Multi-Dimensional Risk Classification and Importance Assessment of Motorway Lane-Changing

1. Research Hypotheses and Objectives：The primary objective of this research is to quantify the subjective risk perceptions of drivers regarding lane-changing maneuvers on high-speed segments. The study is built upon the hypothesis that highway lane-changing risk is a multi-dimensional construct influenced by four primary factors: Driver Factors, Vehicle Characteristics, Road Environment, and Sudden Interference Events. It is hypothesized that human-centric variables, such as fatigue and distraction, carry a higher weight in risk contribution compared to static road infrastructure. Furthermore, the study assumes that sudden, unpredictable events—such as emergency braking or road subsidence—represent a higher level of instantaneous risk than routine driving variables.
2. Data Content and Structure：The dataset consists of 130 valid survey responses collected from drivers with varying levels of experience. The data is structured into three distinct modules:Respondent Demographics: This includes categorical data on age, gender, driving years, and average weekly highway driving duration.Vehicle Profiles: Data regarding the primary vehicle types operated by respondents, ranging from small passenger cars to heavy freight trucks.Risk Preference Matrices: This section contains pairwise comparison values based on the Analytic Hierarchy Process (AHP) and Best-Worst Method (BWM). The values range from "1" (equal importance) to "5" (extreme importance), including reciprocal values (1/2 to 1/5) when the second factor is deemed more critical.
3. Data Collection and Methodology：The data was collected on August 6, 2025, under the academic supervision of Chongqing Transportation University. A structured questionnaire was deployed to gather expert and experienced driver intuition regarding safety risks. To ensure data quality, the survey utilized a "consistency-check" logic common in AHP modeling; for instance, respondents were asked to compare specific sub-factors like "Fatigue Driving" vs. "Distraction" and "Vehicle Speed" vs. "Braking Performance" to ensure a logical hierarchy of risk perception.
4. Significant Findings and Indicators：Initial analysis of the 130 responses indicates several critical safety trends. Within the "Sudden Interference" category, "Emergency Braking of the Lead Vehicle" and "Sudden Lane-Changing by Adjacent Vehicles" were identified as top-tier risk factors. In the "Road Environment" category, "Distance between Vehicles" (Headway) was consistently rated as more critical than "Lane Width" or "Traffic Volume". Additionally, the data highlights a significant perceived risk disparity between "Brake Failure" and "Navigation Errors" within vehicle system malfunctions, with the former reaching extreme importance ratings.

1. 研究假设与研究目标：本研究的核心目标是量化驾驶员对高速路段变道操作的主观风险感知。本研究基于如下假设：高速公路变道风险为多维度构念，受四大核心因素影响，分别为驾驶员因素、车辆特征、道路环境与突发干扰事件。研究进一步假设，诸如疲劳与分心这类以人为核心的变量，相较于静态道路基础设施，在风险贡献中占据更高权重。此外，本研究假定，诸如紧急制动或道路塌陷这类突发且不可预见的事件，相较于常规驾驶变量，会带来更高水平的瞬时风险。
2. 数据内容与结构：本数据集包含130份有效问卷回复，受访者为不同驾龄的驾驶员。数据分为三个独立模块：
（1）受访者人口统计学信息：涵盖年龄、性别、驾龄、平均每周高速驾驶时长等分类数据；
（2）车辆概况：包含受访者主要驾驶车辆的相关数据，覆盖小型乘用车至重型货运卡车的全类别车型；
（3）风险偏好矩阵：本模块包含基于层次分析法（Analytic Hierarchy Process, AHP）与最佳最差法（Best-Worst Method, BWM）得到的两两比较值。取值范围为“1”（同等重要）至“5”（极端重要），当认为第二因素更为关键时，将采用倒数取值（1/2至1/5）。
3. 数据收集与研究方法：本数据采集于2025年8月6日，由重庆交通大学提供学术指导。研究采用结构化问卷，收集专家与资深驾驶员对安全风险的直觉认知。为保障数据质量，本次调研运用了层次分析法建模中通用的“一致性检验”逻辑：例如，要求受访者对“疲劳驾驶”与“分心驾驶”、“车辆行驶速度”与“制动性能”等具体子因素进行比较，以确保风险感知层级的逻辑性。
4. 重要研究发现与指标：对130份回复的初步分析揭示了多项关键安全趋势。在“突发干扰”类别中，“前车紧急制动”与“相邻车辆突然变道”被认定为顶级风险因素。在“道路环境”类别中，“车距（车头时距，Headway）”始终被评定为比车道宽度或交通流量更为关键。此外，数据还凸显了车辆系统故障中“制动失效”与“导航错误”之间显著的感知风险差异，其中前者的评分达到了极端重要性等级。