引用本文:司小胜,胡昌华,张琪,李娟.基于进化信度规则库的故障预测[J].控制理论与应用,2012,29(12):1579~1586.[点击复制]
SI Xiao-sheng,HU Chang-hua,ZHANG Qi,LI Juan.Fault prognosis based on evolving belief-rule-base system[J].Control Theory and Technology,2012,29(12):1579~1586.[点击复制]
基于进化信度规则库的故障预测
Fault prognosis based on evolving belief-rule-base system
摘要点击 2233  全文点击 1511  投稿时间：2012-05-26  修订日期：2012-07-26
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/j.issn.1000-8152.2012.12.CCTA120601
  2012,29(12):1579-1586
中文关键词  专家系统  信度规则库  证据推理  故障预测
英文关键词  expert system  belief-rule-base  evidential reasoning  fault prognosis
基金项目  This work was supported by the National Nature Science Foundation of China (No. 61025014, 61174030, 61104223) and the Shandong High School Science & Technology Fund Planning Project (J09LG26).
作者单位E-mail
司小胜* 第二炮兵工程大学 302教研室
清华大学 自动化系 		sxs09@mails.tsinghua.edu.cn 
胡昌华 第二炮兵工程大学 302教研室  
张琪 第二炮兵工程大学 302教研室  
李娟 青岛农业大学 机电工程学院  
中文摘要
      在假设信度规则库(BRB)的输入为均匀分布的情况下, 已有文献提出了一种序贯自适应的学习算法以实现BRB的参数在线辨识和结构的自适应调整. 然而在实际问题中, 信度规则库的输入一般是未知的、难以得到的, 这在一定程度上限制了序贯自适应学习算法的实用性, 因此就需要研究一种改进的BRB学习算法以实现参数和结构的同时辨识. 本文在序贯自适应方法的基础上, 通过定义BRB的完整性准则, 提出了改进的BRB进化策略. 与现有方法相比, 该方法可以实现信度规则的自动增减, 且无需输入样本的概率密度函数. 此外, 该方法继承了BRB的特点, 仅需要部分的输入输出信息. 基于改进的进化策略, 提出了一种新的故障预测算法, 最后通过陀螺仪故障预测实验验证了本文方法的有效性.
英文摘要
      Recently, a sequential adaptive learning algorithm has been developed for online constructing belief-rulebased (BRB) system. This algorithm is based on the assumption that the sample density function of the inputs to BRB system obeys the uniform distribution. However, in practice, the sample density function is not always available and is difficult to be determined; this really limits the applicability of the above method. As such, it is desired to develop an improved algorithm without requiring the sample density function. In this paper, on the basis of the sequential adaptive learning algorithm, we develop an improved evolving BRB learning algorithm based on the belief-incomplete criterion. Compared with the current algorithms, a belief rule can be automatically added into the BRB or pruned from the BRB without the need of the sample density function. In addition, our algorithm inherits the features of the BRB, in which only partial input and output information are required. Based on the improved algorithm, a fault prognosis method is presented. In order to verify the effectiveness of our algorithm, a practical case study for gyroscope fault prognosis is studied and examined to demonstrate how our algorithm can be implemented.