现代制造工程 ›› 2017, Vol. 438 ›› Issue (3): 71-76.doi: 10.16731/j.cnki.1671-3133.2017.03.012

• 数控加工技术 • 上一篇    下一篇

基于RBR-GA-BP集成的凸轮轴数控磨削工艺智能推理

张晓红1,2, 邓朝晖3, 万林林3, 刘伟3   

  1. 1 湖南理工学院机械工程学院,岳阳 414006
    2 湖南大学机械与运载工程学院,长沙 410082
    3 湖南科技大学机电工程学院,湘潭 411201
  • 收稿日期:2015-10-30 出版日期:2017-03-18 发布日期:2018-01-08
  • 作者简介:张晓红,博士,副教授,研究方向为磨削加工工艺与装备。
    E-mail:jansbomb@126.com
  • 基金资助:
    国家自然科学基金青年项目(51405158);国家863高科技资助项目(2014AA041504);中国博士后科学基金面上一等资助项目(2013M540626)

Research on intelligent reasoning of camshaft NC grinding process based on RBR-GA-BP

Zhang Xiaohong1,2, Deng Zhaohui3, Wan Linlin3, Liu Wei3   

  1. 1 College of Mechanical Engineering,Hunan Institute of Science and Technology,Yueyang 414006,Hunan,China
    2 College of Mechanical and Vehicle Engneering,Hunan University,Changsha 410082,China
    3 College of Electromechanical Engineering,Hunan University of Science and Technology,Xiangtan 411201,Hunan,China
  • Received:2015-10-30 Online:2017-03-18 Published:2018-01-08

摘要: 磨削加工工艺方案智能推理一直是困扰制造企业的难题。当前选择磨削加工工艺方案的方式仍以传统的试切法和经验法为主。为此,提出一种基于规则推理理论(Rule-based Reasoning,RBR)、遗传算法(Genetic Algorithm,GA)和BP神经网络(BP neural network,BP)等多种智能技术集成的凸轮轴数控磨削工艺智能推理模型体系结构。建立了基于变量+事实+规则的三层结构体系,推理策略采用正、反向混合推理、置信度和活性度综合排序的方式来进行冲突消解;构建了5×12×8三层GA-BP网络拓扑模型,采用测试样本对模型进行测试,结果表明,85.42%的预测值的百分比误差处于±10%以内。以某型凸轮轴毛坯加工为例,验证了凸轮轴数控磨削工艺智能推理模型的有效性。

关键词: 规则推理, 遗传算法, BP神经网络, 凸轮轴磨削, 工艺智能推理

Abstract: A problem for manufacturing enterprises is how to obtain grinding process plans intelligently.Currently choosing the grinding process plans is still focus on the traditional “test cut” method and “experience” law.On the basis of,an intelligent reasoning model of camshaft NC grinding process based on rule-based reasoning,genetic algorithm and BP neural network methods is introduced.Its structure is constructed by three layers including variable,fact and rule.Mixed reasoning with positive and negative position,an integrated sort of confidence and activity to realize conflict resolution is used in reasoning strategy.A 5×12×8 GA-BP network topology model is established.After testing,the results show that for 85.42% of the predicted values their percentage errors are within ±10%.Its effectiveness is verified by applying the model to process.

Key words: Rule-based Reasoning (RBR), Genetic Algorithm (GA), BP neural network (BP), camshaft grinding, process intelligent reasoning

中图分类号: 


版权所有 © 《现代制造工程》编辑部 
地址:北京市东城区东四块玉南街28号 邮编:100061 电话:010-67126028 电子信箱:2645173083@qq.com
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn
访问总数:,当日访问:,当前在线: