Research on Micro Traffic Risk at Entrances and Exits of Highway Interchanges Based on a Multi-Dimensional Perspective Method

Interchanges, as crucial traffic nodes in road networks, play a pivotal role in facilitating smooth traffic transitions. However, the frequent lane changes and inflow/outflow of vehicles at interchange entrances and exits often result in an increased risk of traffic accidents. In this study, we constructed a 3D virtual model of a high-density interchange cluster and conducted a multi-person driving simulation experiment to gather analytical data. To quantitatively assess the micro-risks associated with interchange entrances and exits from multiple perspectives, namely “2D God’s view, 2.5D driver’s view, and 3D spatial view,” we employed a multidimensional approach. The 2D risk was evaluated using metrics such as Time-to-Collision (TTC) and Post-Encroachment Time (PET). Driver tension was employed to quantify the 2.5D risk, while driving comfort and stability were used to measure the 3D risk. Weightings for risk factors from each perspective were determined using the Borda ordinal value method and analytic hierarchy process (AHP). Ultimately, a risk tree map was generated to depict the risk levels at interchange entrances and exits. This map serves as a diagnostic tool to identify high-risk scene factors, comprehensively evaluate traffic safety and driving comfort, and provide guidance for the implementation of control measures.