Neural Network-based O-D Matrix Estimation from Link Traffic Counts

Abstract

The Origin—Destination (O—D) trip table is an essential ingredient in a wide variety of travel analysis and planning studies and increasing attention has been paid to methods for the estimation of O—D matrices from traffic counts over past decade. The O—D estimation from link traffic counts has attracted lot of interest and many methods have been proposed to obtain O—D trip tables based on link counts. In this study, we propose a neural network model for O—D estimation from link traffic counts in more complicated urban street networks. Moreover, a training method for the application of a neural network model in O—D estimation has been proposed in order to achieve more accurate predictive results. The experimental results showed that the proposed neural network model can be much more efficient and accurate than currently usual methods based on traffic assignment models.

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 — 7 citations exist upstream.

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