车辆换道的时空影响分析 ：交通安全、效率和油耗视角

[Background] Lane changing, a crucial microlevel behavior in vehicle operation, plays a significant role in traffic flow. Understanding its spatiotemporal impact is essential for effective traffic management optimization. However, existing studies have primarily focused on the instantaneous effects of lane changing, neglecting its sustained influence over time and space, resulting in an incomplete quantification of its impact on traffic flow. [Objective] This study aims to reveal the multidimensional impact of discretionary lane-changing on the safety, efficiency, and fuel consumption of upstream traffic in both the original and target lanes, providing a theoretical foundation for optimizing driver lane-changing decisions and traffic management strategies. [Method] A framework of“data processing-spatiotemporal impact definition-construction of multidimensional evaluation metrics-statistical analysis”is used. By combining metrics such as the time-exposed time-to-collision, fuel consumption, and Edie’s generalized flow, this study quantitatively analyzes the spatiotemporal impact of a single lane change. [Data] Empirical analysis was conducted using the open-source Zen Traffic Data dataset, which includes 16 h of vehicle trajectory data from three typical two-lane highway segments in Japan: Ikeda Route 11, Wangan Route 4, and Higashiosaka Route 13. The vehicles followed the keep-left traffic rule. The right lane is the passing lane and the left lane is the driving lane. [Conclusions] This study shows that lane changes have different impacts depending on the lane type (left vs. right) and target lane. Lane changes from right to left cause a significantly higher safety impact on upstream vehicles in the original lane compared with lane changes from left to right. Lane changes from left to right pose higher safety risks to upstream vehicles in the target lane than lane changes from right to left. In addition, lane changes from right to left demonstrated a greater potential for reducing fuel consumption than lane changes from left to right. Despite the safety risks, lane changes exhibited a flow gain effect, increasing the flow rates by 2.51% and 5.17% in the original and target lanes, respectively. Regarding fuel consumption, most lane-change behaviors contribute to reducing fuel consumption for upstream vehicles in both the original and target lanes. [Application] The proposed method for quantifying the spatiotemporal impacts of lane-changing can support autonomous vehicle lane-changing decision algorithms. By predicting the real-time effects of lane changes on surrounding vehicles, autonomous vehicles can make more rational lane-changing decisions that minimize disruptions while optimizing traffic flow.