Introduction: The data analysed in this pair are primarily time head ways, the elapsed time between the leading edges of successive vehicles. The intent of the analysis is to go beyond the curve fitting stage of headway distributions and use instead these headway measurements to understand the behaviour of the traffic in terms of its component parts. Considerable analysis has been done on the curve fitting of headway distributions, but so far there does not appear to be one completely satisfactory continuous solution with respect to volume. Detailed studies of head ways show a significant difference between traffic flow under congested and non-congested conditions. The search for one distribution to suit all these conditions tends to smooth out the differences rather than acknowledge their presence. It is these differences which will be studied in greater detail throughout the text of these notes. Conclusions: 1) Traffic may be effectively divided into platoons and groups in the study of traffic performance. The platoon definition is a noteworthy parameter. 2) Platoon head ways may be considered independent of both the numerical size of the platoon in which they occur and their relative position within that platoon. 3) Characteristic Volume, derived from time mean headway of platoons, gives a practical and consistent measure of limiting volume and may afford a relative measure of the variation in limiting volumes. The characteristic volume enables comparison of geometric design, location, lane position, weather and possibly various driver groups. 4) The effects of congestion on headway distributions may be accounted for by the use of characteristic volume. 5) Adequate headway distributions may be synthesises from individual platoon and group headway distributions. 6) The independence of platoons and groups infer that prior knowledge of a proceeding platoon or group does not effectively aid in the prediction of the character of a succeeding platoon or group. This result is significant when considering any type of predictive control system. 7) In the observation recorded, it appears that the platoon definition may be adjusted within contain limits without loss of generality in the conclusions drawn. The following conditions may be weighted in the selection of a platoon definition. a) Consistency of the characteristic headway. b) Linearity in the relative proportion of platoons versus volume. c) Consistency in the relative proportion of platoons for congestion and free flow at the same volume levels. e) Goodness of fit for the composite headway distribution of platoon and groups. f) A practical point in the selection of a platoon definition is the need to be with the class interval of the headway distribution. In this study the interval was 0.5 seconds. Platoon Definitions of 2.0 seconds or 2.5 seconds appear to be the best choices.
Abstract