Chain Activity-based Travel Mode Choice Behavior Model

With the development of the city, pollution, traffic congestion and other problems become more and more prominent. From the city traffic demand management, according to the disaggregate thought, through the analysis of the city residents travel behavior, to establish the efficient travel demand forecasting model, and make the supply and demand balanced becomes more and more important for easing traffic congestioa At present although the travel mode choice model is established from the activities from the perspective of disaggregate model, but the study on chaining the travel mode, and the choice of travel mode chaining is less. And as the selected variables, usually chooses the traditional sense of the personal attributes, family attributes and travel attributes, without from the true sense of the activity chain angle to select variable. This article is based on above considerations, from the angle of the travel mode chain, take the conversion and focation OD paired of the travel mode chain into travel attribute, as the common impacted variables, to establish a typical activity patterns and travel mode chain choice model and provide some reference for future related research and application.Based on the related research and the analysis of basis theory, combined with the2011Kunming resident trip survey data, worked in the following aspects:first, from the angel of activities chain, according to the day trip records of the residents of Kunming, divided the activities chain into three types that is commuting activity chain, non-commuting activity chain and school type activity chain, and carried out statistical characteristics analysis of the three types of activities; secondly, on the basis of selecting the four typical patterns HWH. HWHWH, HWH+, HWH+H in highest proportion of commuting activity chain, strung the travel mode chain of each activities mode according to the activity chain principle, and make travel chain transformation as one of the travel attribute, from personal attributes, attributes of family and travel attribute this three respects to analyze the influence factors of commuting activity patterns as well as the7types of trip chain; finally, according to the above analysis of the influence factors, selected the attribute which is greater impact on the activity patterns and the travel mode chain as the variables, a typical activity pattern and travel mode chain choice multinomial Logit model (MNL) is established, and demarcated and tested by using Biogeme software, along with establishing utility function and the interpretation of model.In the activity chain analysis, conclusions are obtained as follows:activity chain type of Kunming residents mainly displayed as commuting activity chain and simple activity chain; four main typical patterns of activity mode in the commuting activity chain, the HWH mode plays the highest proportion; in the7types travel mode chain, electric vehicle mode chain proportion is the highest. Through the construction of the activity patterns and travel mode chain choice model, results obtained as follows:males, young people whose occupation has the travel demand on time and travel mode chain is non-conversion tends to choose the work trip activity patterns; while female, occupation needn't the travel demand on time and travel mode chain is conversion tends to choose other activities trip activity patterns; In HWH mode, male, middle or young people whose trip distance is relatively distant is more inclined to select the motor vehicle chains; while female, people whose family owns non-motor vehicles and income is in the lower level, travel distances is shorter, as well as place of residence and workplace are in the same division within the area are more likely to choose the non-motor vehicle chain; and female, people who has the travel demand on time. and place of residence and workplace are both in the central area tends to choose public transportation chain.