摘要
阐述了用局部放电数字化—计算机测量装置检测放电信号,由放电统计特性来判断设备状态的意义。指出采用模式识别方法提高了放电识别的有效性。介绍了三种基于距离的模式识别方法:置信区间法、最小距离法、趋中心度法,分析了这三种方法的原理、算法和样本构成。最后对这三种用于识别放电的模式识别方法作了比较、归纳。
The significance of condition judgement of electrical apparatus based on statistical characteristics of partial discharge (PD) extracted from PD signals detected by digital and computer-aided PD measuring device is demonstrated. It is also pointed out that the efficiency of discharge recognition is improved by using pattern recognition methods. Three pattern recognition methods based on distance: confidence intreval method,minimum distance method and centour score method are introduced. The basic principles and algorithms, samples constitutions are analysed.
出处
《高压电器》
CAS
CSCD
北大核心
1997年第3期3-7,共5页
High Voltage Apparatus
基金
国家自然科学基金项目(批准号:59637200)
