中国安全科学学报 ›› 2021, Vol. 31 ›› Issue (6): 1-6.doi: 10.16265/j.cnki.issn 1003-3033.2021.06.001

• 安全科学理论与安全系统科学 •    下一篇

基于特征函数和联系数的系统故障模式识别

崔铁军1 副教授, 李莎莎2 讲师   

  1. 1 辽宁工程技术大学 安全科学与工程学院,辽宁 阜新 123000;
    2 辽宁工程技术大学 工商管理学院,辽宁 葫芦岛 125105
  • 收稿日期:2021-03-14 修回日期:2021-05-15 出版日期:2021-06-28 发布日期:2021-12-28
  • 作者简介:崔铁军 (1983—),男,辽宁沈阳人,博士,副教授,主要从事系统可靠性及系统故障演化方面的研究。E-mail: ctj.159@163.com。
  • 基金资助:
    国家自然科学基金资助(52004120);辽宁省教育厅科学研究经费项目(LJ2020QNL018);国家重点研发计划项目(2017YFC1503102);辽宁工程技术大学学科创新团队项目(LNTU20TD-31)。

Research on system fault pattern recognition based on characteristic function and connection number

CUI Tiejun1, LI Shasha2   

  1. 1 College of Safety Science and Engineering, Liaoning Technical University, Fuxin Liaoning 123000, China;
    2 School of Business Administration, Liaoning Technical University, Huludao Liaoning 125105, China
  • Received:2021-03-14 Revised:2021-05-15 Online:2021-06-28 Published:2021-12-28

摘要: 为识别具有确定性和不确定性的系统故障模式,提出一种系统故障模式识别方法。利用已有故障标准模式识别故障样本模式,使用集对分析联系数表示识别的确定性和不确定性;使用空间事故树的特征函数计算联系数的分量系数,通过联系度和识别度识别故障样本模式;借助电气系统实例分析,有效识别在多因素影响下的系统故障模式。结果表明:基于特征函数和联系数的模式识别方法可表示模式识别的确定性和不确定性;该方法可使用特征函数确定联系数分量系数,利用联系度和识别度识别故障模式。

关键词: 特征函数, 联系数, 系统故障模式, 识别度, 集对分析

Abstract: In order to study system fault pattern with certainty and uncertainty, a pattern recognition method was proposed. Firstly, fault sample patterns were identified by using existing fault standard pattern, and connection number of set pair analysis was used to represent certainty and uncertainty of identification. Secondly, component coefficients of connection number were calculated by utilizing characteristic function of space fault tree, andfault sample pattern was recognized by connection degree and recognition degree. Finally, through analysis on an electrical system example, its fault mode under influence of multiple factors was effectively identified. The results show that the proposed method, based on characteristic function and connection number, can express certainty and uncertainty of pattern recognition. It uses characteristic function to determine component coefficient of connection number and uses connection degree and recognition degree for pattern recognition.

Key words: characteristic function, connection number, system fault mode, recognition degree, set pair analysis

中图分类号: