Citation

Ferrari P, Treglia P, Cascetta E, Nuzzolo A, Olivotto P. Transp. Res. B Methodol. 1982; 16(5): 399-418.

Copyright

(Copyright © 1982, Elsevier Publishing)

DOI

10.1016/0191-2615(82)90021-2

PMID

unavailable

Abstract

The paper presents a method of measuring the quality of circulation worked out by processing the data derived from 155, 000 vehicle passages over two sections of two-lane, divided-carriageway motorway of identical geometric characteristics. The mass of observed data was broken down into a sequence of 293 periods of constant flow rate, each characterized by a specific flow rate and percentage of heavy vehicles. The percentage of vehicles passing on lane No. 1 and the suitable transformations of the characteristics of the speed and time headway processes on the two lanes during each of the periods gave rise to a vector of 15 components representing the circulation conditions during that period. By performing a linear transformation of the vectors for the 293 constant flow rate periods, it was possible to single out three non-correlated variables, called the "principal components", which accounted for 75% of the total variance observed for the 15 components, and in turn permitted the representation, without excessive distortion, of the observed circulation conditions within a three-dimensional space by means of a swarm of 293 points. Analyzing the manner in which the three principal components varied jointly within the swarm, a total of four zones were identified, representing respectively free flow, stable flow, unstable flow and forced flow. Moreover, it was possible to identify the numerical value of the quality of circulation with the first principal component, which by itself accounted for 55% of the total variance. This component showed very strong correlation with the flow rate and the percentage of heavy vehicles, and thus permits among other things the mutual comparison of the conditions of circulation caused by different flow rates and compositions.

Language: en