Research On Classification And Influencing Factors Of Nanjing Metro Commuting Patterns Based On Fusion Data

Commute is an important and rigid transportation demand in the urban.Rail transit has received more and more attention in recent years due to its advantages of large capacity,high efficiency,and environmental protection.In order to study the metro commuters' behavior,this paper combined the conventional resident trip survey data and smart card data to cluster metro commuting patterns and analyze the formation mechanism of different metro commuting patterns in Nanjing.Firstly,the commuting recognition rules were drawn from metro commute behavior in the resident trip survey data,and metro commuters were obtained based on metro IC card data.The metro commuting distance,the time when entering the housing station and job station were used as indicators to reflect the individual's characteristic of metro commuting behavior.Based on the Gaussian mixture model,these three indicators were clustered to obtain three typical metro commuting patterns: classic commuting pattern,staggered commuting pattern and long distance commuting pattern.Secondly,the metro commuting distance,the time when entering the residence station and job station were calculated in the resident trip survey database.Based on these indicators,the metro commuters were classified into the three metro commuting patterns.Then the individual and family characteristics,residence and work location characteristics,commuting trip characteristics were analyzed under different metro commuting patterns.Furthermore,the three characteristics which were directly obtained or calculated from the resident trip survey data were used as variables to establish a multinomial Logit model and a mixed Logit model for analyzing the formation mechanism of metro commuting patterns.The research results indicate that the goodness of fit of the mixed Logit model is better,the commuting trip characteristics affect the metro commuting patterns most,the individual social economic attributes include the most significant influencing factors,and the impact of the built environment in the office and residence is weak.Finally,based on the proposed model,the formation mechanism of metro commuting patterns was analyzed from three aspects: individual and family characteristics,built environment characteristics of office and residence and commuting trip characteristics.Based on the analysis,policy suggestions were proposed in order to improve the metro commuting efficiency and service level.