Fault Diagnosis of Nuclear-Powered Equipment Based on HMM&SVM
-
摘要: 核动力设备复杂且积累的资料与故障样本少,传统的诊断方法有待改进。隐马尔可夫模型与支持向量机是一种新的智能诊断技术。本文针对核动力设备机械故障诊断的特点,采用隐马尔可夫模型建模的方式进行故障的初步诊断,再利用支持向量机小样本的强推广能力进行进一步甄别。主泵故障模拟装置上的验证实验表明,HMM&SVM混合模型具有较高的故障识别率。Abstract: For the complexity and the small fault samples of nuclear-powered equipment,a hybrid HMM/SVM method was introduced in fault diagnosis.The hybrid method has two steps: first,HMM is utilized for primary diagnosis,in which the range of possible failure is reduced and the state trends can be observed;then faults can be recognized taking the advantage of the generalization ability of SVM.Experiments on the main pump failure simulator show that the HMM/SVM system has a high recognition rate and can be used in the fault diagnosis of nuclear-powered equipment.
-
计量
- 文章访问数: 19
- HTML全文浏览量: 6
- PDF下载量: 0
- 被引次数: 0