基于卷积原型网络的断路器故障诊断方法研究

doi:10.16628/j.cnki.2095-8188.2024.02.010

摘要/Abstract

摘要：

针对现有断路器故障诊断研究中无法有效区分未知类样本的问题,提出了一种基于卷积原型网络的断路器故障诊断算法。首先,以聚类的思想构建分类函数,通过各类故障的原型样本点特征空间距离约束来划分概率空间,实现对包含未知类故障样本集的识别。同时,以原型样本点为聚类中心,将样本特征的空间距离作为卷积特征自提取网络的优化目标,以有效改善样本特征的类内聚集性及类间的分散性,提高模型对样本的分类准确度。最后,基于110 kV断路器现场实验数据,对所提算法的有效性和准确性进行验证。结果表明,所提算法能够准确区分测试样本中的未知故障,并有效改善了故障样本特征的空间分布。

关键词: 断路器, 故障诊断, 原型卷积网络, 聚类, 未知类, 智能运维

Abstract:

Aiming at the problem that unknown samples cannot be effectively distinguished in existing circuit breaker fault diagnosis research, a circuit breaker fault diagnosis algorithm based on the convolutional prototype network is proposed. Firstly, the classification function is constructed using the clustering approach and the probability space is divided based on the distance constraint of the prototype sample point feature space for various types of faults, which can achieve the recognition of a sample set containing unknown fault classes. At the same time, with each type of prototype sample points as the cluster center, the sample feature space distance is used as the optimization target of the convolution feature self-extraction network, which can effectively improve the intra-class aggregation and inter-class dispersion of fault sample features and improve the classification accuracy of the model. Finally, the validity and accuracy of the proposed algorithm are verified based on the field experiment data of 110 kV circuit breaker. The results show that the proposed algorithm can accurately distinguish the unknown faults in the test samples and effectively improve the spatial distribution of fault sample features.

Key words: circuit breaker, fault diagnosis, prototype convolutional network, clustering, unknown class, intelligent maintenance

中图分类号:

TM561

沙浩源, 刘佩, 王之赫, 孙毅, 赵贺, 邓凯, 朱超. 基于卷积原型网络的断路器故障诊断方法研究[J]. 电器与能效管理技术, 2024, 0(2): 56-65.

SHA Haoyuan, LIU Pei, WANG Zhihe, SUN Yi, ZHAO He, DENG Kai, ZHU Chao. Research on Fault Diagnosis Method of High Voltage Circuit Breaker Based on Convolution Prototype Network[J]. LOW VOLTAGE APPARATUS, 2024, 0(2): 56-65.

图/表 16

图1

图2

表1

图3

图4

图5

表2

图6

图7

图8

图9

图10

图11

图12

图13

表3

参考文献 29

[1]	RAZI-KAZEMI A A, VAKILIAN M, NIAYESH K, et al. Priority assessment of online monitoring investment for power system circuit breakers-part I:qualitative-quantitative approach[J]. IEEE Transactions on Power Delivery, 2013, 28(2):928-938. doi: 10.1109/TPWRD.2013.2243921
[2]	赵莉华, 付荣荣, 荣强, 等. 基于自适应神经模糊推理系统的高压断路器操作机构状态评估[J]. 高电压技术, 2017, 43(6):2007-2015.
[3]	钟声, 李晓洋, 梁胜乐, 等. 基于线圈电流信号及动态时间规整的高压断路器状态评估方法[J]. 高压电器, 2023, 59(4):24-31.
[4]	孙曙光, 张强, 杜太行, 等. 基于振动信号的低压万能式断路器分合闸故障程度评估方法的研究[J]. 中国电机工程学报, 2017, 37(18):5473-5482.
[5]	赵书涛, 马莉, 朱继鹏, 等. 基于CEEMDAN样本熵与FWA-SVM 的高压断路器机械故障诊断[J]. 电力自动化设备, 2020, 40(3):181-186.
[6]	黄新波, 许艳辉, 朱永灿. 基于混合分类器的高压断路器故障诊断[J]. 高压电器, 2022, 58(10):149-157.
[7]	杨秋玉, 阮江军, 张灿, 等. 基于定量递归分析的高压断路器机械缺陷辨识及应用[J]. 电工技术学报, 2020, 35(18):3848-3859.
[8]	廖斌, 梁晨, 陈松, 等. 一起35kV并联电容器装置故障原因分析[J]. 电力电容器与无功补偿, 2022, 43(5):1-6.
[9]	RAOP, HUANG J, HU X G, et al. Testing of circuit breakers using coil current characteristics analysis[C]// IEEE Control Autom.Conf., 2009:185-189.
[10]	JOHAL H, MOUSAVI M J. Coil current analysis method for predictive maintenance of circuit breakers[C]// IEEE/PES Transmission Distribution Conference, 2008:1-7.
[11]	HU X G, CHAO L. Research on the condition parameter tester of high voltage circuit breakers[C]// EEE Conference on Industrial Electronics & Applications, 2008:2389-2393.
[12]	RAZI-KAZEMI A A, VAKILIANM, NIAYESH K, et al. Circuit-breaker automated failure tracking based on coil current signature[J]. IEEE Trans Power Deliv, 2014, 29(1):283-290. doi: 10.1109/TPWRD.2013.2276630
[13]	NATTI S, KEZUNOVIC M. Assessing circuit breaker performance using condition-based data and Bayesian approach[J]. Electric Power Systems Research, 2011, 81(9):1796-1804. doi: 10.1016/j.epsr.2011.04.010
[14]	KEZUNOVIC M, NATTI S. Risk-based maintenance approach:a case of circuit breaker condition-based monitoring:3rd international CIGRE workshop on liberalization and modernization of power systems[C]. 2006.
[15]	靳文娟, 季天瑶, 唐文虎. 基于多参量的高压断路器分/合闸线圈的故障诊断[J]. 高压电器, 2019, 55(3):226-233.
[16]	周小娜. 10 kV智能高压真空断路器在线监测系统研究与设计[D]. 厦门: 厦门理工学院, 2016.
[17]	张佳, 陈志英, 陈丽安, 等. 基于改进集合模态分解的真空断路器分合闸线圈电流特征值提取[J]. 高压电器, 2020, 56(12):116-123.
[18]	梅飞, 梅军, 郑建勇, 等. 粒子群优化的KFCM及SVM诊断模型在断路器故障诊断中的应用[J]. 中国电机工程学报, 2013, 33(36):134-141,19.
[19]	梅飞, 梅军, 郑建勇, 等. 基于KPCA-SVM的断路器故障稳健诊断方法[J]. 电工技术学报, 2014, 29(增刊1):50-58.
[20]	朱萌, 梅飞, 郑建勇, 等. 基于深度信念网络的高压断路器故障识别算法[J]. 电测与仪表, 2019, 56(2):10-15,46.
[21]	PAN Y, MEI F, MIAO H Y, et al. An approach for hvcb mechanical fault diagnosis based on a deep belief network and a transfer learning strategy[J]. Journal of Electrical Engineering & Technology, 2019, 14(1):407-419.
[22]	LECUN Y, BOSER B, DENKER J S, et al. Handwritten digit recognition with a back-propagation network[M]//Advances in Neural Information Processing Systems. San Francisco, CA: Morgan Kaufmann Publishers Inc., 1990:396-404.
[23]	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks[C]// Advances in neural information processing systems. 2012:1097-1105.
[24]	SMIRNOV E A, TIMOSHENKO D M, ANDRIANOV S N. Comparison of regularization methods for imagenet classification with deep convolutional neural networks[J]. Aasri Procedia, 2014, 6:89-94. doi: 10.1016/j.aasri.2014.05.013
[25]	吐松江·卡日, 逯浩坦, 高文胜, 等. 基于系统动力学的高压断路器全寿命成本评估[J]. 高压电器, 2023, 59(1):69-175,184.
[26]	YANG H M, ZHANG X Y, YIN F, et al. Robust classification with convolutional prototype learning[C]// Proceedings of the IEEE conference on computer vision and pattern recognition. 2018:3474-3482.
[27]	ARANGANAYAGI S, THANGAVEL K. Clustering categorical data using silhouette coefficient as a relocating measure[C]// International Conference On Computational Intelligence and Multimedia Applications (ICCIMA 2007).IEEE, 2007, 2:13-17.
[28]	叶昱媛, 沙浩源, 梁君涵, 等. 基于小波包能量的断路器操作机构缺陷诊断技术研究[J]. 电力工程技术, 2018, 37(4):71-77.
[29]	LU Y, Li Y. A novel fault diagnosis method for circuit breakers based on optimized affinity propagation clustering[J]. International Journal of Electrical Power & Energy Systems, 2020, 118:1-11.

故障类型	故障程度	类型标签
正常	—	F11
线圈老化	A	F21
	B	F22
	C	F23
铁芯卡涩	A	F31
铁芯卡涩	B	F32

网络层类型	输入→输出维度	参数
输入层	(1×800)
卷积池化层1	(1×800)→ (1×400×8)	卷积核维度:1×10 激活函数:ReLU 最大池化:1×2
卷积池化层2	(1×400×8)→ (1×200×1)	卷积核维度:1×5 激活函数:ReLU 最大池化:1×2
全连接层	(200×1)→(2)	激活函数:ReLU 损失函数:式(15) 原型损失函数

对比指标		故障类型	方法一 (DBN+TL)	方法二 (特征聚类法)	方法三 (本文方法)
轮廓系数	平均值	F11	0.94	0.94	0.97
		F21	0.97	0.86	0.97
		F22	0.81	0.75	0.97
		F23	0.63	0.72	0.98
		F31	0.94	0.95	0.98
		F32	0.98	0.93	0.98
	最小值	F11	0.86	0.90	0.95
		F21	0.90	0.73	0.93
		F22	0.61	0.37	0.95
		F23	-0.32	-0.63	0.95
		F31	0.91	0.89	0.94
		F32	0.94	0.87	0.96
识别正确率/%		F11	100	100	100
		F21	100	100	100
		F22	90.0	90.0	100
		F23	86.7	83.3	100
		F31	100	100	100
		F32	100	100	100
平均正确率/%			96.1	95.6	100
未知类识别正确率/%			—	90	100