An Application of Shortest Path Methods to A Pre-trip Planning System of Taiwan Railway Companies / 軌道運輸系統行前旅次規劃問題之研究

碩士 / 國立成功大學 / 交通管理(科學)學系 / 85 / In order to improve the utilization of public transportation
services,it is important to establish a pre-trip planing system
such that we canintegrate related service information of all
transportation companies aswell as to generate valuable trip
suggestions to possible trip makers.In this study, a prototype
trip planning system for the planed Taiwanrailway system is
discussed, which includes the western area's high speedrail, the
around island Taiwan railway, and some connecting transit
andcommuter rail lines. In the study, two issues on the trip
planning system are discussedin detail: first, how to create a
network data with sensible meaning inthe real world and
practical size for the computation works; and secondly,how to
select and use appropriate shortest path methods to generate
valuabletrip suggestions. A hierarchical network structure is
proposed for thesystem, in which only one fourth of railway
stations are considered in themain network as the nodes with
transfer possibility, and the possibleconnections between the
other stations and the main stations are consideredin the
second-level network. Only the trip planning problem on the
mainnetwork is investigated with the shortest path methods,
because the solutionof the problem on the second-level network
is in general straight forward.Then we use label setting method
matching with time-space network and DYNETalgorithm matching
with discrete dynamic network to solute the pre-tripplanning
problem of domestic rail transportation system through
thehierarchical network we proposed. Because of the practical
needs, we alsodeveloped k-shortest path method basing on the one
shortest path methodswe used to improve the trip suggestion
qualities. The major findings of the study are as follows:
1.We can reduce thenetwork data by using hierarchical network
structure. 2.Compared with the label setting method matching
with time-space network, DYNET algorithm matching with discrete
dynamic network has less running time. 3.By the means of k-
shortest path method, we can offer more than one kind of trip
suggestion for trip makers, and we can also provide the trip
suggestion with less transferring times for trip makers.

Identiferoai:union.ndltd.org:TW/085NCKU0119012
Date January 1997
CreatorsLee, Szu-Chien, 李思謙
ContributorsLee Chi-Kang, 李治綱
Source SetsNational Digital Library of Theses and Dissertations in Taiwan
Languagezh-TW
Detected LanguageEnglish
Type學位論文 ; thesis
Format107

Page generated in 0.0016 seconds