TRIP MATRIX ESTIMATION FROM TRAFFIC COUNTS USING CUBE

Subha Nair, Nisha Radhakrishnan, Samson Mathew

Abstract: This paper presents a distinguished method of integrating Direct and Indirect Methods of estimating O-D Matrix. Direct method includes household and demographic data, whereas indirect method includes traffic counts. CUBE is highly efficient transport modeling software whereas ArcGIS is a geospatial tool which gives the geographic information of the land. The various factors influencing the trip generation, considered in this paper are household data like employment, population, income and average household occupancy and residential density. The secondary data are collected from the municipal offices and other local government bodies of Tiruchirappalli district of Tamil Nadu. The traditional four-step modeling is done in CUBE, using these factors, to obtain a priori O-D matrix. Further, traffic counts in the concerned network are determined using traffic surveys in major arterial roads in Tiruchirappalli city. For updating the existing Origin- Destination matrix using link volumes the gradient descent approach has been used . In this paper a unique attempt has been made to script a bi-level program in the CUBE software. The traffic counts are used to remodel in Cube using to get an updated matrix

Keywords: Bi-level programming, CUBE 6.0, Gradient method, Link Volume, O-D matrix, Regression, Travel demand

DOI: https://doi.org/10.15623/ijret.2013.0213025