VEHICLE CLASSIFICATION FROM WEIGH-IN-MOTION DATA - THE PROGRESSIVE SIEVING ALGORITHM

A classification procedure has been developed, based on the concept of progressive sieving of weigh-in-motion data through a series of crite ria designed to separate, identify, and classify vehicles by their axl e arrangement. The procedure is demonstrated to be feasible, and is ca pable of identifying all major types of vehicles quite accurately. The procedure together with a nomenclature provides a classification syst em that is open-ended having an infinite number of classes. The system is universal and could be used by any jurisdiction. It facilitates ev aluation of vehicle weight and dimension policy alternatives and their impacts on the highway system. The procedure has the capability of tr anslating data from one classification system into another.