Heterogeneity Analysis Of Passenger Perception And Differential Service Strategies Aimed At Intermodal Service

As a vital part of the integrated transportation system,intermodal service can connect the facilities and service of different modes of transportation,and take full advantage of different modes of transportation,which is of great significance for promoting the supply-side structural reform of transportation and the integrated development of transportation.The study is sponsored by the project of National Key Research and Development Plan(2018YFB1601304),National Natural Science Fund(52072066)and Outstanding Young Scholar Fund(BK20200014).Aimed at intermodal passengers,the heterogeneity in passenger perception of intermodal service is analyzed by using the heterogeneity analysis methods,and differential service strategies are formed.Firstly,structural equation modeling and artificial neural network are combined to explore the influence of single attribute on intermodal passenger perception.The model is established for subgroups to identify the determinants of passenger perception and the importance of factors.The heterogeneity in passenger perception among different subgroups is analyzed.At the same time,structural equation modeling-multiple indicator multiple cause and artificial neural network are combined to explore the influence of multiple attributes on intermodal passenger perception,and identify the influence of key attributes on the latent variables.Decision tree algorithm is used to classify passengers based on the data of key attributes.Combination weighting method is used to calculate the comprehensive evaluation function of each group to measure the overall perception level.The difference in the perception level of each dimension among each group is compared with each other.Secondly,aimed at air-bus intermodal service,the SEM-ANN method is used to explore the influence of single attribute on perception based on the data of passenger perception in Nanjing Lukou Airport,and the priority factors for passengers under the influence of attributes are identified.The SEM-MIMIC-ANN method is used to explore the influence of multiple attributes on perception,and the weight of each attribute on the perception level in each dimension is obtained.Passengers are divided into six groups based on five key attributes with decision tree algorithm.The comprehensive evaluation function value is calculated and the perception difference is compared.Considering the actual conditions of air-bus intermodal service,the differential service strategies are proposed.Finally,aimed at air-rail intermodal service,the SEM-ANN method is used to explore the influence of single attribute on perception based on the data of passenger perception in Shijiazhuang Zhengding Airport,and the priority factors for passengers under the influence of attributes are identified.The SEM-MIMIC-ANN method is used to explore the influence of multiple attributes on perception,and the weight of each attribute on the perception level in each dimension is obtained.Passengers are divided into six groups based on six key attributes with decision tree algorithm.The comprehensive evaluation function value is calculated and the perception difference is compared.Considering the actual conditions of air-rail intermodal service,the differential service strategies are proposed.The results show that the heterogeneity analysis methods are feasible and effective,which can better explain the mechanism of passenger perception heterogeneity,and the differential service strategies have reference significance for improving the quality of intermodal service.