| 基于后验概率的支持向量机在故障诊断中的应用 |
| |
| 引用本文： | 毛荣富,朱海潮,黄映云. 基于后验概率的支持向量机在故障诊断中的应用[J]. 中国机械工程, 2006, 0(Z2) |
| |
| 作者姓名： | 毛荣富 朱海潮 黄映云 |
| |
| 作者单位： | 海军工程大学，海军工程大学，海军工程大学 武汉，430033，武汉，430033，武汉，430033 |
| |
| 摘 要： | 考虑到用后验概率能更有利于表示样本的类别信息和相关后处理,在故障诊断中引入了后验概率,同时针对故障诊断问题中误判造成的损失不同这一特点,通过引入损失函数,使诊断决策的平均损失达到最小,建立了适合故障诊断的模型。通过与其他方法的比较和诊断实例说明了该方法的合理性和有效性。
|
| 关 键 词： | 后验概率 支持向量机 故障诊断 平均损失 |
Application of SVM Based on Posterior Probability in Fault Diagnosis |
| |
| Mao Rongfu Zhu Haichao Huang Yingyun Naval University of Engineering,Wuhan. Application of SVM Based on Posterior Probability in Fault Diagnosis[J]. China Mechanical Engineering, 2006, 0(Z2) |
| |
| Authors: | Mao Rongfu Zhu Haichao Huang Yingyun Naval University of Engineering Wuhan |
| |
| Affiliation: | Mao Rongfu Zhu Haichao Huang Yingyun Naval University of Engineering,Wuhan,430033 |
| |
| Abstract: | Because the posterior probability is better in the sample class information expression and related post-processing,the SVM based on posterior probability was applied in fault diagnosis.Moreover,aiming at different losses caused by error judges in fault diagnosis,risk function was combined to reach least average risk diagnosis result.A reasonable model for fault diagnosis was built.The diagnosis examples prove the va- lidity of this method after the comparisons with other methods. |
| |
| Keywords: | posterior probability support vector machine(SVM) fault diagnosis average risk |
| 本文献已被 CNKI 等数据库收录！ |
