1 MODELLING THE HEAVY VEHICLE DRIVERS’ LANE CHANGING DECISION UNDER HEAVY TRAFFIC CONDITIONS Sara Moridpour, Geoff Rose, Majid Sarvi Department of Civil Engineering, Monash University, Melbourne, VIC, Australia ABSTRACT Lane changing manoeuvres of heavy vehicles have significant influence on traffic flow characteristics of the surrounding vehicles. This influence is due to the physical effects that the heavy vehicles impose on surrounding traffic. This paper presents an exclusive fuzzy logic lane changing decision model for heavy vehicle drivers on freeways. The trajectory data which is applied in this study is under heavy traffic conditions. Then, the efficiency of the calibrated lane changing decision model is examined. The validation results show that the new fuzzy logic lane changing decision model could closely replicate the observed lane changing decisions of the heavy vehicle drivers. Keywords: Lane changing decision; Heavy vehicles; Heavy traffic conditions. 1. INTRODUCTION Lane changing manoeuvres of heavy vehicles have significant influence on traffic flow characteristics of the surrounding vehicles due to the physical effects that the heavy vehicles impose on surrounding traffic (Moridpour et al. 2008; 2009). These effects are the result of heavy vehicles’ length, size, weight and the limitations in their manoeuvrability (Uddin and Ardekani, 2002; Al-Kaisy and Hall, 2003; Al-Kaisy et al., 2005). Considering the specific physical and operational characteristics of heavy vehicles, understanding the lane changing behaviour of heavy vehicle drivers is very important. However, the previous lane changing models are mainly associated with passenger car drivers and the lane changing behaviour of heavy vehicles has received little attention. Several approaches have been applied to model the drivers’ lane changing decision. The conventional lane changing decision models are mainly based on mathematical equations and crisp magnitudes to model drivers’ lane changing decision (Das and Bowles 1999; Das et al. 1999). However, in real world the drivers’ make their driving decisions based on their imprecise perceptions of the surrounding traffic. To overcome this problem, several approaches have recently become popular such as fuzzy logic. Fuzzy logic provides the opportunity to introduce a quantifiable degree of uncertainty into the modelling procedure to reflect the natural or subjective perception of the real variables (Wu et al. 2000).