针对目前对于微型电机故障检测的研究较少及基于单域特征的传统电机诊断方法精度低等问题，提出一种基于集成经验模态分解(ensemble empirical mode decomposition ,EEMD)及蝙蝠算法(bat algorithm ,BA)优化核极限学习机(kernel based extreme learning machine，KELM)的微型电机故障检测方法。所提方法包括样本集构造、模型训练及参数优化和模型测试三个步骤：首先，对采集到的微型电机信号进行EEMD处理并依据相关系数原则筛选出主要的本征模态分量(intrinsic mode fuction,IMF)，结合计算得到的电机信号的时域、频域特征构造多域特征集并进行归一化处理，按一定比例将样本集划分训练集和测试集；其次，输入训练集，以错误率为适应度，并采用蝙蝠算法对KELM模型进行参数优化；最后，输入测试集对优化的BA-KELM模型进行测试，并与其他模型进行对比。试验结果表明，所提方法的准确率达98.75%，高于其余方法。
关键字：微型电机；故障诊断；蝙蝠算法；极限向量机；核极限向量机
Abstract
At present, there are few researches on micro-motor fault detection, and the traditional motor diagnosis methods based on single domain features have low accuracy. So, a fault detection method for micro motor based on ensemble empirical mode decomposition (EEMD) and bat algorithm (BA) is proposed. The proposed method includes three steps: construction of sample sets, model training as well as parameters optimization, and model test. Firstly, EEMD processing is carried out to the collected micro motor signals, and the main intrinsic mode fuction(IMF) components are selected by the principle of correlation coefficient, combined with the calculated time and frequency domain features of the motor signals, multi-domain feature set is constructed and normalized. Then, these features are divided into a training set and a test set in a certain proportion. Secondly, the training set is taken as input, the error rate is employed as fitness, and then parameters of KELM model are optimized with bat algorithm (BA). Finally, the optimized BA-KELM model is tested by test set, the experimental results show that the accuracy of the proposed method is 98.75%, which is higher than other methods.
关键词
微型电机 /
故障诊断 /
蝙蝠算法 /
极限向量机 /
核极限向量机
{{custom_keyword}} /
Key words
micro motor；fault diagnosis；bat algorithm /
extreme learning machine (ELM) /
Kernel Based Extreme Learning Ma chine (KELM)
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 黄琳娣,黄大绪.国内外微电机市场概述[J].微电机,2007(07):75-78.
Huang Lindi, Huang Daxu. Developing trends and market demands of micromotors [J]. Micromotors, 2007(07):75-78.
[2] 王博磊,曹伟,王涛,等.基于EEMD能量特征识别的电动机故障诊断技术研究[J].电机技术,2020(05):23-25+30.
Wang Bolei, Cao Wei, Wang Tao, et al. Research on motor fault diagnosis technology based on the energy feature recognition of EEMD [J]. Electrical Machinery Technology , 2020(05):23-25+30.
[3] 蔡少辉.基于振动信号分析的电机故障诊断应用研究[J].电子测试,2021(06):95-96.
Cai Shaohui. Application Research of motor fault diagnosis based on vibration signal analysis [J]. Electronic Test, 2021(06):95-96.
[4] 李俊卿,朱锦山,沈亮印,等.SVD滤波与改进小生境遗传算法在双馈异步电机转子匝间短路故障量提取中的应用[J].华北电力大学学报(自然科学版),2017,44(02):42-46+54.
LI Junqing, Zhu Jinshan, Shen Liangyin, et al. Application of SVD filter and improved niche genetic algorithm in doubly-fed induction generator rotor Inter-turn short circuit fault analysis [J]. Journal of North China Electric Power University(Natural Science Edition), 2017,44(02):42-46+54.
[5] 何玉灵,王珂,仲昊,等.基于最大相关峭度解卷积算法的发电机特征振动信号增强检测[J].华北电力大学学报(自然科学版),2017,44(03):67-73+89.
He Yulin, Wang Ke, Zhong Hao, et al. Enhanced detection of generator’s characteristic vibration signal based on maximum correlated kurtosis deconvolution [J]. Journal of North China Electric Power University(Natural Science Edition), 2017,44(02):42-46+54.
[6] 李俊卿,李忠徽,仝宗义.基于支持向量机和D-S证据理论的双馈风机定子匝间短路故障诊断[J].电机与控制应用,2018,45(05):99-103+110.
LI Junqing, Li Zhonghui, Tong Zongyi. Fault diagnosis of statorinter-turn short-circuit in DFIG based on support vector machine ＆ D-S evidence theory [J]. Electric Machines & Control Application, 2018,45(05):99-103+110.
[7] Yan Y, Liu Q, Gao X Q, et al. Motor fault diagnosis algorithm based on wavelet and attention mechanism[J]. Journal of Sensors,2021,2021:1-9.
[8] 王栋悦,谷怀广,魏书荣,等.基于机电信号融合的DFIG定子绕组匝间短路故障诊断[J].电力系统自动化,2020,44(09):171-178.
Wang Yuedong, Gu Huaiguang, Wei Shurong, et al. Diagnosis of inter-turn short-circuit fault in stator windings of DFIG based on mechanical and electrical signal fusion [J]. Automation of Electric Power Systems, 2020,44(09):171-178.
[9] 范万里.基于小波分析与BP神经网络的机车牵引电机故障诊断[J].内燃机与配件,2020(09):149-150.
Fan Wanli. The fault diagnosis of locomotive traction motor based on wavelet analysis and artificial neutral network [J]. Internal Combustion Engine & Parts, 2020(09):149-150.
[10] 李强,车文龙.基于改进粒子群优化神经网络的电机故障诊断[J].电气传动,2020,50(01):103-108.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
PDF(1686 KB)
