Leak Detection for Self-Contained Fluid-Filled Cables using Regression Analysis
Leak Detection for Self-Contained Fluid-Filled Cables using Regression Analysis
This paper investigates the methodology of the machine learning technique, namely the Support Vector Machine to assess the condition of fluid-filled high voltage cables based on thermal, pressure and load current information. Field data from a healthy circuit containing pressure, temperature and load current information have been obtained. The data structure has been investigated and a feasible algorithm to restructure the data for further analysis is proposed. The post-processing technique using Support Vector Machine Regression to predict oil pressure in the system is demonstrated. Results obtained using the regression analysis in this paper are very promising. Based on this method, an expert system could give early warning with better sensitivity than the existing system for the cable circuit and implementation of this approach can be achieved without taking the circuit out of service.
978-1-4244-6300-8
CD-ROM
Hao, L
e6006548-3fc1-4a7e-9df4-a4e9a9a05c45
Lewin, P L
78b4fc49-1cb3-4db9-ba90-3ae70c0f639e
Swingler, S G
4f13fbb2-7d2e-480a-8687-acea6a4ed735
Bradley, C
68d9d906-24a7-4cae-af25-bdef07406455
6 June 2010
Hao, L
e6006548-3fc1-4a7e-9df4-a4e9a9a05c45
Lewin, P L
78b4fc49-1cb3-4db9-ba90-3ae70c0f639e
Swingler, S G
4f13fbb2-7d2e-480a-8687-acea6a4ed735
Bradley, C
68d9d906-24a7-4cae-af25-bdef07406455
Hao, L, Lewin, P L, Swingler, S G and Bradley, C
(2010)
Leak Detection for Self-Contained Fluid-Filled Cables using Regression Analysis.
IEEE 2010 International Symposium on Electrical Insulation, San Diego, California, United States.
06 - 09 Jun 2010.
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
This paper investigates the methodology of the machine learning technique, namely the Support Vector Machine to assess the condition of fluid-filled high voltage cables based on thermal, pressure and load current information. Field data from a healthy circuit containing pressure, temperature and load current information have been obtained. The data structure has been investigated and a feasible algorithm to restructure the data for further analysis is proposed. The post-processing technique using Support Vector Machine Regression to predict oil pressure in the system is demonstrated. Results obtained using the regression analysis in this paper are very promising. Based on this method, an expert system could give early warning with better sensitivity than the existing system for the cable circuit and implementation of this approach can be achieved without taking the circuit out of service.
Text
036.pdf
- Version of Record
Restricted to Registered users only
Request a copy
More information
Published date: 6 June 2010
Additional Information:
Event Dates: 6 - 9 June 2010
Venue - Dates:
IEEE 2010 International Symposium on Electrical Insulation, San Diego, California, United States, 2010-06-06 - 2010-06-09
Organisations:
Electronics & Computer Science, EEE
Identifiers
Local EPrints ID: 271224
URI: http://eprints.soton.ac.uk/id/eprint/271224
ISBN: 978-1-4244-6300-8
PURE UUID: 532aec5f-1b55-46bc-ac9d-ccc67ced64d4
Catalogue record
Date deposited: 06 Jun 2010 19:50
Last modified: 15 Mar 2024 02:43
Export record
Contributors
Author:
L Hao
Author:
P L Lewin
Author:
S G Swingler
Author:
C Bradley
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics