Thesis (M.Eng.) - Asian Institute of Technology, 1998

A thesis proposal submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Civil Engineering

Macroscopic models of traffic situations, which are able to handle the large size of freeway system with the fast simulation time, are utilized for various purposes in designing and operating freeways. They provide an economical and effective way to evaluate alternative system design and control strategies. The models contain a set of parameters, which have to be estimated according to real traffic data. Model parameters have significant effects on the performance of the model. Various techniques had been used to identify the model parameters. This study focuses on the comparison of parameter estimation techniques. Three . techniques, which are Nonlinear Least Square method, Box Complex method, and Kalman filtering method, were applied to a macroscopic model. Traffic data using in this study were collected from the second stage expressway of Bangkok. The estimation performances of each technique were investigated. Then the results were compared. Finally, the relationships between macroscopic parameters and traffic situation were investigated.