抗不良杠杆量测的稳定收敛状态估计方法.pdf

《抗不良杠杆量测的稳定收敛状态估计方法.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43卷第4期电子科技大学学报Vol.43No.42014年7月JournalofUniversityofElectronicScienceandTechnologyofChinaJul.2014·自动化技术·抗不良杠杆量测的稳定收敛状态估计方法113121余娟，廖龙飞，朱黎丽，颜伟，赵雪骞，赵霞(1.重庆大学输配电装备及系统安全与新技术国家重点实验室重庆沙坪坝区400044；2.国网北京电力科学研究院北京丰台区100075；3.重庆工程职业技术学院电气工程学院重庆沙坪坝区400037)【摘要】不良杠杆量测使得

2、传统的含有不良数据辨识的状态估计方法失效，严重降低了状态估计的精度，为了能够自动排除不良杠杆量测的干扰，同时保证状态估计计算的数值稳定性和收敛性，提出了抗不良杠杆量测的稳定收敛状态估计方法。在利用等价权原理并统一量测残差尺度的基础上，提出了采用标准化残差的指数型权函数，并基于此得到抗差状态估计模型。该模型能够根据量测残差自动修正量测权重，具有较强地抵抗不良杠杆量测的能力，能有效保证算法的数值稳定性和收敛速度。分别采用3节点和IEEE118节点算例对该方法的特点进行了验证。关键词不良杠杆量测;收敛性;指数型权函数

3、;数值稳定性;抗差状态估计中图分类号TM744文献标志码Adoi:10.3969/j.issn.1001-0548.2014.04.014StableConvergenceStateEstimationMethodforBadLeverageMeasurements113121YUJuan,LIAOLong-fei,ZHULi-li,YANWei,ZHAOXue-qian,andZHAOXia(1.StateKeyLaboratoryofPowerTransmissionEquipment&SystemSecu

4、rityandNewTechnology,ChongqingUniversityShapinbaChongqing400044;2.StateGridBeijingElectricPowerResearchInstituteFengtaiBeijing100075;3.CollegeofElectricalEngineering,ChongqingInstituteofEngineeringShapingbaChongqing400037)AbstractTraditionalstateestimationme

5、thodswhichincludebaddataidentificationfailinbadleveragemeasurements,resultingthedegradationofstateestimationprecision.Inordertoexcludetheinterferenceofbadleveragemeasurementsautomatically,whileensuringthenumericalstabilityandconvergenceofstateestimationcalcu

6、lation,thispaperproposesastableconvergencestateestimationmethodforbadleveragemeasurements.Onthebaseofutilizingtheprincipleofequivalentweightandunifyingthescalesofmeasurementresiduals,thispaperpresentstheexponentialweightfunctionbyusingthestandardizedresidual

7、andbuildsthemodelofrobuststateestimation.Thismodelcancorrecttheweightofmeasurementautomaticallyaccordingtothemeasurementresidual,therefore,showsthestrongabilitytoresistbadleveragemeasurementsandeffectivelyguaranteethenumericalstabilityandconvergencespeed.The

8、characteristicsoftheproposedmethodaredemonstratedonthe3-busandthe118-bussystems.Keywordsbadleveragemeasurements;convergence;exponentialweightfunction;numericalstability;robuststateestimation[2-3