Fuzzy logic and enhanced MOBIL model for a type lane-changing decisions in mixed traffic

The study of lane-changing behaviour is crucial for road efficiency, safety, and traffic flow, especially at highway interchanges where lane-changing actions significantly impact capacity and safety. In mixed traffic with human-driven and autonomous vehicles (AVs), the traditional model of Minimising Overall Braking Induced by Lane changes (MOBIL) fails to account for human decision making imprecision. This study is an analysis of lane-changing characteristics using naturalistic driving data and enhances MOBIL by integrating dynamic fuzzy thresholding and collaborative decision making. A fuzzy logic-based lane-changing model for human drivers is developed, incorporating multivehicle data and lane-changing urgency, and has been adapted for AVs. Using Simulation of Urban Mobility (SUMO)–Python microsimulation, results show that a 50% AV penetration rate improves interchange efficiency by 15% and reduces conflicts by 20%, compared with human-only scenarios. These findings aid in optimising traffic management for mixed-traffic interchanges.