10kV电缆接头典型缺陷局部放电类型模式识别  被引量:2

Pattern recognition of partial discharge types for typical defects in 10 kV cable joints

在线阅读下载全文

作  者:桂俊峰[1] 刘岳楠 逯华 耿民 邱超军 王喜莲[1] GUI Junfeng;LIU Yuenan;LU Hua;GENG Min;QIU Chaojun;WANG Xilian(School of Electrical Engineering,Beijing Jiaotong University,Beijing 100044,China;System Engineering Re-search Institute of China Shipbuilding Corporation,Beijing 100094,China;Motor Train Maintenance Department of CRRC Tangshan Locomotive and Rolling Stock Co.,Ltd.,Tangshan Hebei 063035,China)

机构地区:[1]北京交通大学电气工程学院,北京100044 [2]中国船舶集团有限公司系统工程研究院,北京100094 [3]中车唐山机车车辆有限公司动车检修部,河北唐山063035

出  处:《北京交通大学学报》2023年第3期140-148,共9页JOURNAL OF BEIJING JIAOTONG UNIVERSITY

基  金:国家自然科学基金(50907004)。

摘  要:针对配电网10 kV电缆接头典型缺陷常见放电类型识别精度低问题,根据电缆接头常见放电类型建立4种放电模型,搭建局放试验平台并获取模型的放电谱图.对谱图的分布特征、矩特征、纹理特征进行提取,对比反向传播(Back Propagation,BP)神经网络算法和网格搜索参数优化的支持向量机算法(Support Vector Machine,SVM)对放电模型特征序列的识别精度.制作4种电缆接头典型缺陷实物模型并获取其放电谱图,提取谱图分布特征、矩特征及纹理特征.对比BP神经网络、基于网格搜索参数优化的SVM和基于遗传算法参数优化的SVM等3种算法对电缆接头缺陷放电类型的识别精度.研究结果表明:基于遗传算法参数优化的SVM算法能对电缆接头缺陷放电谱图的组合特征进行准确识别,且识别率高于其他传统算法,在电缆局部放电模式识别领域具有一定的工程应用价值.To address the issue of low identification accuracy of common discharge types of typical de⁃fects in 10 kV cable joints in distribution network,this study established four discharge models based on the common discharge types of cable joints.A partial discharge test platform is built and the dis⁃charge spectrum of the model is obtained.The distribution,moment,and texture features of the spec⁃trum are extracted,and the recognition accuracy of the discharge model feature sequence is compared with the Back Propagation(BP)neural network and the Support Vector Machine(SVM)algorithm op⁃timized by the grid search parameters.Physical models of four typical defects in cable joints are cre⁃ated,and their discharge spectra are obtained.The distribution,moment,and texture features of the spectra are extracted.The recognition accuracy of three algorithms,namely BP neural network,SVM based on grid search parameter optimization,and SVM based on genetic algorithm parameter optimization,for cable joint defect discharge types is compared.The results show that the support vector ma⁃chine algorithm based on genetic algorithm parameter optimization can accurately identify the com⁃bined features of cable joint defect discharge spectrum,and the recognition rate is higher than other tra⁃ditional algorithms.This algorithm has engineering application value in the field of cable partial dis⁃charge pattern recognition.

关 键 词:电力电缆 中间接头 典型缺陷 局部放电 放电类型 特征提取 

分 类 号:TM247[一般工业技术—材料科学与工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象