铣削刀具循环神经网络预测寿命方法  

Prediction Method of Milling Cutter Life Based on RNN

作　　者：陆浩天 冀晨宇 张晟玮 沈彬[1] LU Haotian;JI Chenyu;ZHANG Shengwei;SHEN Bin(School of Mechanical Engineering,Shanghai Jiaotong University,Shanghai 200240,China;AECC AERO Science and Technology Co.,Ltd.,Sichuan Chengdu 610503,China)

机构地区：[1]上海交通大学机械与动力工程学院,上海200240 [2]中国航发航空科技股份有限公司,四川成都610503

出　　处：《机械设计与制造》2025年第3期66-70,共5页Machinery Design & Manufacture

基　　金：上海航天科技创新基金(SAST2018-055)。

摘　　要：在开展新的钛合金铣削工艺加工前,总是离不开刀具的选用,一把适应该工艺的刀具能有效提高生产效率,降低加工成本。然而由于钛合金的难加工性,在选刀过程中总会消耗很多时间和人力在试切过程中。为了降低试切消耗的时间和人力成本,提出了一种利用铣削试切过程中采集到的扭矩信号来评估和预测刀具在其寿命中的可用次数,并实现不同刀具间的横向对比以获得最佳刀具的方法。通常加工时扭矩的有效值已经能表征了铣刀的磨损状态,但不同刀具在加工时由于刃倾角、涂层等参数不同导致同样磨损量的刀具在加工时扭矩的有效值不同。该方法将提取扭矩的高频信号来表征加工状态,以避免刀具参数不同带来的影响;继而利用循环神经网络,把多次试切时的信号特征时序转化为刀具的寿命预测。实验结果表明,该方法能有效地通过前几次试切来预测刀具使用寿命,使得试切消耗的次数缩短到原来的1/3;能适用于不同刃倾角和涂层的刀具,并由此横向比对出最适应本次加工的刀具。Before the new titanium alloy milling process,it is always inseparable from the selection of tools.A tool suitable for the process can effectively improve the production efficiency and reduce the processing cost.However,due to the difficult machinability of titanium alloy,it will always consume a lot of time and manpower in the process of tool selection.In order to reduce the time and labor cost of trial cutting,a method using the torque signal collected in the process of milling trial cutting to evaluate and predict the available times of tools in their life,and realize the horizontal comparison between different tools to obtain the best tools is proposed.In this method,the high-frequency signal of torque is extracted to characterize the machining state,so as to avoid the influence of different tool parameters;Then the RNN is used to transform the signal characteristic time series of multiple trial cutting into tool life prediction.The experimental results show that this method can effectively predict the service life of the tool through the previous trial cuts,and reduce the number of trial cuts to 1/3 of the original;It can be applied to tools with different blade inclination and coating,and the most suitable tool for this machining can be compared laterally.

关 键 词：刀具寿命预估 扭矩信号 循环神经网络 刀具优选 

分 类 号：TH16[机械工程—机械制造及自动化]