| 177 | 0 | 85 |
| 下载次数 | 被引频次 | 阅读次数 |
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。
Abstract：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.
[1]陶亮，王进同.面向能耗优化的切削工艺参数多目标优化[J].组合机床与自动化加工技术，2023(3):136-139.
[2]武煜航，张华，鄢威.面向低能耗高质量的数控加工参数优化与决策方法[J].制造技术与机床，2022(7):101-108.
[3]代维.基于历史数据的型腔铣削中主轴功率仿真及进给速度优化方法[D].武汉：华中科技大学，2018.
[4]尹洋.基于加工过程主轴功率模型的三轴铣削粗加工进给速度优化[D].武汉：华中科技大学，2020.
[5] MEIJER A,KREBS E,WIEDERKEHR P,et al. Prediction of feed-rate slowdowns in precise micromilling processes[J].Journal of Manufacturing and Materials Processing,2018,2(1):19.
[6]高威，葛广言，杨赟，等.基于前馈灰色预测和模糊控制的恒功率切削控制方法[J].机械设计与研究，2022,38(3):147-151.
[7] XIE J J,ZHAO P Y,HU P C,et al. Multi-objective feed rate optimization of three-axis rough milling based on artificial neural network[J]. The International Journal of Advanced Manufacturing Technology,2021,114(5):1323-1339.
[8] ZUPERL U,CUS F,MILFELNER M. Fuzzy control strategy for an adaptive force control in end-milling[J]. Journal of Materials Processing Technology,2005,164:1472-1478.
[9] LIN J,LIAN R J. Hybrid self-organizing fuzzy and radial basis-function neural-network controller for constant cutting force in turning[J]. The International Journal of Advanced Manufacturing Technology,2011,53(9):921-933.
[10]付中涛，杨文玉，张彦辉，等.基于切削力预测模型的复杂曲面铣削进给速度优化[J].中国科学：技术科学，2016,46(7):722-730.
[11]赵慧娟，魏忠斌，胡亚南.多轴铣削加工进给速度离线优化方法[J].机械设计与制造，2022(7):183-186.
[12]周吉敏.基于主轴功率切削负载约束和进给轴运动学约束的五轴粗加工进给速度优化方法[D].武汉：华中科技大学，2021.
[13] YIN R X,WANG Z. Optimization of milling cutting parameters under carbon efficiency target[J]. Journal of Physics:Conference Series,2021,1888(1):012008.
[14]AKIYAMA Y,IWAKI M,KOMAGAMINE Y,et al. Effect of spindle speed and feed rate on surface roughness and milling duration in the fabrication of milled complete dentures:an in vitro study[J]. Applied Sciences,2023,13(24):13338.
[15]刘恒丽，王勇，董靖川.基于动态特性的数控铣削参数自适应在线优化[J].制造业自动化，2019,41(9):82-88,120.
[16]艾兴，肖诗纲.切削用量简明手册[M]. 3版.北京：机械工业出版社，1994.
基本信息:
DOI：10.19573/j.issn2095-0926.202501005
中图分类号:TG54;TP18;TP273
引用信息:
[1]王威,赵巍,魏国家等.CNN-LSTM预测和PID控制的恒功率铣削方法[J].天津职业技术师范大学学报,2025,35(01):27-32.DOI:10.19573/j.issn2095-0926.202501005.
基金信息:
辽宁省教育厅基本科研项目（LJKMZ20222111）