Research On Choice Model Of Car Commuter’s Departure Time

As the travel of commuter is the most basic and important behavior in individual life, its travel ratio has up to40%to60%.Thus, researching on commuter traffic characteristics and activity patterns, and strengthening the commuter traffic management has become one of the key tasks that effectively solve the city’s traffic congestion. Commuter departure time indicators, which can effectively describe the commuting time distribution and directly reflects the commuter traffic activity patterns, needs further study.Supported by Beijing major social philosophical projects of integrated transport system research under sustainable development, based on the intention survey of the car commuter basic properties and morning rush hour traffic volume, through analyzing the characteristics of the car commuter trip and the influence factors of car commuter departure time choice, this paper built the time-dependent covariate Cox proportional hazards model for car commuter departure time, analyzed the influence of car commuter departure time choice, and predicted a part of car commuter departure time through the model, which provided a reference basis for formulating and implementing policies to relieve urban traffic congestion. In this paper, the main researches are as follows:1、The paper analyzed the characteristics of commuting trips and the factors of commuter departure time;2、The paper Investigated the commuters in Tongzhou residential area, and g et a lot of first-hand data for the latter study. Then the paper screened、 sorted and input the data and analyzed the effective data.3、 The paper analyzed the applicability and reliability of the model, and Weeded out the outliers and influential points, then proved the Cox regression model is applicable to analysis of departure time choice;4、 The paper constructed car commuter model whose main factors included commuter gender, age, trip distance, work time, road traffic volume, etc. and realized the dynamic analysis initial dynamic analysis.5、 According to the commuter family, work and other properties, the paper predicted a part of car commuter departure time by the use of Cox regression model.The results showed a good predictive model that can accurately predict commuter departure time. The paper had the following innovations:(1) The paper proved the applicability of the model from theoretical and empirical aspects respectively and analyzed the reliability of the model.(2) The paper increased an independent variable of real-time road traffic volume into the time-dependent covariates Cox regression model, made the model more effectively in analyzing and predicting car commuter departure time.