A Novel Road Sectional Characteristic Cluster Matching Algorithm

This paper aims to develop a novel road sectional characteristic cluster matching algorithm for selecting the highlycorrelated road sectional samples and studying the relationship between Physical Characteristics (PC) and Traffic Flow Characteristics (TFC). After analyzing the mapping relationship between PC and TFC of road sections, a system optimization-based cluster matching algorithm is proposed. Compared with the individual optimization-based algorithm, the proposed algorithm could provide more accurate matching results, especially when the road sectional sample size is large and the cluster numbers of both physical characteristic matrix and traffic flow characteristic vector are small.