基于概率神经网络的数控机床滚珠丝杠副的故障诊断研究  

Research on Fault Diagnosis of CNC Machine Ball Screw Movement Pair Based on Probabilistic Neural Network

作　　者：孙晋美[1] 张兴波[1] 张霞[1] SUN Jin-mei, ZHANG Xing-bo, ZHANG Xia(Qindao College, Qingdao Technological University, Qingdao 266106, Chin)

机构地区：[1]青岛理工大学琴岛学院,山东青岛266106

出　　处：《煤矿机械》2018年第5期147-149,共3页Coal Mine Machinery

摘　　要：采用概率神经网络对数控机床滚珠丝杠副的故障状态进行模式识别。首先对数控机床的故障类型及诊断方法进行了介绍,重点介绍了滚珠丝杠副的故障类型和诊断方法;然后基于概率神经网络在模式识别上的广泛应用,结合MATLAB软件对滚珠丝杠副的工作状态信息进行了模式识别,通过有限样本的网络训练和识别测试,验证了概率神经网络在滚珠丝杠副故障诊断中的准确性和适用性。Probabilistic neural network is applied to pattern recognition for the fault state of ball screw movement pair of CNC machine tools. First, the fault types and diagnosis methods of CNC machine tools are introduced, and the introduction focused on fault types and diagnosis methods of ball screw movement pair. Based on the widespread application of probabilistic neural network in pattern recognition and MATLAB software,the working state information of ball screw movement pair is identified,and the limited sample is used,the network training and recognition test verified the accuracy and applicability of probabilistic neural network in fault diagnosis of ball screw movement pair.

关 键 词：概率神经网络 滚珠丝杠副 故障诊断 模式识别 

分 类 号：TG659[金属学及工艺—金属切削加工及机床]