Modeling and Mitigation of Car-Truck Interactions on Freeways

Abstract

Trucks represent the predominant mode of domestic freight movement. Because of the substantial increase in freight truck traffic on the nation’s highways, its influence on traffic flow performance, safety, and the quality of the travel experience is receiving increased attention. The behavior of non–truck drivers is modeled in terms of their interactions with trucks in the traffic stream; existing microscopic freeway traffic flow modeling logic is extended to incorporate these interactions; and alternative strategies to mitigate them are evaluated. The car–truck interactions are modeled by associating a “discomfort level” for every non–truck driver in the vicinity of trucks. This discomfort is affected by driver socioeconomic characteristics and situational factors such as time of day, weather, and ambient traffic congestion levels. Stated-preference surveys of non–truck drivers were used to elicit the factors that influence their behavior when they interact with trucks on highways. A fuzzy logic–based model was used to determine non–truck driver discomfort level. The model characterizes non–truck driver behavior near trucks by using if-then rules constructed with the survey data. The discomfort level is used in conjunction with the car-following and lane-changing logics of a traditional traffic flow model to generate a truck-following model and a corresponding lane-changing model. This redresses a key methodological gap in the literature and provides a capability to analyze alternative strategies to mitigate car–truck interactions. An agent-based freeway segment traffic flow simulator was constructed by using these extended microscopic flow models. Simulation experiments with data from the Borman Expressway (I-80 and I-94) in northwest Indiana were used to analyze the sensitivity of the model to the various parameters and evaluate the effectiveness of alternative mitigation strategies.