Estimation of origin‐destination matrix from traffic counts: a comparison of entropy maximizing and information minimizing models

Abstract

The models based on entropy maximization (EM) and information minimization (IM) techniques are the most commonly used methods for the estimation of origin‐destination (O‐D) matrix from traffic counts. However, the accuracy of these models have not yet been well defined and there is no valid proof that can be used to give preference to either of these two models. This paper investigates how data affect the EM/IM model performance and compares the effects of data variability on model accuracy. A computer algorithm has been developed to solve the EM and IM problems for the estimation of O‐D matrix from traffic counts. The validation of the EM and IM models is based on a comprehensive dataset collected in Shenzhen (a special economic zone in China) using roadside interviews; this included O‐D and route choice information as well as traffic counts. A comparison of results given by the two models shows that the EM model performs better when prior information is not available, but that the performance of the IM model is better if a good prior matrix can be used. However, both the EM and IM models would under‐estimate the total number of O‐D flows.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.