Driving behavior inertia in urban tunnel diverging areas: New findings based on task-switching perspective

Urban tunnel diverging areas are crucial for enhancing overall traffic efficiency. However, the complexity of driving tasks and drivers’ inaccurate perception and response to these tasks are primary contributors to accidents. This study categorizes tunnel diverging areas into three task segments: the approach segment, ramp discovery segment, and entry and navigation segment, with task switching occurring during segment transitions. The objective is to investigate drivers’ responses to driving tasks and their behavior during task switching, providing a foundation for optimizing traffic engineering in diverging areas. Data on speed, longitudinal acceleration, vehicle position, and steering wheel angle were collected from 44 drivers in field tests. Initially, a comparative analysis of driving behavior within task segments was conducted. ANOVA was then used to identify critical change points for each indicator during task switching. Finally, K-means clustering was employed to analyze multiple driving behavior indicators and explore response delays during task transitions. The results reveal that drivers exhibited a higher speeding ratio and delayed lane-change responses within task segments. During task switching, drivers tended to continue their previous driving state—termed ‘driving behavior inertia’, which led to later entries into deceleration lanes and more abrupt deceleration, highlighting drivers’ reduced task sensitivity and lower risk perception in tunnel diverging areas. Furthermore, driving behavior inertia was significantly influenced by route and gender. Drivers in the left lane showed weaker inertia than those in the right lane, while female drivers exhibited stronger inertia. These findings offer valuable insights for the design and optimization of traffic engineering facilities in tunnel diverging areas.