Strategic Lane changes for Autonomous Buses: A Clustering Method to Approaching Vehicles

Abstract

Globally, automated buses are anticipated to replace conventional buses but face challenges in lane changing due to intersection demands, requiring manual control. This study proposes a lanechanging strategy for fully autonomous operation by clustering approaching vehicles and utilizing spaces between them for smoother transitions. The strategy’s effectiveness was evaluated using metrics such as lane-changing time, conflict rates, and speed variability of following vehicles. Results show a 4.7-second reduction in lane-changing time, a 24.3% decrease in conflict rates, and a 3.17 km/h reduction in speed variability, enhancing operational efficiency, safety, and traffic flow stability. These improvements are expected to increase passenger satisfaction and support traffic continuity, contributing to more stable and efficient automated bus operations.