基于SVD降噪法对齿轮点蚀故障特征的试验及分析
|
摘要
根据奇异值分解法(SVD)的基本原理,采用数值模拟分析法验证了SVD法在信号处理上具有良好的降噪效果。再利用该法对齿面点蚀故障振动试验数据进行了奇异值和奇异值差分谱计算,得出了真实信号与噪音信号的相空间及真实信号的重构阶数,并对真实信号进行重构,得出了经降噪后真实振动情况的时域信号。在此基础上,将真实振动加速度信号进行FFT变换,得出了齿轮箱上各测点处的频响特性;最后,分别研究了转速、载荷与齿轮箱振动频响特性之间的变化关系,得出了输出轴的振动特性对转速或载荷的波动情况较为敏感。
According to the basic principle of singular value decomposition(SVD), the SVD has good noise reduction effect is verified in signal processing by the numerical simulation method. Using this method to calculate the singular value and singular value difference spectrum of tooth surface point erosion fault vibration test data, the phase space of real signal and noise signal and the reconstruction order of real signal are obtained, and then the real signal is reconstructed. The time domain signal of real vibration after noise reduction is acquired. On the basis of this, FFT transformation of the real vibration acceleration signal is carried out, and the frequency response characteristics of each measuring point on the gearbox are got. Finally, the relationship between the speed, load and the vibration frequency response of the gearbox is studied, and the vibration characteristics of the output shaft are sensitive to the fluctuation of the speed or load.
According to the basic principle of singular value decomposition(SVD), the SVD has good noise reduction effect is verified in signal processing by the numerical simulation method. Using this method to calculate the singular value and singular value difference spectrum of tooth surface point erosion fault vibration test data, the phase space of real signal and noise signal and the reconstruction order of real signal are obtained, and then the real signal is reconstructed. The time domain signal of real vibration after noise reduction is acquired. On the basis of this, FFT transformation of the real vibration acceleration signal is carried out, and the frequency response characteristics of each measuring point on the gearbox are got. Finally, the relationship between the speed, load and the vibration frequency response of the gearbox is studied, and the vibration characteristics of the output shaft are sensitive to the fluctuation of the speed or load.
引文
[1]冯志鹏,赵镭镭,褚福磊.行星齿轮箱齿轮局部故障振动频谱特征[J].中国电机工程学报,2013,33(5):119-127.
[2]程俊,王硕,武通海,等.基于拓展有限元的齿轮点蚀磨粒形态学特征模拟[J].机械工程学报,2016,52(15):99-105.
[3]冯松,毛军红,谢友柏.齿面磨损对齿轮啮合刚度影响的计算与分析[J].机械工程学报,2015,51(15):27-32.
[4]崔伟成,李伟,孟凡类,等.奇异值分解降噪结合局部特征尺度分解的轴承故障诊断[J].机械传动,2016,40(5):128-133.
[5]刘亚琼,王铁,张瑞亮,等.基于EEMD能量熵的齿轮接触疲劳试验点蚀分析[J].机械强度,2014,36(3):363-367.
[6]SAWALHI N,RANDALL R B,ENDOH.The enhancement of fault detection and diagnosis in rolling element bearings using minimum entropy deconvolution combined with spectral kurtosis[J].Mechanical Systems and Signal Processing,2007,21:2616-2633.
[7]WANG Z J,HAN Z N,GU F S,et al.A novel procedure for diagnosing multiple faults in rotating machinery[J].ISA Trans,2015,55:208-218.
[8]熊炘,杨世锡,甘春标,等.齿轮点蚀的多通道数据融合识别方法[J].振动、测试与诊断,2014,34(1):63-68.
[9]张仰明,骆志高,赵俊丽.矿用拉伸件裂纹声发射信号的SVD降噪研究[J].矿山机械,2015,43(12):139-144.
[10]刘佳音,于晓光,王琦,等.基于Hankel矩阵与奇异值分解降噪方法的齿轮故障诊断研究[J].机床与液压,2018,46(1):158-162.
[11]雷达,钟诗胜.基于奇异值分解和经验模态分解的航空发动机健康信号降噪[J].吉林大学学报:工学版,2013,43(3):764-770.
[2]程俊,王硕,武通海,等.基于拓展有限元的齿轮点蚀磨粒形态学特征模拟[J].机械工程学报,2016,52(15):99-105.
[3]冯松,毛军红,谢友柏.齿面磨损对齿轮啮合刚度影响的计算与分析[J].机械工程学报,2015,51(15):27-32.
[4]崔伟成,李伟,孟凡类,等.奇异值分解降噪结合局部特征尺度分解的轴承故障诊断[J].机械传动,2016,40(5):128-133.
[5]刘亚琼,王铁,张瑞亮,等.基于EEMD能量熵的齿轮接触疲劳试验点蚀分析[J].机械强度,2014,36(3):363-367.
[6]SAWALHI N,RANDALL R B,ENDOH.The enhancement of fault detection and diagnosis in rolling element bearings using minimum entropy deconvolution combined with spectral kurtosis[J].Mechanical Systems and Signal Processing,2007,21:2616-2633.
[7]WANG Z J,HAN Z N,GU F S,et al.A novel procedure for diagnosing multiple faults in rotating machinery[J].ISA Trans,2015,55:208-218.
[8]熊炘,杨世锡,甘春标,等.齿轮点蚀的多通道数据融合识别方法[J].振动、测试与诊断,2014,34(1):63-68.
[9]张仰明,骆志高,赵俊丽.矿用拉伸件裂纹声发射信号的SVD降噪研究[J].矿山机械,2015,43(12):139-144.
[10]刘佳音,于晓光,王琦,等.基于Hankel矩阵与奇异值分解降噪方法的齿轮故障诊断研究[J].机床与液压,2018,46(1):158-162.
[11]雷达,钟诗胜.基于奇异值分解和经验模态分解的航空发动机健康信号降噪[J].吉林大学学报:工学版,2013,43(3):764-770.