【摘要】 钻铣数控机床是现代造纸过滤筛加工的主要设备，一旦该设备出现故障，过滤筛钻孔精度就会受到极大影响。面对这种情况，研究一种基于多维贝叶斯的造纸过滤筛钻铣数控机床故障诊断方法。采集钻铣数控机床工作过程中的振动信号和电流信号并降噪处理。提取信号时域特征和频域特征，分别为波形因子、偏度、重心频率、频率均方根，组成特征向量。构建多维贝叶斯模型并进行训练，输入特征向量到模型中，计算似然函数值，找出其中极大值，极大值对应的故障类型就是诊断结果。结果表明：所研究方法用于，钻铣数控机床1存在机床装置故障；钻铣数控机床2存在伺服系统故障；钻铣数控机床3存在润滑装置故障；钻铣数控机床4存在液压装置故障；钻铣数控机床5无故障。检测结果与实际情况一致，由此证明了所研究诊断方法的准确性。更多还原

【Abstract】 The drilling and milling CNC machine tool is the main equipment for the processing of the modern paper filter screen. Once the equipment breaks down,the drilling accuracy of the filter screen will be greatly affected. In this case,a multi-dimensional Bayesian based fault diagnosis method for paper filter screen drilling and milling CNC machine tool is studied. Collect the vibration signal and current signal of the drilling and milling CNC machine tool in the working process and reduce the noise. The time-domain and frequency-domain features of the signal are extracted,which are waveform factor,skewness,center of gravity frequency,and root mean square of frequency,respectively,to form the feature vector. Construct and train a multidimensional Bayesian model,input the eigenvector into the model,calculate the likelihood function value,and find out the maximum value. The fault type corresponding to the maximum value is the diagnosis result. The results show that the drilling and milling NC machine tool 1 has machine tool device failure when the method is applied;The CNC drilling and milling machine tool 2 has servo system fault;The lubrication device of the drilling and milling CNC machine tool 3 is faulty;The drilling and milling CNC machine tool 4 has hydraulic device failure;The drilling and milling CNC machine tool 5 has no fault. The detection results are consistent with the actual situation,which proves the accuracy of the diagnosis method studied.更多还原