Find in Library

In this paper, a comprehensive analysis of merging behavior on the freeway under different traffic densities has been conducted. First, the influence of traffic density on merging parameters (namely merging position, merging speed, merging duration, and accepted gap) was analyzed by mathematical statistics, Mann-Whitney-U Test, and distribution fitting. It was found that the merging position, merging speed, merging duration, and accepted gap are all significantly influenced by traffic density and the best-fitted distributions of merging parameters could be different under different traffic densities. Then, to incorporate the dynamic effects of traffic density on merging behavior, a cell-based traffic density was introduced in this paper and logistic regression was used to model the gap selection behavior. The results showed that traffic density has a significant influence on gap selection behavior during the merging process. The effects of traffic density in logistic regression showed that merging drivers tend to choose a location with lower traffic density during the gap selection process. Moreover, introducing the cell-based traffic density can also improve the predictive accuracy of the logistic regression model.