Data Processing Framework for Development of Driving Cycles with Data from SHRP 2 Naturalistic Driving Study

Abstract

In the modeling of vehicle operation costs, a driving cycle is a representative speed–time profile to describe the speed–acceleration pattern of a specific road scenario. Driving cycles are important input for estimation of fuel consumption and polluting emissions. Existing driving cycles are either from out-of-date driving data or without detailed consideration of influencing road properties because of the limitations of available data sets. As part of a project sponsored by FHWA, this research developed a data processing framework for development of driving cycles with data from both the SHRP 2 Naturalistic Driving Study (NDS) and the SHRP 2 Roadway Information Database (RID). The framework included data processing of NDS and RID data and a new synthetic optimization method to generate optimized representative driving cycles. The documented data processing framework was applied to develop the driving cycles of light-duty vehicles for 395 road scenarios with consideration of 10 road properties that could have influenced traffic speed patterns. The 4,400 NDS trips, each of which was at least 20 min long, were used for the development of driving cycles. This data processing frame can be applied for development of driving cycles for more road scenarios with data similar to those in the SHRP 2 database.