Loading...

A Heuristic Method to Estimate Time-Dendent Origin-Destination Matrix Based on Fuzzy Theory

Khavarian, Khashayar | 2014

681 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46510 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Shafahi, Yusef
  7. Abstract:
  8. One of the most important required data in order to consult transportation studies is the amount of trip demand between different zones in the study region. Trip demand is usually presented in form of a table. Due to dependence of the demand on time, it gets classified in two categories: Static and Time dependent. In Time dependent problems, which the scope of this research is, trip demand is affected by network condition and the amount of traffic supply in short periods of time, e.g. 15 minutes. Using Traffic count and prior trip demand table in order to estimate the time dependent Trip demand is among the newest methods of travel demand estimation. In this research despite of older studies in this topic, efforts have been made to consider the vagueness of values using Fuzzy theories and present a model for time dependent trip demand estimation. The presented method is a bi-level optimization model. The lower level is a dynamic traffic assignment problem and the upper level is an optimization problem to estimate trip demand. To do the dynamic assignment of traffic, simulation software called Mezzo has been used. In order to solve the upper level problem a metaheuristic method based on Particle swarm optimization is provided. The proposed method and solution algorithm has been used for two made-up networks and the test network of Sodermalm, which is part of the Sweden capital city, and obtained good results. As an example, in Sodermalm case the consumed time for solving the demand estimation problem has been reduced from 10 days to 5. In this research two groups of indexes has been used for evaluating the goodness of results. The first group of indexes is from previous researches, which are RMSE, RMNSE and MNE. The second group is proposed in this research and is for comparing fuzzy results based on Fatourechi fuzzy distance index. In this research the first group of indexes is reduced about 40 percent for the Sodermalm city
  9. Keywords:
  10. Fuzzy Theory ; Origin-Destination Matrix ; Travel Demand Estimation ; Particles Swarm Optimization (PSO) ; Time Dependent Origin-Destination Demand

 Digital Object List

 Bookmark

No TOC