English

中国农机化学报

中国农机化学报 ›› 2021, Vol. 42 ›› Issue (11): 103-109.DOI: 10.13733/j.jcam.issn.20955553.2021.11.16

• 农业信息化工程 • 上一篇    下一篇

基于粗糙遗传算法在打捆机齿轮箱故障诊断中的应用*

任彬1, 白东1, 谢虎2, 韩柏和2, 肖苏伟3   

  1. 1.石家庄铁道大学机械工程学院,石家庄市,050043;
    2.农业农村部南京农业机械化研究所,南京市,210014;
    3.中国热带农业科学院橡胶研究所,海口市,571101
  • 收稿日期:2020-11-25 修回日期:2021-09-12 出版日期:2021-11-15 发布日期:2021-11-15
  • 通讯作者: 谢虎,男,1983年生,安徽合肥人,博士,副研究员;研究方向为农作物秸秆与牧草收集装备。E-mail: tmjamexh@163.com
  • 作者简介:任彬,女,1982年生,河北石家庄人,博士,副教授;研究方向为旋转机电设备状态检测、故障诊断及预警。E-mail: renbin@stdu.edu.cn
  • 基金资助:
    *国家重点研发计划资助项目(2016YFD0701304);石家庄铁道大学研究生创新资助项目(YC2020035)

Application of rough genetic algorithm in fault diagnosis of gearbox of bundling machine

Ren Bin1, Bai Dong1, Xie Hu2, Han Baihe2, Xiao Suwei3   

  1. 1. School of Mechanical Engineering, Shijiazhuang Tiedao University, Shijiazhuang, 050043, China;
    2. Nanjing Institute of Agricultural Mechanization, Ministry of Agriculture and Rural Affairs, Nanjing, 210014, China;
    3. Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Haikou, 571101, China
  • Received:2020-11-25 Revised:2021-09-12 Online:2021-11-15 Published:2021-11-15

摘要: 自走式连续作业打捆机是一款实现不停机连续打捆作业的新型秸秆收集装备,其关键功能部件齿轮箱发生故障会严重影响正常打捆工作。针对齿轮箱故障的防控和监测,提出一种结合粗糙集和遗传算法的故障诊断方法。该方法使用时域频域分析得到的多项故障特征参数作为条件属性,故障类型作为决策属性,并利用自适应遗传算法得到决策规则表,实现无需先验信息的属性约简和故障诊断。在齿轮箱故障诊断试验中,分别对不同故障类型进行信号采集和诊断分析,结果显示:该方法在无先验信息的条件下将12项故障特征参量约简为3项,根据决策规则表进行故障诊断的准确率为100%,结果表明该方法能准确判断故障的发生和故障类型,对实现故障监测和防控具有重要意义。

关键词: 打捆机, 齿轮箱, 故障诊断, 粗糙集, 遗传算法, 属性约简

Abstract: The self-propelled continuous operation baler is a new type of straw collection equipment that realizes continuous baling operations. The failure of one of its key functional components will seriously affect the regular baling work. This paper proposed a fault diagnosis method combining rough sets and genetic algorithms to monitor, prevent, and control gearbox faults. This method used the multiple fault characteristic parameters obtained from time-domain and frequency-domain analysis as conditional attributes and the fault type as decision attribute to obtain a decision rule table using the genetic algorithm and realize the attribute reduction without prior information. In the gearbox fault diagnosis experiment, signal acquisition and diagnosis analysis were carried out on different fault types. The results showed that the method reduced the 12 fault characteristic parameters into 3 when no prior information was given. The accuracy of fault diagnosis according to the decision rule table was 100%. The results showed that this method could accurately determine the occurrence and type of faults, which was of great significance to fault monitoring, prevention, and control.

Key words: baler, gearbox, fault diagnosis, rough set, genetic algorithm, attribute reduction

中图分类号: