Modeling rural road transport demand based on genetic algorithm

Abstract

The paper describes the use of stochastic search process that is the basis of Genetic Algorithms (GAs) in developing transport demand in rural roads of Turkey. Travel demand, vehicle movements and demand of goods transport are estimated based on the socioeconomic indicators that are the Gross National Product (GNP), population and number of vehicles. Various forms of the Genetic Algorithm Travel Demand (GATT) models are developed. Weighting parameters of the GATT models are estimated using the historical data. Available data is partly used for estimating weighting parameters of the GATT and partly for testing the models. GATT models are validated with observed data, and then future estimation of travel demand is projected until 2025. Results are compared with European Union (EU) countries and suggestions are made in the long term in terms of transport demand planning.