基于K-奇异值分解和层次化分块正交匹配算法的滚动轴承故障诊断

利用层次化分块正交匹配算法（HBW-OOMP）的高稀疏性和运算速度快等优点，提出了一种基于K-奇异值分解（K-SVD）字典和HBW-OOMP算法的故障轴承诊断方法。首先利用K-SVD自学习训练方法得到包含冲击成分的冗余字典，克服了固定结构字典适应性不强的缺点。然后采用基于分块思想的HBW-OOMP算法进行原子的选取和稀疏系数的求解，以重构信号包络谱峭度最大为终止条件，自适应确定分解次数。最后应用所提方法对仿真信号和故障轴承实验信号进行故障特征提取，结果表明该方法能够有效提取强背景噪声下故障特征成分，具有一定的应用前景。

Bearing Fault Diagnosis Based on K-SVD and HBW-OOMP

A new method for fault diagnosis of bearings was presented based on K-SVD and HBW-OOMP. Firstly,K-SVD dictionary training algorithm was utilized to construct a redundant dictionary containing impulsive components and the disadvantages of less adaptability of fixed structure dictionary were overcome. Then, HBW-OOMP algorithm was employed in selecting the best atom and solving the sparse coefficients. The signals were decomposed adaptively with the maximum principle of the envelop spectrum kurtosis. The fault features were extracted by the proposed method from simulated and experimental signals respectively. The results show that the method may achieve  the extraction of impulsive components from strong noise, which demonstrates the effectiveness and practicability.