面向转子故障特征提取的多尺度拉普拉斯特征映射方法

摘要/Abstract

摘要： 融合多尺度分解理论和流形学习思想，提出了一种面向转子故障特征提取的多尺度拉普拉斯特征映射算法。首先对转子故障振动信号进行多尺度小波包分解，提取各独立频带信号的最优尺度小波熵，构建特征参量矩阵并估计其固有维数，然后通过拉普拉斯特征映射将特征参量数据嵌入到低维本征空间，得到故障的最敏感特征，最后融合决策实现故障的准确识别。实验表明，相对于主成分分析算法、局部线性嵌入算法和拉普拉斯特征映射算法，多尺度拉普拉斯特征映射方法提取的转子故障信号特征更容易识别。

关键词: 转子系统, 拉普拉斯特征映射, 多尺度, 特征提取

Abstract: Based on theory of multi-scale decomposition and manifold learning thought, a multi-scale Laplasse feature map algorithm for fault feature extraction was proposed. Firstly, the multi-scale wavelet packet decomposition of the rotor fault vibration signals was carried out. The optimal scale wavelet entropy of each independent frequency band signals was extracted, and the characteristic parameter matrix was constructed and the intrinsic dimension was estimated. Then the characteristic parameters of data were embedded into a low dimensional eigenspace by Laplasse feature mapping to get the most sensitive feature of faults. Lastly, the accurate identification of faults was realized by the fusion decision. Experiments show that, compared with the principal component analysis, local linear embedding and Laplacian eigenmap algorithm, rotor fault feature signal extraction of multi-scale Laplasse feature mapping method is more easily identify.

Key words: rotor system, Laplacian eigenmap, multi-scale, feature extraction

中图分类号:

TP277

王广斌, 杜晓阳, 罗军, . 面向转子故障特征提取的多尺度拉普拉斯特征映射方法[J]. 中国机械工程.

Wang Guangbin, Du Xiaoyang, Luo Jun, . Multi-scale Laplace Feature Mapping for Rotor Fault Feature Extraction[J]. China Mechanical Engineering.

参考文献

[1]Randall R B. The Relationship between Spectral Correlation and Envelope Analysis in the Diagnostics of Bearing Faults and Other Cyclostationary Machine Signals [J]. Mechanical System and Signal Processing,2002, 15(5): 945-962.

[2]Ough J R. Development and Analysis of Time Variant Discrete Fourier Transform Order Tracking[J]. Mechanical System and Signal Processing, 2003, 17(6): 185-199.

[3]李晓虎, 贾民平，许飞云. 频谱分析法在齿轮故障诊断中的应用[J]. 振动、测试与诊断, 2003, 23(3): 168 -170.

Li Xiaohu, Jia Minping, Xu Feiyun.Virtual Instruments Based Remote Equipment Diagnosis[J].Journal of Vibration, Measurement & Diagnosis, 2003, 23(3): 168-170.

[4]Seung H S, Daniel D L. The Manifold Ways of Perception[J].Science(S0036-8075),2000,290(5500):2268-2269.

[5]Roweis S, Saul L.Nonlinear Dimensionality Reduction by Locally Linear Embedding[J].Science(S0036-8075),2000,290(5500):2323-2326.

[6]Belkin M, Niyogi P. Laplacian Eigenmaps for Dimensionality Reduction and Data Representation[J].Neural Computation， 2003, 15(6):1373-1396.

[7]Belkin M, Niyogi P. Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering[J]. Advances in Neural Information Processing Systems, 2002，14（6）：585-591.

[8]Zhang Zhenyue, Zha Hongyuan. Principal Maino-flds and Nonlinear Dimension Reduction via Tangent Space Alignment[J]. SIAM Joural on Scinetific Computeing,2005,26 (1):313-338.

[9]刘继伟. 基于大数据的多尺度状态监测方法及应用[D]. 北京: 华北电力大学, 2013.

[10]李鹏. 多尺度分析方法在旋转机械状态监测中的应用研究[D]. 合肥: 中国科学技术大学, 2012.

[11]张龙, 张磊, 熊国良. 基于多尺度熵和神经网络的滚动轴承故障诊断[J]. 机械设计与研究, 2014,30(5): 96-99.

Zhang Long, Zhang Lei, Xiong Guoliang. Based on the Multi-scale Entropy of the Neural Network and the Rolling Bearing Fault Diagnosis[J]. Mechanical Design and Research, 2014,30(5): 96-99.

[12]Sinha J K, Lees A W, Friswell M I. Estimating Unbalance and Misalignment of a Flexible Rotating Machine from a Single Run-down[J]. Journal of Sound and Vibration, 2004,30(9): 967-989.

[13]张靖, 闻邦椿. 两端支座松动转子系统的频率特性分析[J]. 中国机械工程, 2008, 19(1): 68-71.

Zhang Jing,Wen Bangchun. On Both Ends of the Pedestal Looseness Frequency Characteristic Analysis of Rotor System[J].China Mechanical Engineering, 2008, 19(1): 68-71.

[1]	李胜远, 郑龙席. 磁轴承激励下转子系统动力学特性[J]. 中国机械工程, 2021, 32(08): 901-907.
[2]	许鹏, 方舟, 王平, 耿明, 许勇. 基于脉冲漏磁暂态特征的缺陷量化评估方法[J]. 中国机械工程, 2021, 32(07): 860-866,881.
[3]	江志农1;王子嘉1;张进杰2;黄翼飞3;茆志伟2. 基于能量算子梯度邻域特征提取的核电应急柴油发电机组故障诊断方法[J]. 中国机械工程, 2021, 32(05): 617-623.
[4]	丁嘉鑫;王振亚;姚立纲;蔡永武. 广义复合多尺度加权排列熵与参数优化支持向量机的滚动轴承故障诊断[J]. 中国机械工程, 2021, 32(02): 147-155.
[5]	唐贵基;李楠楠;王晓龙. 综合改进奇异谱分解和奇异值分解的齿轮故障特征提取方法[J]. 中国机械工程, 2020, 31(24): 2988-2996.
[6]	万书亭;彭勃;王晓龙. 基于双时域变换和稀疏编码收缩的滚动轴承早期故障诊断方法[J]. 中国机械工程, 2020, 31(23): 2829-2836.
[7]	詹梅, 雷煜东, 郑泽邦, . [成形过程仿真优化与集成计算材料工程]集成计算材料工程在精确塑性成形中的应用现状与发展趋势[J]. 中国机械工程, 2020, 31(22): 2663-2677.
[8]	刘志强, 赵杰, 王克环, 武永, 吕良星, 刘钢, 苑世剑, . [成形过程仿真优化与集成计算材料工程]钛合金热成形工艺形变与组织演变耦合多尺度仿真研究进展[J]. 中国机械工程, 2020, 31(22): 2678-2690,2698.
[9]	王振亚;姚立纲. 广义精细复合多尺度样本熵与流形学习相结合的滚动轴承故障诊断方法[J]. 中国机械工程, 2020, 31(20): 2463-2471.
[10]	何彦1;凌俊杰1;王禹林2;李育锋1;吴鹏程1;肖圳1. 基于长短时记忆卷积神经网络的刀具磨损在线监测模型[J]. 中国机械工程, 2020, 31(16): 1959-1967.
[11]	余石龙, 鲍劲松, 李婕, 张启华. 基于误差注意力的晶圆制造数据异常检测[J]. 中国机械工程, 2020, 31(14): 1686-1692.
[12]	姜少飞;邬天骥;彭翔;李吉泉;李治;孙涛. 基于XGBoost特征提取的数据驱动故障诊断方法[J]. 中国机械工程, 2020, 31(10): 1232-1239.
[13]	王志永;杜伟涛. 螺旋锥齿轮铣齿机主轴\\|轴承系统的动力学分析[J]. 中国机械工程, 2019, 30(18): 2211-2216,2223.
[14]	刘朋;任工昌;杨力鹏;胡小龙. 基于斜率差的激光雷达环境特征提取方法[J]. 中国机械工程, 2019, 30(14): 1679-1683.
[15]	李从志;郑近德;潘海洋;刘庆运. 基于精细复合多尺度散布熵与支持向量机的滚动轴承故障诊断方法[J]. 中国机械工程, 2019, 30(14): 1713-1719,1726.