Detection of Power System Fault Events and Location with Support Vector Machine-based Approach

• Main Authors: Lai, ChangChia, 賴昶嘉
• Other Authors: Chang, WenKung
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/44787165233725843672

碩士 === 國立中正大學 === 電機工程研究所 === 100 === Voltage and current deviations from their nominal values may result in serious damage or equipment malfunction. Therefore, the transmission line fault detection and classification are very important to the power system study. Commonly seen failures in the power system are three-phase balanced and unbalanced faults. There are several types of three-phase unbalanced faults, including single line-to ground, line-to-line, and double line-to-ground faults. Tracking the location of the fault line is also very important in the power system. Rapid diagnosis of the fault location can save manpower, reduce outage times, as well as losses. This thesis aims at applications of the support vector machine (SVM) classification techniques. The SVM is used to identify the fault event and the fault location. The proposed new feature selection can reduce the SVM training time. Also, this feature selection amplifies the characteristics of each fault event. Thus, it improves the SVM classification accuracy.