Connectivity and automation have accelerated the development of algorithms that use route and real-time traffic information for improving energy efficiency. The evaluation of such benefits, however, requires establishing a reliable baseline that is representative of a real-world driving environment. In this context, virtual driver models are generally adopted to predict the vehicle speed based on route data and presence of lead vehicles, in a way that mimics the response of human drivers.
This paper proposes an Enhanced Driver Model (EDM) that forecasts the human response when driving in urban conditions, considering the effects of Signal Phasing and Timing (SPaT) by introducing the concept of Line-of-Sight (LoS). The model was validated against data collected on an instrumented vehicle driven on public roads by different human subjects.
Using this model, a Monte Carlo simulation is conducted to determine the statistical distribution of fuel consumption and travel time on a given route, varying driver behavior (aggressiveness), traffic conditions and SPaT. This allows one to quantify the impact of uncertainties associated to real-world driving in fuel economy estimates.