Traffic simulation with cognitive behavioural mechanisms

Incorporation of more sophisticated human factors (HF) in mathematical models for driving behaviour has become an increasingly popular and important research direction in the last few years. Such models enable us to simulate under which conditions perception errors and risk-taking lead to interactions that result in unsafe traffic conditions and ultimately accidents. In a range of papers from research on this topic, a generic multi-level microscopic traffic modelling and simulation framework has been presented that supports this important line of research. In this framework, the driving task is modelled in a multi-layered fashion. At the highest level, we have idealized (collision-free) models for car following and other driving tasks. These models typically contain HF parameters that exogenously “govern the human factor”, such as reaction time, sensitivities to stimuli, desired speed, etc. At the lowest level, we define HF variables (task demand and capacity, awareness) with which we maintain what the in- formation processing costs are of performing driving tasks as well as non-driving related tasks such as distractions.