在分析常用小波去噪方法基础上提出基于振动信号峭度的阈值降噪方法.阈值大小取决于不同尺度上故障信号的小波分解系数.在综合软、硬阈值函数优点基础上提出新的参数可调阈值函数.与常用小波去噪方法进行对比分析结果表明,该方法可有效提取信号中冲击成分.为正确识别故障特征提供有力保证.%The commonly used wavelet de-noising methods were analyzed, and a comprehensive threshold denoising method based on kurtosis of vibration signals was proposed. The threshold depended on wavelet decomposition coefficients of fault signals with different scales. The soft and hard threshold functions were combined based on their merits, and a new threshold function with adjustable parameters was proposed. Comparative analysis with other commonly used wavelet de-noising methods was performed. It was shown that the proposed method can effectively extract shock components of vibration signals to detect fault features.
展开▼