引用本文:刘佳俊,胡昌华,周志杰,张鑫,王鹏.基于证据推理和置信规则库的装备寿命评估[J].控制理论与应用,2015,32(2):231~238.[点击复制]
LIU Jia-jun,HU Chang-hua,ZHOU Zhi-jie,ZHANG Xin,WANG Peng.Life assessment approach of equipment based on belief-rule-base and evidential reasoning[J].Control Theory and Technology,2015,32(2):231~238.[点击复制]
基于证据推理和置信规则库的装备寿命评估
Life assessment approach of equipment based on belief-rule-base and evidential reasoning
摘要点击 3438  全文点击 2363  投稿时间：2014-05-06  修订日期：2014-09-25
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2015.40406
  2015,32(2):231-238
中文关键词  置信规则库  证据推理  寿命评估  环境因子
英文关键词  belief-rule-base  evidential reasoning  life assessment  environment factor
基金项目  国家杰出青年科学基金项目(61035014), 国家自然科学基金项目(61370031)资助.
作者单位E-mail
刘佳俊 第二炮兵工程大学 302教研室 Liujiajunxian@163.com 
胡昌华 第二炮兵工程大学 302教研室  
周志杰* 第二炮兵工程大学 302教研室  
张鑫 第二炮兵工程大学 302教研室  
王鹏 第二炮兵工程大学 302教研室  
中文摘要
      针对航天产品试验样本少, 寿命评估难的特点, 结合产品在研制阶段多种工作环境的失效数据, 提出了一 种基于证据推理(evidential reasoning, ER)和置信规则库(belief-rule-base, BRB)进行装备寿命评估的新方法. 首先, 分 析了模型的合理性并使用多维BRB模型将多种环境下的寿命数据折合为标准工作环境下的寿命数据, 然后通过ER 算法将折合后数据和实际工作环境数据进行融合. 其次, 详细说明了BRB--ER模型的推理过程和寿命评估的步骤. 最后, 采用某航天产品的失效数据对该方法进行了验证, 并用已有的产品寿命的固定值进行BRB的参数更新. 研究 结果表明, 在专家知识准确合理时, 该模型能够准确地评估产品寿命, 并可根据已有的产品的固定寿命进行训练, 建 立更加准确的寿命预测模型.
英文摘要
      It is difficult to assess the life of equipment by using few sample data. As such, based on the belief-rule-base (BRB) and evidential reasoning (ER) approach, a new BRB-ER model is proposed, which uses effectively few failure data and expert knowledge. Firstly, the rationality of employing the model is analyzed, where BRB is used to calculate the environmental factor between one environment and the standard working environment by using few life data and expert knowledge, and the ER approach is adopted to aggregate the data under various environments to assess the life of products. Secondly, the process of reasoning by using BRB-ER model and steps of assessing the life of the equipment is expounded. Finally, a case study is carried out to illustrate the ability and the training process of BRB-ER model. This study proves that the model is effective for the life assessment by using expert knowledge and update parameters for a better model.