DocumentCode
2907824
Title
Support Vector Machine and Polynomial Fitting Approach for radio effective zones under Power Lines in Taiwan Railway
Author
Whei-Min Lin ; Ting-Chia Ou ; Chiung-Hsing Chen ; Rong-Ching Wu
Author_Institution
Nat. Sun Yat-Sen Univ., Kaohsiung
fYear
2007
fDate
5-8 Nov. 2007
Firstpage
1
Lastpage
6
Abstract
The purpose of this paper is to research the radio effective zones related to the service of a radio system for the dispatch of trolleys and the operation of level-crossings in the service of Taiwan Railways. The service uses radio and radio propagation to dispatch messages to trolleys and level-crossings throughout the railway system when the critical situation is occurred or the emergency button is pressed by users to allow the train driver to break subsequently. In order to estimate the radio effective zone, we perform some realistic measurements related to level-crossings and trolleys. Then, the data is analyzed using a combination of support vector machine (SVM) and polynomial approximation to estimate the radio effective zone, and create a new optimal model specific to the needs of Western Line, Pingtung Line and the Kaohsiung City local line in the Taiwan Railway context.
Keywords
polynomial approximation; power overhead lines; radiowave propagation; railway communication; railway electrification; support vector machines; trolleys; Kaohsiung City local line; Pingtung Line; Taiwan Railway power lines; Western Line; level-crossing operation; polynomial fitting approach; radio effective zones; radio propagation; support vector machine; trolley dispatch; Data analysis; Flowcharts; MIMO; Polynomials; Protection; Radio propagation; Rail transportation; Support vector machine classification; Support vector machines; Wireless communication; Polynomial approximation; Support vector machine;
fLanguage
English
Publisher
ieee
Conference_Titel
Intelligent Systems Applications to Power Systems, 2007. ISAP 2007. International Conference on
Conference_Location
Toki Messe, Niigata
Print_ISBN
978-986-01-2607-5
Electronic_ISBN
978-986-01-2607-5
Type
conf
DOI
10.1109/ISAP.2007.4441682
Filename
4441682
Link To Document