针对采煤机在正常工作中经常发生轴承故障的问题,提出一种基于经验模态分解(EMD)和隐半马尔可夫模型(HSMM)的故障识别方法。该方法先用EMD对故障信号进行分解,形成一系列平稳的本征模函数(IMF),并提取包含故障信息的IMF的能量作为特征向量输入到HSMM中对其进行训练。之后用训练好的HSMM分类器对故障进行识别。实验结果表明,采用该方法可以有效地提高故障识别率。 Aiming at the bearing fault of the shearer in normal operation, a fault identification method based on Empirical Mode Decomposition (EMD) and Hidden Markov Markov Model (HSMM) is proposed. The method first decomposes the fault signal by EMD to form a series of stationary IMFs and extracts the energy of the IMF containing the fault information as an eigenvector to input into the HSMM for training. Afterwards, the trained HSMM classifier is used to identify the fault. Experimental results show that this method can effectively improve the fault recognition rate.