基于KPCA-LSSVM的单向阀故障诊断研究

doi:10.16180/j.cnki.issn1007-7820.2019.03.003

摘要/Abstract

摘要：

针对高压隔膜泵单向阀的故障振动信号特征难以提取及诊断的问题,文中采用KPCA和LSSVM相结合的方法进行故障诊断研究。对单向阀各状态信号运用双稳SR方法和DEMD算法进行信号预处理,并利用K-L散度选择分解后的主分量进行时频域特征参数的提取以构建特征向量集。运用KPCA对向量集进行二次特征提取,并将提取的特征向量输入到LSSVM诊断系统中,以完成单向阀故障诊断及分类。经实验验证,该方法的故障诊断率可达到90%,能够较好的诊断出单向阀故障特征。

关键词: 单向阀, DEMD, 特征向量, KPCA, LSSVM, 故障诊断

Abstract:

It was difficult to extract and diagnose the characteristic of the fault vibration signal of the check valve of the high pressure diaphragm pump. To solve the problem, KPCA combined with LSSVM were performed for fault diagnosis research. The bi-stable SR method and DEMD algorithm were used to preprocess the signal of each state of the one-way valve, and the principal component of the decomposed K-L divergence was used to extract feature parameters of the time-frequency domain to construct the feature vector set. The second feature extraction of vector set was carried out by KPCA, which were further input into the LSSVM diagnostic system to complete check value fault diagnosis and classification. The experimental data proved that the fault diagnosis rate of this method was 90%, and the fault feature of the one-way valve could be diagnosed well.

Key words: check value, DEMD, feature vector, KPCA, LSSVM, fault diagnosis

中图分类号:

TN911.7

牟竹青. 基于KPCA-LSSVM的单向阀故障诊断研究[J]. 电子科技, 2019, 32(3): 10-15.

MU Zhuqing. Study on the Fault Diagnosis of Check Valve Based on KPCA-LSSVM[J]. Electronic Science and Technology, 2019, 32(3): 10-15.

图/表 9

图1

图2

图3

图4

表1

图5

图6

图7

表2

参考文献 17

[1]	Wu S D, Wu P H, Wu C W , et al. Bearing fault diagnosis based on multiscale permutation entropy and support vector machine[J]. Entropy, 2012,14(8):1343-1356. doi: 10.3390/e14081343
[2]	桑琦, 陈浩, 李卫华 . 基于小波变换的图像融合方法仿真[J]. 电子科技, 2017,30(3):1-3,7.
	Sang Qi, Chen Hao, Li Weihua . Image fusion method simulation based on wavelet transform[J]. Electronic Science and Technology, 2017,30(3):1-3,7.
[3]	李文峰, 戴豪民, 许爱强 . 时域新指标和PNN在滚动轴承故障诊断中的应用[J]. 机械科学与技术, 2016,35(9):1382-1386.
	Li Wenfeng, Dai Haomin, Xu Aiqiang . New Time domain index and probabilistic neural network and their application in fault diagnosis of rolling bearing[J]. Mechanical Science and Technology for Aerospace Engineering, 2016,35(9):1382-1386.
[4]	徐卫晓, 宋平, 谭维文 . 基于KPCA-BP网络模型的滚动轴承故障诊断方法研究[J]. 煤矿机械, 2014,25(8):265-267.
	Xu Weixiao, Song Ping, Tan Weiwen , et al. Fault diagnosis methods of rolling bearing based on kernel principal component analysis and BP neural network[J]. Coal Mine Machinery, 2014,25(8):265-267.
[5]	叶远芹, 原思聪, 魏笑笑 , 等. 基于混沌遗传算法的计算机辅助动态布局[J]. 计算机工程与设计, 2017,38(9):2562-2566.
	Ye Yuanqin, Yuan Sicong, Wei Xiaoxiao , et al. Computer-aided dynamic facility layout based on chaos genetic algorithm[J]. Computer Engineering and Design, 2017,38(9):2562-2566.
[6]	李怡弘, 裘炅 . 粒子群蚁群融合算法的火灾救援路径研究[J]. 电子科技, 2018,31(1):58-62.
	Li Yihong, Qiu Jiong . The particle swarm optimization algorithm merged with ant colony optimization algorithm of fire rescue way research[J]. Electronic Science and Technology, 2018,31(1):58-62.
[7]	黄潮 . 云计算环境下的海量光纤通信故障数据挖掘算法研究[J]. 激光杂志, 2017,38(1):96-100.
	Huang Chao . Research on data mining algorithm of massive optical fiber communication in cloud computing environment[J]. Laser Journal, 2017,38(1):96-100.
[8]	Benzi R, Sutera A, Vul Piana A . The mechanism of stochastic resonance[J].Journal of Physics A, 1981(14):L453-L457.
[9]	焦尚彬, 李鹏华, 张青 , 等. 采用知识的粒子群算法的多频微弱信号自适应随机共振检测方法[J]. 机械工程, 2014,50(12):1-10.
	Jiao Shangbin, Li Penghua ,Zhang qing, et al. Multi-frequency weak signal detection method based on adaptive stochastic resonance with knowledge-based PSO[J]. Journal of Mechanical Engineering, 2014,50(12):1-10.
[10]	冷永刚, 赖志慧, 范胜波 , 等. 二维Duffing振子的大参数随机共振及微弱信号检测研究[J]. 物理学报, 2012,61(23):71-80.
	Leng Yonggang, Lai Zhihui, Fan Shengbo , et al. Large parameter stochastic resonance of two-dimensional Duffing oscillator and its application on weak signal detection[J]. Acta Physica Sinica, 2012,61(23):71-80.
[11]	张东, 李金海, 欧松林 , 等. 一种适用于GPS软件信号源的信噪比模拟算法[J]. 微电子学与计算机, 2017,34(3):10-14.
	Zhang Dong, Li Jinhai, Ou Songlin , et al. A signal to noise ratio simulation algorithm applied to gps software signal source[J]. Microelectronics & Computer, 2017,34(3):10-14.
[12]	孟宗, 季艳 . 基于DEMD和对称差分能量算子解调的滚动轴承故障诊断[J]. 中国机械工程, 2015,26(12):1658-1664.
	Meng Zong, Ji Yan . Fault diagnosis of rolling bearings based on DEMD and symmetric difference energy operator demodulation[J].China Mechanical Engineering2015, 26(12):1658-1664.
[13]	刘海波, 玄志武 . 利用微分算子增强EMD算法频带分解能力[J]. 振动与冲击, 2013,32(18):133-135,145.
	Liu Haibo, Xuan Zhiwu . Improving frequency-band separating ability of EMD with a differential operator[J]. Journal of Vibration and Shock, 2013,49(8):10-18.
[14]	赵永威, 周苑, 李弼程 , 等. 基于近义词自适应软分配和卡方模型的图像目标分类方法[J]. 电子学报, 2016,44(9):2181-2188.
	Zhao Yongwei, Zhou Yuan, Li Bicheng . Image object classification method with ho moionym based adaptive soft-assignment and chi-square model[J]. Acta Electronica Sinica, 2016,44(9):2181-2188.
[15]	钱琨 . 基于组合KPCA与改进ELM的工业过程故障诊断研究[D]. 重庆:重庆大学, 2016.
	Qian Kun . Research on industrial process fault diagnosis based on combined KPCA and improved ELM[D]. Chongqing:College of Automation of Chongqing University, 2016.
[16]	王辞 . 基于LCD和PSO-LSSVM的城轨列车滚动轴承故障诊断研究[D]. 北京:北京交通大学, 2016.
	Wang Ci . Research on fault diagnosis of urban rail train rolling bearing based on LCD and PSO-LSSVM[D]. Beijing:Beijing Jiaotong University, 2016.
[17]	赵海君 . 语音识别的SVM模型选择分析[J].电脑知识与技术, 2015(25):133-134.
	Zhao Haijun . Speaker recognition model selection analysis based on SVM[J]. Computer Knowledge and Technology, 2015(25):133-134.

样本类别描述	样本规模	训练样本数	类别标签
CN	3 072×20	3 072×20	1
CZ	3 072×20	3 072×20	2
CM	3 072×20	3 072×20	3
CK	3 072×20	3 072×20	4

算法种类	训练样本	故障错分个数				故障准确率
算法种类	训练样本	CN	CZ	CM	CK	故障准确率
KPCA-LSSVM	20	0	4	3	1	90.00%
PCA-SVM	20	2	4	4	1	86.25%