欢迎来到天天文库
浏览记录
ID:57750728
大小:475.01 KB
页数:7页
时间:2020-03-28
《考虑数据聚类特性的电能质量分级预警阈值设置方法.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第37卷第3期:0088—0094电力电容器与无功补偿Vo1.37,No.3:0o88—00942016年6月PowerCapacitor&ReactivePowerCompensationJun.2016DOI:IO.14044/j.1674-1757.pcrpc.2016.03.018考虑数据聚类特性的电能质量分级预警阈值设置方法卢验锋,郜幔幔,钟红梅,李凤珍s(1.广东电网公司惠州供电局,广东惠州516001;2.华南理工大学电力学院,广东广州510640;3.广州市奔流电力科技有限公司,广东广州510640
2、)摘要:针对传统电能质量分级预警的等级区间1分及阈值设置僵硬.难以反映和体现不同预警对象对电能质量的敏感程度和差异化诉求的问题.提出一种考虑数据聚类特性的电能质量分级预警阈值设置方法。首先。在系统分析了国家电能质量标准及传统电能质量等级划分的基础上.引入聚类方法对各预警对象进行聚类划分.将同类预警对象的各电能质量指标值的算术平均值作为该类对象的个性化预警限值,以深度挖掘、体现和反映各类预警对象对电能质量的个性化要求;然后,依据各类预警对象的个性化预警限值与国标限值及传统电能质量各等级区间的关系.分6种情况重新设置“
3、因情而异”的电能质量等级区间及阈值.有效避免因各级阈值设置不当造成漏报警或频繁报警现象;最后,以某供电局20个10kV电压等级的公共连接点的电能质量分级预警为例.验证了本文方法的灵活可调性和广泛适用性关键词:电能质量:分级预警;预警阈值:模糊聚类GradedEarlyWarningThresholdSettingMethodofPowerQualityConsideringtheDataClusteringFeaturesLUYanfeng,GAOManman:,ZHONGHongmei,LIFengzhen(1.
4、HuizhouPowerSupplyBureau,GuangdongPowerGrid,Huizhou516001,China;2.SchoolofElectricPower,SouthChinaUniversityofTechnology,Guangzhou510640,Province,China;3.GuangzhouPowerElectricTechnologyCo.,Ltd.,Guangzhou510640,China)Abstract:Asforsuchissueasinflexiblegradedin
5、tervalclassificationandthresholdsettingoftraditionalpowerqualityanddificultyinreflectingtheappealofdifferentearlywarningobjectstothesensitivityanddifferenceofpowerquality,akindofgradedearlywarningthresholdsettingmethodofpowerqualityconsideringthedataclustering
6、featuresisproposed.Firstly,theclusteringmethodisintroducedtomakeclusteringclassificationofeachearlywarningobjectandtotakearithmeticmeanvalueofeachpowerqualityindexofthesametypeofearlywarningobjectasindividualearlylimitvalueoftheobjectonthebasisofanalyzingsyste
7、maticallyChinanationalpowerqualitystandardandgradeclassificationoftraditionalpowerquality,SOtoexploitandreflectdeeplytheindividualrequirementofeachearlywarningobjecttothepowerquality.Then,powerqualitygradeintervalandthresholdvalueissetas6casesbasedondiferentsi
8、tuationinaccordancewiththerelationsbetweenindividualearlywarningvalueandinternationallimitofeachtypeofearlywarningobjectandeachgradeintervaloftraditionalpowerquality,whichavoidseff
此文档下载收益归作者所有