运用EMD和GA SVM的齿轮故障特征提取与选择

针对齿轮故障特征提取,首先将齿轮箱振动信号进行经验模态分解,得到一组固有模态函数.计算各固有模态函数的能量和矩阵的奇异值,采用Shannon熵和Renyi熵度量能量和奇异值分布,构成原始特征子集.再采用遗传算法和最小二乘支持向量机的Wrapper方法选择最优特征子集.该方法能够利用较少的特征参数集准确判别齿轮故障,提高了齿轮故障诊断的精度与效率.

Gear Fault Diagnosis Using Empirical Mode Decomposition, Genetic Algorithm and Support Vector Machine

In order to extract the gear fault features, firstly, the gearbox vibration signal was decomposed as intrinsic model functions (IMF) by using the empirical mode decomposition (EMD) method. The ener gy of every IMF and the singular value of the IMF matrix were chosen as features. The Shannon and Renyi entropy of the energy and singular value distribution were also extracted. Secondly, a wrapper feature se lection method employing the genetic algorithm and the least square support vector machine (LS-SVM) was used to search the optimal feature subsets for the gear fault diagnosis. The results demonstrate that the proposed approach can detect the gear faults by only using a small feature set with high accuracy and efficiency.