COST EFFECTIVE METHODS OF ESTIMATING O-D FLOWS FROM TRAFFIC COUNT AND LICENCE PLATE SURVEY INFORMATION
Itzhak Geva, Ezra Hauer, Uzi Landau
Abstract
Itzhak Geva, Ezra Hauer, Uzi Landau
Abstract
A coherent method for the estimation of o-d flows from traffic count and license plate survey information is presented. The estimation method identifies the most likely set of flows which agrees with the observed approach counts on a stretch of road and the results obtained by the sample license plate survey. The solution procedure is relatively simple and the method is illustrated by numerical example. The effect of the sample size on estimation accuracy is examined, using data from a comprehensive license plate survey conducted on a stretch of road in Toronto. It was expected to find that accuracy increases with the sampling ratio. It seems that good estimates can be obtained for small systems for flows in one direction without any sample when turning movements at each intersection are also known. The main reason for that is that the assumption of equiprobable linkage is satisfactory for such systems. For larger systems, however, when there are two directions of flow, estimates obtained without any sample are very poor compared to those obtained with small sample ratios. It appears that o-d flows for these systems may be well estimated by this method. For the covering abstract of the conference see TRIS 378581. (TRRL)
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A coherent method for the estimation of o-d flows from traffic count and license plate survey information is presented. The estimation method identifies the most likely set of flows which agrees with the observed approach counts on a stretch of road and the results obtained by the sample license plate survey. The solution procedure is relatively simple and the method is illustrated by numerical example. The effect of the sample size on estimation accuracy is examined, using data from a comprehensive license plate survey conducted on a stretch of road in Toronto. It was expected to find that accuracy increases with the sampling ratio. It seems that good estimates can be obtained for small systems for flows in one direction without any sample when turning movements at each intersection are also known. The main reason for that is that the assumption of equiprobable linkage is satisfactory for such systems. For larger systems, however, when there are two directions of flow, estimates obtained without any sample are very poor compared to those obtained with small sample ratios. It appears that o-d flows for these systems may be well estimated by this method. For the covering abstract of the conference see TRIS 378581. (TRRL)
Key concepts: Intersection (aeronautics), Sample (material), License, Sampling (signal processing), Statistics, Sample size determination, Traffic count, Traffic flow (computer networking)