论文部分内容阅读
研究了多值分类支持向量机在机械故障诊断中的应用,以滚动轴承振动信号进行了分类实验。实验表明,在小学习样本条件下SVM比RBF人工神经网络具有更好的分类性能和推广能力。SVM方法的应用为以计算机技术为基础的设备监测、智能故障诊断提供技术保障。
The application of Multivalued classification SVM in mechanical fault diagnosis is studied. The classification experiment of rolling bearing vibration signal is carried out. Experiments show that SVM has better classification performance and promotion ability than RBF artificial neural network under the condition of small learning samples. The application of SVM method provides technical support for equipment monitoring and intelligent fault diagnosis based on computer technology.