一种基于支持向量机的齿轮箱故障诊断方法

提出了一种基于多分类支持向量机(简称MSVM)的齿轮箱故障诊断方法。先根据齿轮箱故障机理和振动特点,探讨了齿轮箱故障诊断试验方案。再测取齿轮箱振动信号,并提取了能反映齿轮箱运转信息的时频域特征参数。通过结合投票法和决策树的基本思想,有针对性地构造了多分类支持向量机决策结构并将其应用于齿轮箱故障诊断。实际齿轮箱故障诊断试验结果表明,该决策结构较好地解决了小样本学习问题,避免了人工神经网络进行诊断时出现的过学习、收敛速度慢、泛化能力弱等缺点,能有效应用于齿轮箱故障诊断。

Gearbox Fault Diagnosis Based on SVM

A gearbox fault diagnosis method based on multi-class support vector machine (MSVM) was proposed in this paper. Firstly, based on the fault mechanism and vibration characteristics of the gearbox, an experiment system of fault diagnosis was designed. Then, the vibration signals of the gearbox were acquired, and the characteristic parameters of the signals both in the time and frequency domains were extracted, which contained the operating information. Combined the basic thought of voting method and decision tree, a special decision-structure of MSVM was designed and applied to gearbox fault diagnosis. Experimental results demonstrate that the decision-structure proposed in the paper solves the small sample learning problems quite well and overcomes the shortcomings of over-fitting, longtime training and the weakness in generalization of ANN in fault diagnosis. The MSVM based decision-structure is an effective method for gearbox fault diagnosis.