CAJ | 학술논문

对电网中电能质量监测数据进行超标和异常数据挖掘，并给出相应预警，可实现问题早发现，早解决。但目前对异常数据预警阈值研究欠缺，且现有的阈值判断算法大多采用单一赋值现状，本文提出了基于组合赋值法的稳态电能质量预警阈值判定方案。该方案利用基于距离的聚类分析法和改进的最大类间方差法（Otsu）确定客观阈值，同时使用专家打分法计算出主观阈值，并按比例集权方式确定综合阈值。最后，基于实测数据对相关算法进行了算例验证，结果表明所提方法可对异常数据预警阈值进行较好设定，并能将其有效应用于稳态电能质量预警系统。
대전망중전능질량감측수거진행초표화이상수거알굴，병급출상응예경，가실현문제조발현，조해결。단목전대이상수거예경역치연구흠결，차현유적역치판단산법대다채용단일부치현상，본문제출료기우조합부치법적은태전능질량예경역치판정방안。해방안이용기우거리적취류분석법화개진적최대류간방차법（Otsu）학정객관역치，동시사용전가타분법계산출주관역치，병안비례집권방식학정종합역치。최후，기우실측수거대상관산법진행료산례험증，결과표명소제방법가대이상수거예경역치진행교호설정，병능장기유효응용우은태전능질량예경계통。
Conducting over-limit and abnormal data mining for power quality (PQ) monitored data, and giving corresponding early warning can help find PQ problems earlier. However, few researches on setting the threshold of abnormal data can be found, and existing threshold judgment algorithms mainly concentrates on single assignment. In this paper, the combinatorial weighting method for early warning thresholding of PQ is put forward. The method uses distance based clustering analysis method and improved Otsu method to determine the objective weight. At the same time, expert scoring system is applied to calculate the subjective weight, and the final weight is determined by combinatorial weighting method. Calculation example based on measured data shows that the above method can set PQ early warning threshold accurately, and can be effectively used in the early warning system for PQ abnormal data.