作者：邵宇鹰¹, 王枭², 彭鹏¹, 袁国刚², 郑申辉¹

作者单位：1. 国网上海市电力公司，上海 200122;
2. 上海睿深电子科技有限公司，上海 201108

关键词：变压器;故障诊断;声信号;听觉感知;基底膜

摘要：

精准的变压器故障诊断方法，对于保证变压器可靠运行具有重要意义。该文提出一种基于变压器声信号并考虑人耳听觉感知的变压器故障诊断方法。首先，通过模拟变压器不同工况条件，采集变压器正常工况和不同故障条件下的噪声样本；然后，基于人耳传声集总参数模型，计算各个样本的基底膜位移响应的统计平均值(statistical mean value of basilar membrane displacement response，SMVBMDR)，并构建基于SMVBMDR的特征向量；最后，基于所构建的听觉特征向量，采用遗传算法优化的支持向量机对变压器状态进行识别，并通过测试样本验证所提出方法的有效性。结果显示，以SMVBDMDR为特征向量的故障诊断平均准确率高达98.1%。以上研究表明，基于变压器声信号并考虑人耳听觉感知特性的变压器故障诊断方法可以有效地应用于变压器故障诊断和监测中。

Research on transformer fault diagnosis method based on auditory perception of sound field
SHAO Yuying¹, WANG Xiao², PENG Peng¹, YUAN Guogang², ZHENG Shenhui¹
1. State Grid Shanghai Municipal Electric Power Company, Shanghai 200122, China;
2. Shanghai Rhythm Electronic Technology Co., Ltd., Shanghai 201108, China
Abstract: Accurate transformer fault diagnosis method is of great significance to ensure the reliable operation of the transformer. This paper presents a transformer fault diagnosis method based on the transformer acoustic signal and the human auditory perception. Firstly, by simulating the different working conditions of the transformer, the noise samples under the normal working conditions and different fault conditions of the transformer were collected. Then, the statistical mean value of basilar membrane displacement response (SMVBMDR) of each sample was calculated in a lumped-parameter model of the human ear, and thereby a SMVBMDR-based feature matrix was established. Finally, based on the constructed auditory feature vector, the state of the transformer was identified by genetic algorithm optimization support vector machine, and the effectiveness of the proposed method was verified by test samples. The results showed that the average accuracy of the fault diagnosis method taking SMVBMDR as feature vector was 98.1%. The above research shows that the transformer fault diagnosis method based on the transformer acoustic signal and considering the characteristics of the human auditory perception can be effectively applied to the transformer fault diagnosis and monitoring.
Keywords: transformer;fault diagnosis;acoustic signal;auditory perception;basilar membrane
2021, 47(3):92-97  收稿日期: 2020-06-04;收到修改稿日期: 2020-07-08
基金项目: 国网上海市电力公司科技项目资助（B30970190002）
作者简介: 邵宇鹰(1977-),男,江苏常州市人,高级工程师,博士,主要从事电力设备状态监测和新能源技术研究与应用
参考文献
[1] 张重远, 罗世豪, 岳浩天, 等. 基于Mel时频谱-卷积神经网络的变压器铁芯声纹模式识别方法[J]. 高电压技术, 2020, 46(2): 413-422
[2] 马叶芝, 焦彦军, 王东升, 等. 变压器故障诊断中溶解气体的PSO-WFCM算法研究[J]. 高压电器, 2014, 50(1): 72-76
[3] 张镱议, 彭鸿博, 李昕, 等. 基于DGA特征量优选与改进磷虾群算法优化支持向量机的变压器故障诊断模型[J]. 电测与仪表, 2018, 56(21): 110-116
[4] 颜秋容, 刘欣, 尹建国. 基于小波理论的电力变压器振动信号特征研究[J]. 高电压技术, 2007, 33(1): 165-168
[5] 魏晓莹, 宋仕江, 郭谋发, 等. 基于振动信号二维特征向量的配变铁心故障诊断[J]. 电气技术, 2016, 17(1): 16-15
[6] 陶志国. 小型电力配电变压器空载条件下声功率级的测定试验[J]. 中国新技术新产品, 2010, 17(20): 158-160
[7] BRUMMUND M K, SGARD F, PETIT Y, et al. Three-dimensional finite element modeling of the human external ear: Simulation study of the bone conduction occlusion effecta[J]. Journal of the Acoustical Society of America, 2014, 135(3): 1433-1444
[8] LIU Y, NEELY S T. Distortion product emissions from a cochlear model with nonlinear mechanoelectrical transduction in outer hair cells[J]. Journal of the Acoustical Society of America, 2010, 127(4): 2420-2432
[9] BRUCE I C, SACHS M B, YOUNG E D. An auditory-periphery model of the effects of acoustic trauma on auditory nerve responses[J]. Journal of the Acoustical Society of America, 2003, 113(1): 369-378
[10] 高延峰, 王齐胜, 黄林然, 等. 基于人耳听觉模型的MIG焊熔滴过渡状态识别[J]. 机械工程学报, 2019, 55(17): 68-76
[11] XU Z, XIA X J, LAI S Y, et al. Improvement of interior sound quality for passenger car based on optimization of sound pressure distribution in low frequency[J]. Applied Acoustics, 2018, 130(1): 43-51
[12] 张天宇, 鲁义, 施式亮, 等. 基于支持向量机分类算法的多煤种煤自燃危险性预测[J]. 湖南科技大学学报(自然科学版), 2019, 34(2): 11-17
[13] 戴栋, 黄筱婷, 代州, 等. 基于支持向量机的输电线路覆冰回归模型[J]. 高电压技术, 2013, 39(11): 2822-2828