摘要
针对煤矿井下工作环境恶劣,摇臂轴承经常发生故障,提出一种基于小波能谱熵-BP_Adaboost的采煤机摇臂轴承故障诊断方法。对振动信号进行小波分解,提取各小波系数的能谱熵,以此作为故障特征信息。利用Adaboost算法联合多个BP神经网络,构建BP_Adaboost强分类器。通过实验验证,该方法相比单BP神经网络识别率有很大提高,总体识别精度达到92%,其中外圈轴承故障识别率达96%,具有较好的采煤机摇臂轴承故障诊断能力。
For the poor working environment of the coalface, the rocker bearing fault often occur, a fault diagnosis method of shearer rocker bearing based on wavelet energy spectrum entropy and BP_Adaboost. The vibration signal is decomposed by the wavelet basis, and to extract the energy spectrum entropy of every wavelet coefficients, that are selected as the fault feature information. The BP_Adaboost strong classifier is build by the Adaboost algorithm to joint multiple BP neural network. Experiments show that the recognition rate of using this method has a greatly improve compared with single BP neural network, the overall recognition rate is 92%, and the outer bearing fault recognition rate reach 96%, it has a good fault diagnosis ability of shearer rocker bearing.
出处
《煤矿机械》
北大核心
2014年第11期302-304,共3页
Coal Mine Machinery
