Title :
SA-SVM incremental algorithm for GIS PD pattern recognition
Author :
Di-bo Wang ; Ju Tang ; Ran Zhuo ; Jun-yi Lin ; Jian-rong Wu ; Xiao-Xing Zhang
Author_Institution :
State Key Lab. of Power Transm. Equip. & Syst. Security & New Technol., Chongqing Univ., Chongqing, China
Abstract :
With changes in insulated defects, the environment, and so on, new partial discharge (PD) data are highly different from the original samples. It leads to a decrease in on-line recognition rate. Using ultra-high frequency (UHF) cumulative energy and its corresponding apparent discharge as inputs, a support vector machine (SVM) incremental method based on simulated annealing (SA) is constructed. Examples show that the new method speeds up the data update rate and improves the adaptability of the classifier.
Keywords :
gas insulated switchgear; learning (artificial intelligence); partial discharges; pattern classification; power engineering computing; simulated annealing; support vector machines; GIS; PD; SA; SVM incremental method; UHF cumulative energy; insulated defect; online recognition rate; partial discharge; pattern classifier; pattern recognition; simulated annealing; support vector machine; ultra high frequency; Discharges (electric); Gas insulation; Optimization; Partial discharges; Pattern recognition; Support vector machines; Training;
Conference_Titel :
High Voltage Engineering and Application (ICHVE), 2012 International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4673-4747-1
DOI :
10.1109/ICHVE.2012.6357131
