| Author | Paramate Leartsakulpanitch |
| Call Number | AIT Thesis no. TE-98-04 |
| Subject(s) | Traffic estimation--Simulation methods
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Civil Engineering |
| Publisher | Asian Institute of Technology |
| Abstract | Traffic simulation techniques are standard tools to analyze complex traffic engineering problems. The use of traffic simulation techniques enables traffic engineers to determine the performance of a traffic network. However, the accuracy of simulation results continues to be a significant drawback to the use of these techniques. Up to now, much research has been conducted on traffic states estimation for freeway system. The treatment of urban road traffic has been limited to the simplification of freeway models due to the complexities of urban traffic. Nevertheless, such simplifications can cause adverse effect on the estimation precision. To improve the accuracy of estimation, a new scheme for the surveillance of traffic states of an urban road network is proposed. The method is based upon the application of Kalman Filtering Method to the nonlinear-dynamic, macroscopic model, which was proposed by Cremer and Putensen (1992). However, due to the characteristics of traffic states in urban road networks, the macroscopic model used here is modified in accordance with where segment is, particularly the first and last segment of each road/link, and intersection segments and with what traffic signal is. To investigate the applicability of this technique, a simulation program, KNWSML (Kalman Filtering for Network Simulation) is developed by combining macroscopic model and Kalman Filter method. All in all, firstly, model parameters used in KNWSML are calibrated by using the simulation data generated from the other reliable simulation program, TRAF-NETSIM. Then, by using the simulation data for extensive traffic situations, two verifications are conducted to test these two subprograms of KNWSML. The first verification illustrated that the macroscopic subprogram in KNWSML can predict the traffic estimation in different situations quite well. For the second verification, it investigates how effective Kalman Filtering technique is on the estimation of traffic states. The results show that this technique can estimate traffic states in urban road networks with reliability and consistency for each traffic situation. |
| Year | 1998 |
| Type | Thesis |
| School | School of Civil Engineering |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Transportation Engineering (TE) |
| Chairperson(s) | Nakatsuji, Takashi; |
| Examination Committee(s) | Yordphol Tanaboriboon;Pannapa Herabat; |
| Scholarship Donor(s) | Government of Japan.; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1998 |