机构地区: 西南石油大学电气信息学院
出 处: 《测控技术》 2013年第3期22-25,共4页
摘 要: 针对非线性模拟电路故障诊断中软故障诊断的难题,提出了Volterra级数结合隐马尔科夫模型(HMM)进行故障诊断的方法。首先利用梯度搜索算法求解Volterra级数并提取出故障特征,然后利用提取出来的故障特征构造出观察变量对隐马尔科夫模型进行训练,最后用训练好的隐马尔科夫模型完成故障诊断。实验结果表明,该方法能有效提取故障特征,提高故障诊断效果。 Considering the problem of soft fault diagnosis of nonlinear analog circuits, a fault diagnosis approach combining the Volterra series with the hidden Markov model (HMM) is proposed. Firstly, the Voherra series are calculated by gradient search algorithm and the fault features are extracted. Then, the extracted fault features are used to form the observation variables to train the HMM. Finally, the well-trained HMM is used to diagnose faults. The experiment shows that the proposed method can extract the fault feature effectively and the fault diagnosis is improved.
关 键 词: 非线性模拟电路 故障诊断 软故障 级数 隐马尔科夫模型
领 域: [电子电信]