Estimating intersection turning movement proportions from less-than-complete sets of traffic counts

Gary A Davis, Chang Jen Lan

Research output: Contribution to journalArticlepeer-review

16 Scopus citations

Abstract

Estimated turning movement proportions are used in a number of traffic simulation and traffic control procedures to predict the turning movement flows at intersections. Historically, these proportions have been estimated by manual counting, but the ongoing deployment of real-time adaptive traffic control strategies indicates that the ability to automatically estimate these proportions from traffic detector data is becoming increasingly important. When it is possible to count the vehicles both entering and exiting at each of an intersection's approaches, methods based on ordinary least squares can produce usable estimates of the turning movement proportions, but when the number or placement of the detectors does not support complete counting, these methods fail. The feasibility of estimating turning movement proportions from less-than-complete sets of traffic counts is assessed, and the statistical properties of less-than-complete count estimates are compared with estimates generated from complete counts. It turns out that estimation from less-than-complete counts can be done as long as the detector configuration satisfies an identifiability condition. A numerical test is presented to assess whether or not this condition is satisfied, along with some simple rules for designing detector configurations that are likely to satisfy this condition. A Monte Carlo experiment suggests that estimates generated from less-than-complete counts can be more variable than those generated from complete counts.

Original languageEnglish (US)
Pages (from-to)53-59
Number of pages7
JournalTransportation Research Record
Issue number1510
StatePublished - Jul 1995

Fingerprint

Dive into the research topics of 'Estimating intersection turning movement proportions from less-than-complete sets of traffic counts'. Together they form a unique fingerprint.

Cite this