Commodity-Based Truck Origin–Destination Matrix Estimation Using Input–Output Data and Genetic Algorithms

Abstract

A commodity-based model to estimate a truck origin–destination (O-D) matrix is presented. The model takes advantage of the genetic algorithm global search method to find the best O-D matrix that when assigned to the network gives the minimum deviation between observed and estimated data. The model is flexible with respect to the type of data used in estimating the O-D matrix; however, the case study presented in this paper takes into consideration only two sets of information: commodity flow on specific links and column and row sums of the O-D matrix. Flows are treated as commodity dollar value; therefore, the estimated O-D matrix entries consist of the value of the commodity shipped by truck from the origin zone to the destination zone. The method is composed of two submodels. The first submodel, the trip generation model, uses input–output data with employment and population data to estimate the zonal level of commodity attraction and production. The second submodel, the genetic algorithm model, searches globally for the optimum O-D matrix. The model and its application to a case study of a region in Ontario, Canada, are presented. Directions for future research are provided.