A Study on Mass-Events Evacuation Network Design Model

碩士 === 國防管理學院 === 後勤管理研究所 === 96 === The run-up of the mass-event often attracts a large number of trips gathering once. Therefore, how to effectively evacuate trips due to attraction from mass-event would be an important task. In order to benefit the trip evacuation, we must propose a comprehensive...

Full description

Bibliographic Details
Main Authors: Tsung-Yun Wu, 吳宗昀
Other Authors: Chung-Yung Wang
Format: Others
Language:zh-TW
Published: 2008
Online Access:http://ndltd.ncl.edu.tw/handle/50019560437621172705
Description
Summary:碩士 === 國防管理學院 === 後勤管理研究所 === 96 === The run-up of the mass-event often attracts a large number of trips gathering once. Therefore, how to effectively evacuate trips due to attraction from mass-event would be an important task. In order to benefit the trip evacuation, we must propose a comprehensive plan including a perfect evacuation network in advanced so that the transportation network of the entire event area would be operated smoothly. When the mass-event is held, whether or not activities can succeed will be highly influenced by the design of entire transportation network system. As a result, how to set up a transportation network for mass-event evacuation which gives consideration to both event attendees and local residents at the same time is the primary subject of this research. This research is all about network-design problems, it constructs bi-level programming models. The upper level model is to confer the strategy of how to build an evacuation network with minimum total traveling cost, and the lower level model considers constraints for the evacuating routes and positions. Because the solution sets of this research belong to non-convex and mixed 0-1 integer nonlinear programming, the examination process is extremely difficult. We worked on solutions by using Variational Inequality Method, Gradient Projection Method and Simulated Annealing Method. Moreover, the development and innovation of research methods is another emphasis of this research. At last, we used numerical samples to demonstrate the accuracy and practicability of the models. The result of this research is an important reference which can contribute the decision-making process for government, military units and other administrations when holding mass-event related activities in the future.