改进小波包多阈值去噪法及其工程应用
发布时间：2018-08-15 18:35
【摘要】：针对机械振动信号提取时面临的去噪问题,在小波包多阈值准则去噪法的基础上,提出一种改进的小波包多阈值准则综合去噪方法(改进FMC去噪法)。该方法首先采用探测插值法对机床原始振动信号进行预处理,剔除受外界干扰产生的突变噪声信号;再以小波包分析为基础,根据有用信号的最小频率确定最大分解层数,并按最小代价原理确定信号分解的最佳小波包基;最后采用小波包多阈值降噪准则对振动信号进行重构,得到去噪后的机床振动信号。针对含噪blocks信号、doppler信号及模拟的含噪振动信号进行的仿真实验结果表明,改进后的FMC去噪法去噪效果优于传统方法。将该方法应用于气囊修整机振动信号分析中,结果表明,改进FMC去噪法能够有效剔除振动信号各频段的噪声,提高信号特征的可分离性。
[Abstract]:Aiming at the problem of de-noising in mechanical vibration signal extraction, an improved wavelet packet multi-threshold criterion comprehensive denoising method (improved FMC denoising method) is proposed based on wavelet packet multi-threshold criterion de-noising method. In this method, the original vibration signal of the machine tool is preprocessed by detecting interpolation method, and the sudden noise signal caused by external interference is eliminated, and then the maximum decomposition layer number is determined according to the minimum frequency of the useful signal based on wavelet packet analysis. The optimal wavelet packet basis for signal decomposition is determined according to the principle of minimum cost. Finally, the vibration signal is reconstructed by wavelet packet multi-threshold de-noising criterion, and the vibration signal of machine tool after denoising is obtained. The simulation results of the noisy blocks signal and the simulated noisy vibration signal show that the improved FMC denoising method is better than the traditional method. The method is applied to the vibration signal analysis of air bag repair machine. The results show that the improved FMC denoising method can effectively eliminate the noise of the vibration signal and improve the separability of the signal characteristics.
【作者单位】： 厦门大学航空航天学院;
【基金】：国家自然科学基金资助项目(51675453) 深圳科技计划资助项目(JCYJ20160517103720819)
【分类号】：TN911.7
本文编号：2185059
[Abstract]:Aiming at the problem of de-noising in mechanical vibration signal extraction, an improved wavelet packet multi-threshold criterion comprehensive denoising method (improved FMC denoising method) is proposed based on wavelet packet multi-threshold criterion de-noising method. In this method, the original vibration signal of the machine tool is preprocessed by detecting interpolation method, and the sudden noise signal caused by external interference is eliminated, and then the maximum decomposition layer number is determined according to the minimum frequency of the useful signal based on wavelet packet analysis. The optimal wavelet packet basis for signal decomposition is determined according to the principle of minimum cost. Finally, the vibration signal is reconstructed by wavelet packet multi-threshold de-noising criterion, and the vibration signal of machine tool after denoising is obtained. The simulation results of the noisy blocks signal and the simulated noisy vibration signal show that the improved FMC denoising method is better than the traditional method. The method is applied to the vibration signal analysis of air bag repair machine. The results show that the improved FMC denoising method can effectively eliminate the noise of the vibration signal and improve the separability of the signal characteristics.
【作者单位】： 厦门大学航空航天学院;
【基金】：国家自然科学基金资助项目(51675453) 深圳科技计划资助项目(JCYJ20160517103720819)
【分类号】：TN911.7
【相似文献】
相关期刊论文 前1条
1 许刚;巩稼民;梁猛;;基于光纤中拉曼散射原理实现多阈值神经网络[J];微计算机信息;2007年31期
，本文编号：2185059
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2185059.html
