作者：杜伟, 房立清, 吕岩, 齐子元
作者单位：军械工程学院火炮工程系, 河北 石家庄 050003
关键词：故障诊断;多尺度关联维数;流形学习;自动机
摘要：
针对自动机振动信号非平稳、非线性的特点，提出基于多尺度关联维数和线性局部切空间排列（linear local tangent space alignment，LLTSA）相结合的自动机故障诊断方法。首先，利用局部特征尺度分解（local characteristic-scale decomposition，LCD）将自动机振动信号分解为不同尺度下的内禀尺度分量（intrinsic scale component），提取出反映状态信息的主要分量并计算各分量的关联维数。然后，利用线性局部切空间排列算法挖掘出可区分度更高的特征子集。最后，将得到的低维特征输入支持向量机进行识别，自动机故障诊断实验表明，所提方法具备较高的诊断准确率。此外，将LCD与经验模态分解（empirical mode decomposition，EMD）和局部均值分解（local mean decomposition，LMD）方法的诊断结果进行比较，验证所提方法的优势。
Fault diagnosis of automaton based on multiscale correlation dimension and manifold learning
DU Wei, FANG Liqing, LÜ Yan, QI Ziyuan
Department of Artillery Engineering, Ordnance Engineering College, Shijiazhuang 050003, China
Abstract: Aiming at the non-stationary and non-linear characteristics of automaton vibration signal,a fault diagnosis method of automaton based on multiscale correlation dimension and linear local tangent space alignment (LLTSA) was proposed.Firstly,the vibration signal of automaton was decomposed with local characteristic-scale decomposition (LCD) to obtain intrinsic scale components in different scales,and the correlation dimension of each principal component reflected the state information was calculated.Then,the mining performance of the feature subset with higher distinguishability was further implemented by using linear local tangent space alignment.Finally,low-dimensional feature was put into SVM to recognize the state types.The results of automaton fault test indicate that the proposed method is of high accuracy.In addition,the diagnostic results calculated with LCD,empirical mode decomposition and local mean decomposition were compared,verifying the advantage of the method.
Keywords: fault diagnosis;multiscale correlation dimension;manifold learning;automaton
2017, 43(10): 102-108 收稿日期: 2017-01-12;收到修改稿日期: 2017-02-08
基金项目: 河北省自然科学基金资助项目（E2016506003）
作者简介: 杜伟(1992-),男,山东临沂市人,硕士研究生,专业方向为兵器性能检测与故障诊断。
参考文献
[1] 潘宏侠,都衡,马春茂. 局域波信息熵在高速自动机故障诊断中的应用[J]. 振动、测试与诊断,2015(6):1159-1164,1205.
[2] 戴涌,张国平,王茂林,等. 某转膛自动机异常发射故障分析[J]. 火炮发射与控制学报,2013(3):67-71.
[3] 兰海龙,潘宏侠,龚明. 局域波分形方法在柴油机故障诊断中的应用[J]. 机床与液压,2013(5):180-183.
[4] ZHANG Z, WU J, MA J, et al. Fault diagnosis for rolling bearing based on lifting wavelet and morphological fractal dimension[C]//Control and Decision Conference. IEEE,2015.
[5] ARGYRIS J, ANDREADIS I, PAVLOS G, et al. The influence of noise on the correlation dimension of chaotic attractors[J]. Chaos Solitons and Fractals,1998,9(3):343-361.
[6] 杨宇,王欢欢,喻镇涛,等. 基于ITD改进算法和关联维数的转子故障诊断方法[J]. 振动与冲击,2012(23):67-70,76.
[7] 李琳,张永祥,明廷涛. EMD降噪的关联维数在齿轮故障诊断中的应用研究[J]. 振动与冲击,2009,28(4):145-148.
[8] HUANG N E, SHEN Z, LONG S R. A new view of nonlinear water waves:the Hilbert spectrum[J]. Annual Review of Fluid Mechanics,1999(31):417-457.
[9] 程军圣,郑近德,杨宇. 一种新的非平稳信号分析方法:局部特征尺度分解法[J]. 振动工程学报,2012(2):215-220.
[10] 程军圣,杨怡,杨宇. 局部特征尺度分解方法及其在齿轮故障诊断中的应用[J]. 机械工程学报,2012(9):64-71.
[11] SMITH J S. The local mean decomposition and its application to EEG perception data[J]. Journal of the Royal Society Interface,2005,2(5):443-454.
[12] LEI Y G, HE Z J, ZI Y Y. Application of the EEMD method to rot or fault diagnosis of rotating machinery[J]. Mechanical Systems and Signal Processing,2009,23(4):1327-1338.
[13] 李锋,汤宝平,陈法法. 基于线性局部切空间排列维数化简的故障诊断[J]. 振动与冲击,2012(13):36-40.
[14] 张良,张前图. 基于LCD模糊熵和流行学习的故障特征提取方法[J]. 机械强度,2016(2):225-230.
[15] GRASSBERGER P, PROCACCIA I. Characterization of strange attractors[J]. Physical Review Letters,1983,50(5):346-349.
[16] 李奎为,胡瑾秋,张来斌,等. 关联维数无标度区确定方法及诊断应用[J]. 石油机械,2007(4):43-45.
[17] PENG J E, QIAO H, XU Z B. A new approach to stability of neural networks with time-varying delays[J]. Neural Networks,2002,15(1):95-103.
[18] CAO L Y. Practical method for determining the minimum embedding dimension of a scalar time series[J]. Physica. Section D:Nonlinear Phenomena,1997,110(1/2):43-50.
