机构地区: 浙江大学机械工程学系
出 处: 《浙江大学学报(工学版)》 2002年第6期642-645,共4页
摘 要: 如果转子机械系统存在潜在的缺陷,那么在转子启动过程的振动信号中就会出现异常现象.因此研究转子启动过程的故障诊断方法对于选择合理的应变措施是至关重要的.该过程的故障可以看作是若干主变量依赖于时间的动态模式.隐Markov模型(HMM)已经证明是学习动态时间序列的概率模型的最广泛应用的工具之一,它可以使用一个隐变量来模拟系统的动态行为的变化.实验证明,HMM可以对转子启动过程的故障进行有效的模拟诊断. If latent defects exist in the rotor machinery system,rotor vibration signals in the whole runup process can present abnormal phenomenon. Therefore study on metbods for diagnosing rotor faults in the runup process is crucial for the selection of appropriate response actions. The rotor machine faults in this process can be conridered as timedependent dynamical patterns related to the principal variables. Hidden Markov models have proven to be one of the most widely used tools for learning probabilistic models of dynamics time series. HMM can model dynamical behavior variation existing in the system through a latent variable (hidden states). It was verified that HMM can effectively model and identify the rotor machine faults in the runup process.