刀具磨损特征参数提取与状态识别方法
|
摘要
为提高刀具磨损状态识别准确率,文章提出了S变换时频图纹理特征参数提取方法和基于隐马尔科夫模型的磨损状态识别方法。以声发射信号为敏感信号,设计了刀具磨损实验方案;基于EEMD算法,提出了互相关系数与鞘度相结合的综合降噪方法;使用S变换处理声发射信号得到等高线灰度图,通过灰度共生矩阵提取等高线灰度图的纹理特征参数;将类内散布矩阵和类间散布矩阵结合,提出了基于散布矩阵的特征参数敏感度分析和降维方法;采用基于隐马尔科夫模型的磨损状态识别方法,分别将全维特征参数和降维特征参数用于磨损状态识别,实验结果表明,全维特征参数的识别准确率为88.34%,降维特征参数的识别准确率为100%。
To improve recognition accuracy of cutting tool wear state, s-transform time-frequency graph texture feature parameters extraction method and wear state recognition method based on hidden Markov model are proposed. Acoustic emission signal as sensitive signal, cutting tool wear experimental scheme is designed. based on EEMD algorithm, comprehensive denoise method combined cross correlation coefficients and sheath degree is put forward. Contour line grey-scale image of acoustic emission signal is gotten by S-transform, and textual feature parameters of contour line grey-scale image is calculated by grey-level co-occurrence matrix. Intra-class scattering matrix and inter-class scattering matrix are combined, so that characteristic parameters sensitivity analysis method and dimensionality reduction method is raised. Hidden Markov model is employed to recognize wear state. Full-dimensional characteristic parameters and reduced--dimensional characteristic parameters are used to recognize wear state, and the experiment shows that accuracy rate by full-dimensional characteristic parameters is 88.34%, and 100% by reduced-dimensional characteristic parameters.
To improve recognition accuracy of cutting tool wear state, s-transform time-frequency graph texture feature parameters extraction method and wear state recognition method based on hidden Markov model are proposed. Acoustic emission signal as sensitive signal, cutting tool wear experimental scheme is designed. based on EEMD algorithm, comprehensive denoise method combined cross correlation coefficients and sheath degree is put forward. Contour line grey-scale image of acoustic emission signal is gotten by S-transform, and textual feature parameters of contour line grey-scale image is calculated by grey-level co-occurrence matrix. Intra-class scattering matrix and inter-class scattering matrix are combined, so that characteristic parameters sensitivity analysis method and dimensionality reduction method is raised. Hidden Markov model is employed to recognize wear state. Full-dimensional characteristic parameters and reduced--dimensional characteristic parameters are used to recognize wear state, and the experiment shows that accuracy rate by full-dimensional characteristic parameters is 88.34%, and 100% by reduced-dimensional characteristic parameters.
引文
[1] Nouri M,Fussell B K,Ziniti B L,et al.Real-time tool wear monitoring in milling using a cutting condition independent method[J].International Journal of Machine Tools & Manufacture,2015,89:1-13.
[2] Rmili W,Ouahabi A,Serra R,et al.An automatic system based on vibratory analysis for cutting tool wear monitoring[J].Measurement,2016,77:117-123.
[3] 祝和意,李宏波,张合沛,等.基于改进CRITIC法的TBM刀具声发射信号研究[J].振动与冲击,2016,35(6):197-202.
[4] 彭爬.基于混沌理论的刀具磨损识别研究[D].吉林:东北电力大学,2016.
[5] 张锴锋,袁惠群,聂鹏.基于广义分形维数的刀具磨损状态监测[J].振动与冲击,2014,33(1):162-164.
[6] 王利伟.基于云理论和支持向量机的数控机床刀具状态诊断技术研究[D].大连:大连交通大学,2012.
[7] 谢楠,马飞,段明雷,等.基于主成分分析与C-支持向量机的刀具磨损状态监测[J].同济大学学报(自然科学版),2016,44(3):434-439.
[8] 陆启宇,李腾飞,王备,等.基于自适应EEMD算法的变压器绕组状态检测仿真[J].计算机仿真,2017,34(1):453-457.
[9] Haraclick R M.Texture Features for Image Classification[J].IEEE Trans Smc,1990,3(6):610-621.
[10] 唐发明,王仲东,陈绵云.一种新的二叉树多类支持向量机算法[J].计算机工程与应用,2005,41(7):24-26.
[2] Rmili W,Ouahabi A,Serra R,et al.An automatic system based on vibratory analysis for cutting tool wear monitoring[J].Measurement,2016,77:117-123.
[3] 祝和意,李宏波,张合沛,等.基于改进CRITIC法的TBM刀具声发射信号研究[J].振动与冲击,2016,35(6):197-202.
[4] 彭爬.基于混沌理论的刀具磨损识别研究[D].吉林:东北电力大学,2016.
[5] 张锴锋,袁惠群,聂鹏.基于广义分形维数的刀具磨损状态监测[J].振动与冲击,2014,33(1):162-164.
[6] 王利伟.基于云理论和支持向量机的数控机床刀具状态诊断技术研究[D].大连:大连交通大学,2012.
[7] 谢楠,马飞,段明雷,等.基于主成分分析与C-支持向量机的刀具磨损状态监测[J].同济大学学报(自然科学版),2016,44(3):434-439.
[8] 陆启宇,李腾飞,王备,等.基于自适应EEMD算法的变压器绕组状态检测仿真[J].计算机仿真,2017,34(1):453-457.
[9] Haraclick R M.Texture Features for Image Classification[J].IEEE Trans Smc,1990,3(6):610-621.
[10] 唐发明,王仲东,陈绵云.一种新的二叉树多类支持向量机算法[J].计算机工程与应用,2005,41(7):24-26.