Stratified Measurement Model Of Service Levels For Public Transportation Based On Multi-Source Data

With the increasingly important role that the public transportation plays in the urban transportation system, the public transportation services have received the increased attention from both transit operational departments and the general public. Most of the existing studies on the public transportation services have been conducted at the macro level, the method in establishing the evaluation index system has been simple, and most of the selected indicators have been difficult to be quantified, resulting in a lack of practicality in the studies. Therefore, this thesis focuses on the quantitative study on the evaluation method of the public transportation service level based on the multi-source bus operational data. It evaluates the service levels of the public transportation from three levels, the bus stop, the bus line section, and the whole bus line, which are stratified, progressive, inter-correlated, and jointly reflect the level of bus services.First, at the stop level of bus lines, five factors are selected as the explanatory variables, including dwelling times at stops of bus line, volume of boarding passengers, volume of alighting passengers, number of bus lines serviced, and difference between boarding and alighting passengers, the feeling of passengers about the level of service at the stop is selected as the response variable. Then, the service evaluation model at the stop level is established based on the cumulative logistic regression.Then, at the segment level of bus lines, the service levels both at the stop and in the road section are taken into the consideration. The service level at the stop is calculated by the proposed evaluation model based on the cumulative logistic regression. For the service level in the section, the thesis selects the bus occupancy rate and the average vehicle speed in the section as the independent variables to build the calculation model using the expert evaluation survey. Thereby, the weighted service evaluation model in the segment level is established using the time as the weight, based on the service level at the stop and the service level in the section.Afterward, at the level of the whole bus line, by using the comprehensive evaluation method, the study chooses the service levels of all segments as independent variables and the Passenger-Hours Traveled (PHT) index as the weights to establish the bus line service level evaluation model based on the demand of the transit passengers.Finally, some regular bus lines in Beijing are selected in a case study with the established models. The service levels in the morning peak hours, non-peak hours, and evening peak hours are calculated by the corresponding models established in the thesis. The results are then comparatively analyzed to validate the practical application of the models.