基于多源信息融合的电网故障诊断方法研究  

Study of a fault diagnosis approach for power grid with information fusion based on multi-data resources

作　　者：高振兴[1] 郭创新[1] 俞斌[1] 骆玉海 彭明伟[3] 杨健[4] 

机构地区：[1]浙江大学电气工程学院,浙江杭州310027 [2]烟建集团有限公司,山东烟台264000 [3]浙江省电力设计院,浙江杭州310027 [4]上海市电力公司市区供电公司,上海200080

出　　处：《电力系统保护与控制》2011年第6期17-23,共7页Power System Protection and Control

基　　金：国家自然科学基金(50677062);新世纪优秀人才支持计划资助(NCET-07-0745);浙江省自然科学基金资助(R107062);国家863计划(2008AA05Z210);教育部博士点基金(20090101110058)

摘　　要：考虑到电力系统自动化水平及通信、广域测量技术的发展,提出了一种综合SCADA开关量、故障录波器电气量及WAMS系统电气量的多源信息融合电网故障诊断方法。该方法将蕴含故障信息的电气量分析与开关量诊断相结合,对电网故障采集的电气量通过小波能量分析提取故障特征,采用蕴含时序贝叶斯网络对保护、断路器开关量进行故障推理,定义了能量畸变故障度、能量故障度、改进RBF神经网络故障度及时序贝叶斯故障度衡量线路故障程度,并作为证据体采用改进D-S证据理论进行信息融合,进而通过模糊C-均值聚类方法给出故障诊断决策。PSCAD仿真及Matlab与Java混合编程计算表明,所提出的电网故障辅助诊断新方法相对传统开关量诊断,准确度得到了提高,具有工程实用价值和良好的应用前景。Considering the development of power system automation and communications and wide area measurement technology,this paper presents a fault diagnosis method for power grid with information fusion based on multi-data resources which contain the switching-status data from SCADA and continuous-time data derived from fault recorder and WAMS.Based on electrical value analysis with fault information and switch value diagnosis,this method employs the wavelet energy analysis to extract features from the continuous-time data and uses Bayesian networks with temporal order to perform fault reasoning on switching-time data of protection relays and breakers.During this process,fault degree of energy distortion（FDED）,energy fault degree（EFD）,fault degree based on improved RBF network（FDIR） and fault degree based on bayesian networks with temporal order（FDBT） are defined to indicate the situation of the fault on the lines,which is taken as evidences for information fusion by using improved D-S theory.Then the fuzzy C-means clustering method is involved to handle the fusion result and give decision-making.PSCAD simulation and calculations based on MATLAB programming with Java show that this proposed new approach for fault diagnosis significantly improves the diagnostic accuracy compared with the conventional way based merely on switching-time data,and has practical value and good application prospects.This work is supported by National Natural Science Foundation of China（No.50677062）,New Century Excellent Talents in University（NCET-07-0745）,Zhejiang Provincial Natural Science Foundation of China（No.R107062）,National High-tech Research and Development Program of China（863 Program）（No.2008AA05Z210）,and Doctorate Fund of the Ministry of Education（No.20090101110058）.

关 键 词：故障诊断 小波能量谱 改进RBF神经网络 时序贝叶斯网络 改进D-S理论 信息融合 FCM 

分 类 号：TM711[电气工程—电力系统及自动化]