Characterization of real world bus driving behavior for emission evaluation

This study aims to characterize real world bus driving behavior from data obtained in experimental campaigns using instrumented buses. An integrated statistical method to determine bus driving behavior, based on analysis of time series of bus speed and GPS location data, is illustrated in this paper. Kinematic features of bus operating conditions are characterized by multivariate analysis of trip speed time series. Each trip is analyzed within a multi-level hierarchical structure: sequence, sub-cycle, cycle (part of trip between two successive bus stops). Finally, a preliminary application of the proposed method is shown, based on data obtained with buses of different size and homologation class in Naples and Palermo. Reference driving cycles were built, for one line in both cities, taking the most statistically representative cycles of each line with regard to different traffic and road situations. The ensuing results are useful in evaluating emissions measured in on-road testing at different space and time scales and their association to driving operations.