基于决策树与多元支持向量机的齿轮箱早期故障诊断方法

Early Fault Diagnosis of Gearbox Based on Multiclass Support Vector Machine and Decision Tree

摘要:

齿轮箱部件的故障形式多样,典型故障训练样本数量有限；为了提高齿轮箱故障诊断的精度和效率,提出了基于决策树与多元支持向量机的齿轮箱早期故障诊断方法；利用决策树分类速度快、效率高的优点和支持向量机在小样本二元分类方面突出的特点构建多元分类识别模型,在不同故障情形下提取齿轮箱振动信号典型特征参数作为故障特征向量训练模型,并对样本进行测试；实验结果表明,该方法在小样本情况下识别效果明显优于神经网络方法,同时在识别效率方面比常规多元支持向量机方法有了较大的提高。

Abstract:

Failure form of gear box parts is varied, and typical faults have limited training samples. In order to improve the accuracy and efficiency of gearbox fault diagnosis, decision tree with multiple support vector machine (SVM) was proposed based on the early gearbox fault diagnosis methods. Classification based on decision tree on the advantage of fast speed, high efficiency and support vector machine (SVM) in binary classification has outstanding characteristics of small sample build multivariate classification model, in different typical fault case to extract the gearbox vibration signal characteristic parameters as the fault feature vector training model, and testing samples. The results show that this method not only can complete the model learning training in the case of small samples ,but also has been greatly improved over the neural network method in terms of the recognition performance,and can be effectively applied to gearbox fault diagnoisis. The results show that this method not only can complete the model learning training in the case of small samples,but also ha.