CNN-LSTM预测和PID控制的恒功率铣削方法  

Constant power milling method based on CNN-LSTM prediction and PID control

作　　者：王威 赵巍[1] 魏国家 陈娴琦 赵冠 张宇[1] WANG Wei;ZHAO Wei;WEI Guojia;CHEN Xianqi;ZHAO Guan;ZHANG Yu(School of Mechanical Engineering,Tianjin University of Technology and Education,Tianjin 300222,China;Department of Mechanical and Electrical Engineering,Liaoning Provincial Transportation College,Shenyang 110122,China)

机构地区：[1]天津职业技术师范大学机械工程学院,天津300222 [2]辽宁省交通高等专科学校机电系,沈阳110122

出　　处：《天津职业技术师范大学学报》2025年第1期27-32,共6页Journal of Tianjin University of Technology and Education

基　　金：辽宁省教育厅基本科研项目(LJKMZ20222111).

摘　　要：为解决传统固定参数加工和控制系统响应速度慢导致的控制滞后问题,提出了一种旨在结合深度学习和自适应PID控制器,实现对台阶工件恒功率铣削的智能优化方法。该方法利用CNN-LSTM模型预测主轴功率的变化,捕捉铣削过程中的动态特性和复杂关系,并基于预测结果,通过自适应PID控制器实时调整进给倍率,实现恒功率切削。实验结果表明:基于深度学习和自适应PID控制器的进给倍率优化方案效果较好,能够在设定的进给倍率调节范围内保持恒定的功率切削,加工效率显著提高。In order to solve the control hysteresis caused by the slow response of traditional fixed-parameter machining and control systems,this paper proposes an intelligent optimization method combining deep learning and adaptive PID controllers to achieve constant-power milling of step workpieces.A CNN-LSTM model is used to predict variations of spindle power,capture the dynamic characteristics and complex relationships in the milling process,and based on the prediction results,the feed override is adjusted online by the adaptive PID controller to realize constant power cutting.The experimental results show that the feed multiplier optimization scheme based on deep learning and adaptive PID controller performs more effectively,maintains constant power cutting within the set feed multiplier adjustment range,ensures the machining quality and stability,and significantly improves the machining efficiency through continuous learning and intelligent adjustment.

关 键 词：CNN-LSTM模型 自适应PID控制 恒功率切削 

分 类 号：TG54[金属学及工艺—金属切削加工及机床]