摘要
基于统计学和模糊数学原理,提出一种用于电能质量可疑数据清理的云模型方法。以基波电压和基波电流为特征量,对于特征量的每一相日监测数据,采用逆向云发生器计算云模型的三个数字特征(Ex,En,He),根据云模型的"3En'规则"确定可疑数据的上、下阈值,通过遍历数据来检测可疑数据及其对应时间点。由于电能质量各项指标具有相位相关性和采样时间点同步性,故利用特征量可疑数据对应的相位和时间点来剔除其他指标的可疑数据。与正态分布相比,该方法考虑了样本方差的随机波动性,其确定的阈值不易造成正常数据的误判,算例分析验证了该方法在工程应用中的有效性。
Based on the principles of statistics and fuzzy mathematics, a method for cleaning suspicious data of power quality based on cloud model is proposed. The fundamental voltage and current are considered as characteristics; for the daily monitoring data of each characteristic phase, the three numerical characteristics （Ex, En, He） of cloud model are.calculated by the reverse cloud generator; the upper and lower threshold values are determined based on the cloud model ＂3En＇ rule＂ ; the suspicious data and its corresponding time are detected by traversing the data. Due to the phase correlation and sampling time synchronization of power quality indicators, the suspicious data of the other indicators are weeded out using the corresponding time of the characteristics. Compared with the normal distribution, this method takes into account the random volatility of the sample variance, so the threshold values are not easy to make the normal data misjudged. Calculation example verifies the effectiveness of this method in engineering applica- tion.
出处
《华东电力》
北大核心
2013年第8期1597-1600,共4页
East China Electric Power
作者简介
曾庆辉（1985-），男，工程师，主要从事电力测量、电能质量技术工作。
