Abstract
Traffic congestion on today's freeways is a serious problem, causing significant delays for both passengers and goods. Freeway traffic congestion also results in increased vehicle emissions; however, this increase has not been quantified using current vehicle emission models. Current models use emission factors based on driving cycles that do not properly represent freeway driving characteristics. This paper presents a new methodology for relating the macroscopic speed, flow, and density parameters measured by traffic sensors with statistics of microscopic driving traces under different levels of congestion. This approach can be used to better estimate freeway emissions when combined with an appropriate modal emissions model. Preliminary experimentation has been carried out with a vehicle equipped with global positioning system (GPS) instrumentation, allowing for precise localization in both space and time. With the GPS, second-by-second velocity traces are acquired and matched with simultaneously measured freeway traffic data obtained by embedded traffic sensors. Statistical measures of velocity variation are derived from the velocity traces and are functionally related to the macroscopic traffic parameters of speed, flow, and density. Given a known distribution of vehicle types, models, and model years, vehicle emissions can be related to these statistical measures of velocity variation using a modal emission model, and, thus given speed-flow-density measures of freeway traffic, localized emissions estimates can be made.
Get full access to this article
View all access options for this article.