摘要
为提高铣削过程监测与刀具故障诊断精度 ,通过测量铣床的频响函数和在铣削加工中的铣床振动加速度响应信号 ,用载荷识别的方法计算铣削力 ,分别得到了用 4刀齿和 2刀齿加工时横向铣削力的识别结果 ,所得到的铣削力曲线与加工工况吻合良好。以所识别铣削力为特征参量 ,用ART2神经网络进行了铣削过程监测与铣刀故障诊断 ,其结果比直接用振动响应信号进行监测与诊断更可靠 ,从而得到较好的监测诊断结论。
In this work,the transverse milling forces were adopted as the inputs of the ART2 neural networks to monitor the milling process and to diagnose the fault of its cutter.The forces were identified by measured frequency response function and the vibration response of the milling machine.Cutting experiments with a 4-teeth cutter and a 2-tooth cutter seperately were conducted.Results show that the number of peaks in a cutting period at the measured time history of the forces is exactly the same as the number of working tooth of the cutter,and the force-input networks give more reliable and accurcy results than those obtained directly from the vibration response of the machine.
出处
《振动.测试与诊断》
EI
CSCD
2002年第2期89-92,共4页
Journal of Vibration,Measurement & Diagnosis
基金
国家自然科学基金资助项目 (编号 :5 9975 0 87)
关键词
铣削力
铣刀
故障诊断
频响函数
铣削
载荷识别
状态监测
frequency response function vibration milling load identification fault diagnosis