Modeling the relationships between microscopic and macroscopic travel activity on freeways: Bridging the gap between current travel demand models and emerging mobile emission models.

摘要

Metropolitan Planning Agencies currently use Travel Demand Models (TDMs) to forecast traffic and provide the inputs to the approved mobile emission model---MOBILE or EMFAC in California. MOBILE and EMFAC use macroscopic measures of emissions-related vehicular activity as input (average speed and flow). However, these models are insensitive to the underlying distribution of microscopic activity (called modal activity), an important factor in the estimation of mobile emissions. A new generation of mobile emission models are under development that are sensitive to modal activity and require it as input. However, TDMs do not forecast modal activity. This creates a gap between outputs from the TDMs used to forecast traffic and the inputs required by the new generation of mobile emission models. This research bridges this gap.; The methodology validated by this research includes procedures for experiment design, data collection, data processing, and model estimation. Car-following techniques are used to sample vehicles in freeway traffic, collecting speed-time traces (microscopic activity). Data is collected simultaneously from vehicle detectors (macroscopic activity). Both data are mapped temporally and spatially. Data are post-processed using a software management program (MAKE). In final form, modal activity is sampled according to Highway Capacity Manual (HCM) sections. Models are estimated using a regression tree statistical method and cross-validated.; Two measures of modal activity are modeled. These models are statistically valid and support the research hypothesis that freeway microscopic modal activity can be forecast as a function of macroscopic activity and facility geometric properties. The contribution of this research is the validated methodology, not the estimated models, which are too restricted for general use. Recommendations are made regarding the best chase-vehicle instrumentation, traffic detectors, and data management strategy.; An application was also developed. It is a protocol for developing driving cycles based on mutually exclusive and collective exhaustive emission-related activity partitions of the speed-flow regime. This methodology does not rely on the unsupported assumption that HCM Level of Service can provide these partitions. The current updating of MOBILE is based on this assumption. Concluding recommendations are made for experiments to reduce measurement error, determine sample size, and develop methodologies for arterials.