摘要
针对机匣电解加工工艺复杂、加工环境封闭且故障发生时特征不明显、诊断不及时等问题,提出一种利用遗传算法优化BP神经网络的电解加工状态预测与故障诊断方法,不依赖于加工人员的诊断经验。通过对机匣电解加工过程进行故障诊断试验,表明正常情况下优化后的BP神经网络的识别准确率可达95%,该方法为机匣电解加工故障诊断提供了一种新的模式,有效提高机匣电解加工的可靠性。
Aiming at the problems of complex electrochemical machining process,closed machining environment,inconspicuous features and untimely diagnosis when faults occur,a method of electrochemical machining state prediction and fault diagnosis based on genetic algorithm optimized BP neural network was proposed,which doesn’t depend on the diagnosis experience of workers. Through the fault diagnosis test of the case in the electrochemical machining process,the result showed that the recognition accuracy of the optimized BP neural network can reach 95% under normal conditions,which provides a new mode for the fault diagnosis of the case electrochemical machining and effectively improves the reliability of the case electrochemical machining.
作者
顾聪聪
王福元
王娟
方洵
GU Congcong;WANG Fuyuan;WANG Juan;FANG Xun(College of Mechanical Engineering,Yancheng Institute of Technology,Yancheng 224051,China)
出处
《电加工与模具》
2023年第1期38-43,共6页
Electromachining & Mould
关键词
电解加工
BP神经网络
状态预测
故障诊断
electrochemical machining
BP neural network
state prediction
fault diagnosis