Early-onset Breakdown of Bottleneck Traffic at Expressway Merging Area, Part I: Causes Analysis

Expressway merging areas are typical recurring bottlenecks, and traffic breakdown seriously restricts operating efficiency and affects traffic safety. Previous research has proved that many merging bottlenecks show early-onset breakdown, which further aggravates bottleneck effects. However, it is difficult to reproduce and analyze this unique phenomenon using current modeling and simulation methods. In the first part of this study, three new types of merging behavior were defined and modeled. They were integrated with car-following and lane-changing models to build a high-precision simulation prototype for bottleneck traffic flow. In the second part, the focus was on the causes of the early-onset breakdown of bottleneck traffic flow - that is, analyzing the factors influencing bottleneck breakdown based on controlled simulation experiments and investigating the internal mechanism between different micro driving behavior factors and macro traffic flow breakdowns. First, a simulation prototype for merging bottlenecks was calibrated with refined parameters as the basis for controlled experiments. Second, an orthogonal test was designed based on the multifactor analysis-of-variance method to eliminate the interaction between factors and analyze the main effects of various driving behaviors on bottleneck efficiency. Finally, causal inference theory was used to build a cause diagram of early-onset breakdown and to analyze the interaction chain for the above driving behaviors, as well as the interaction relationship with bottleneck breakdown. The results show that the most important causes of early-onset breakdowns are consistent with the influencing factors, in which parallel merging and mainline forced lane changing negatively affect traffic efficiency, whereas active-response merging has a positive effect. The causal diagram shows that the driving behavior event chain leading to bottleneck breakdown is active-response merging-"-parallel merging -"-mainline forced lane changing-"-productivity, and the driving characteristic parameters (time headway and merging motivation generation distance), as exogenous variables, affect the frequency of the above behaviors. © 2024 Chang'an University. All rights reserved.