摘要
介绍了一种基于振动信号隐马尔可夫模型(HMM)的新的齿轮故障检测和诊断方案。首先从振动信号中提取特征,这些信号既包括正常齿轮也包括故障齿轮,特征以振动信号自回归模型的多项式传递函数的反射系数为基础。这些特征用来训练HMM归类各种齿轮状况。经过试验验证,用这些特征判断故障的准确性很高。
A new gear lault detection and diagnosis scheme based on Hidden Markov Model （HMM） of vibration signals is introduced. Features are extracted from vibration signals obtained from both normal and faulty gears firstly. The features are based on the reflection coefficients of the polynomial transfer function of the autoregressive model of the vibration signal. Then the features are used to train HMMs to represent various gear conditions. After verifing by testing, can ensure that the accuracy using these features to estimate faults is very high.
出处
《煤矿机械》
北大核心
2007年第3期194-196,共3页
Coal Mine Machinery
