghai and explores the commuting temporal and spatial law and the relationship between work and residence space.The main research contents and achievements of this paper are summarized as follows:(1)A multi-period commuting identification and workplace recognition method based on subway swipe card data is designed.The general idea of multi-period commuting recognition rules is given,the spatio-temporal characteristics of commuting activities are analyzed,and the commuting period determination method based on time unit is proposed,the process of extracting one-day morning and evening commuting OD is proposed,the commuting day determination method based on information entropy is proposed,the commuting day number index is automatically determined,and the commuting individual identification rule based on multi-day commuting OD and the workplace and residence identification method are constructed.(2)A method for measuring the similarity of commuting travel time is proposed.The general idea of analyzing the spatio-temporal law mining of commuting groups is analyzed,a method of commuting trip time similarity extraction is defined,a commuting group classification method based on commuting time similarity clustering is proposed,and the commuting space law mining relationship analysis method based on the spatial analysis model is discussed to realize the spatio-temporal law mining of commuting groups.(3)An empirical analysis of the spatio-temporal characteristics of work-residence relationship and commuting in Shanghai is carried out.A case study was conducted using the one-week subway card swipe data in Shanghai.The results show that: 1)The Shanghai employment single center model is obvious,different clusters of employment hotspots are distributed in the city center,and the settlements are characterized by the spatial organization of "hot in the west and cold in the east".2)The main circulation period is7:00-8:30 and 17:00-19:00,nearly half of the commuters commute during the early mainstream period,and 90% carry out off-work activities before 19:30.3)The travel time characteristics of different commuting types are generally consistent with the distribution of their work and housing hotspots.