Incorporating Biobehavioral Architecture into Car-Following Models: A Driving Simulator Study

Abstract

Mathematical models of car-following, lane changing, and gap acceptance are mostly descriptive in nature and lack decision making or error tolerance. Including additional driver-related information with respect to behavior and cognitive characteristics would account for these lacking parameters and incorporate a human aspect to these models. Car-following, particularly in relation to the Intelligent Driver Model (IDM), was the primary component of this research. The major objectives of this research were to investigate how psychophysiological constructs can be modeled to replicate car-following behavior, and to correlate subjective measures of behavior with actual car-following behavior. This dissertation presents a thorough literature review into car-following models and existing driving and biobehavioral relationships that can be capitalized to improve the calibration and predicting capabilities of these models. A framework was theorized to utilize the task-capability interface to incorporate biobehavioral parameters such as cognitive workload, situation awareness, and level of activation in order to better predict changes in driving performance. Ninety drivers were recruited to validate the framework by participating in virtual scenarios within a driving simulator environment. The scenarios were created to capture all the necessary parameters by varying the situation complexity of individual tasks. A biobehavioral extension to the IDM was developed to easily calibrate predicted and observed values by grouping individual driver performance and behavioral traits. The model was validated and found to be an effective way of utilizing behavioral and performance variables to efficiently predict car-following behavior.