改进的卷积型小波包分解及在故障诊断中的应用

发布时间：2018-12-09 19:50

【摘要】：针对卷积型小波包分解存在频带错位与频带重叠缺陷,提出了一种改进的卷积型小波包分解算法。该算法通过交换偶数位置节点小波包分解后的两节点顺序来消除频带错位缺陷,引入两算子分别从频域除去低、高频子带理想通带范围外的频率成分以消除频带重叠缺陷。由构造的故障信号进行仿真实验,并使用某直升机中减速器疲劳实验的故障数据进行了验证。结果表明:由于消除了卷积型小波包和内积型小波包分解算法中广泛存在的频率折叠、频带重叠和频带错位缺陷,改造的卷积型小波包分解算法能更方便、更有效地提取隐藏在强噪声和其他强干扰背景下的故障特征信息,从而为机械故障的诊断提供了一种强有力的分析手段。
[Abstract]:An improved wavelet packet decomposition algorithm based on convolutional wavelet packet is proposed to overcome the defects of frequency band dislocation and frequency band overlap in convolutional wavelet packet decomposition. The algorithm eliminates the frequency band misalignment defects by exchanging the order of two nodes after the wavelet packet decomposition of even position nodes. Two operators are introduced to remove the frequency components from the frequency domain and the frequency components outside the ideal passband range of the high frequency subband respectively to eliminate the frequency band overlap defects. Simulation experiments were carried out with the constructed fault signals, and the fault data of reducer fatigue test in a helicopter were used to verify the results. The results show that the improved wavelet packet decomposition algorithm is more convenient because of eliminating the frequency folding, frequency band overlap and frequency band dislocation in convolution wavelet packet decomposition and inner product wavelet packet decomposition algorithms. The fault feature information hidden in strong noise and other strong interference background is extracted more effectively, which provides a powerful analysis method for mechanical fault diagnosis.
【作者单位】：海军工程大学兵器工程系;
【基金】：国家自然科学基金资助项目(50775218) 国家部委基金资助项目(4010801020202)
【分类号】：TN911.6

【参考文献】