中国机械工程

• 机械基础工程 • 上一篇    下一篇

基于迁移学习的数控机床节能控制决策方法

张朝阳;吉卫喜;彭威   

  1. 1. 江南大学机械工程学院,无锡,214122
    2. 江苏省食品先进制造装备技术重点实验室,无锡,214122
  • 出版日期:2020-12-10 发布日期:2020-12-18
  • 基金资助:
    国家自然科学基金资助项目(51805213);
    江苏省自然科学基金青年基金资助项目(BK20170190);
    山东省重大科技创新工程项目(2019JZZY020111);
    江苏省精密与微细制造技术重点实验室开放基金资助项目(JS2020021)

Decision-making Method for Energy-saving Control of CNC Machine Tools Based on Transfer Learning

ZHANG Chaoyang;JI Weixi;PENG Wei   

  1. 1. School of Mechanical Engineering, Jiangnan University, Wuxi, Jiangsu, 214122
    2. Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and  Technology, Wuxi, Jiangsu, 214122
  • Online:2020-12-10 Published:2020-12-18

摘要: 为降低机床等待过程的能耗,提出了基于迁移学习的数控机床节能控制决策方法。通过分析机床的节能控制条件,提出了机床等待过程的5种节能控制策略。考虑生产过程数据的多样性与复杂性,建立了融合深度置信网络与迁移学习的机床节能决策方法,电梯零部件制造车间的案例分析表明该方法的机床节能决策误差仅为3.2%,机床等待过程总能耗降低了52.5%。

关键词: 迁移学习, 节能控制, 数控机床, 深度置信网络

Abstract: In order to reduce energy consumption of machine tools in waiting processes, a energy-saving control decision-making method for CNC machine tools was proposed based on transfer learning. By analyzing energy-saving control conditions of machine tools, 5 energy-saving control strategies of the machine tools in waiting processes were proposed. Considering diversity and complexity of production data, an energy-saving decision-making method was established based on deep belief networks and transfer learning. Through analyzing a case of elevator part manufacturing workshops, it is found that energy saving decision errors of this method are as 3.2%, and the total energy consumption in machine tools waiting processes is reduced by 52.5%.

Key words: transfer learning, energy-saving control, computer numerical control(CNC) machine tools, deep belief networks

中图分类号: