基于半定量信息的复杂动态系统故障预测与最优维护研究

It is important and significant to predict the failure and carry out the optimal maintenance of complex dynamic systems such as gyro. Supported by the National Science Foundation of China under Grant 61004069, methods for online updating the belief rule base (BRB) parameters are systematically and deeply. Moreover, in order to solve the problems of failure prognosis and optimal maintenance under environmental influence, a new HMC-BRB model is proposed and an algorithm for online updating HMC-BRB model is further developed. At present, some periodical results have been obtained in the youth fund. 1 monograph has been published. 26 academic papers have been published and 6 papers have been indexed by SCI in which two long papers are published by IEEE TSMCB and TSMCA respectively. The applicant has obtained the national 100 excellent doctoral dissertation award nomination and the military outstanding doctoral dissertation award in 2012 and 2011, respectively. Based on the youth fund, some problems about failure prognosis and optimal maintenance will be studied further in this project on the basis of semi-quantitative information. In particular, based on extending the basic theory of BRB (mainly refers to the methods for determining the BRB attributes and their reference values, and the method for reasoning of BRB when the attributes are related), methods for online predicting the hidden failure under testing influence and competitive failures will be proposed. Moreover, methods for determining the optimal maintenance scheme and the optimal testing instant will be developed. At same time, the proposed methods will be applied in the inertial platform of missile.

对诸如陀螺仪等动态系统进行故障预测和最优维护具有重要意义。申请人在青年基金资助下，系统深入地研究了置信规则库（BRB）在线更新算法，基于此，提出一种新的HMC-BRB模型及其参数在线更新算法，解决了存在环境影响下动态系统隐含故障在线预测和最优维护问题。目前该项目已取得阶段性成果，出版专著1部，发表学术论文26篇，其中SCI论文6篇，包括IEEE TSMCB和TSMCA长文各1篇，分别于2012年和2011年获全国百优提名奖和军队优秀博士学位论文奖。本项目在青年基金的基础上，进一步研究基于半定量信息的动态系统故障预测和最优维护若干问题。特别地，在拓展BRB基本理论的基础上（主要提出BRB前提属性及其参考值确定方法和前提属性相关时的推理方法），提出存在测试影响和竞争失效条件下隐含故障的在线预测方法，并提出基于故障预测的最优维护方案和最佳测试时机确定方法，同时将所提方法应用于导弹惯性平台。

由于在国家自然科学青年基金研究过程中取得了较为突出的成果，申请人于2013年又获得了该青年连续面上基金的资助。本基金的主要研究目标是：.1）在理论上，通过突破置信规则库（BRB）前提属性及其参考值确定和前提属性相关时推理等难题，解决复杂动态系统在线故障预测和最优维护所涉及的核心基础理论问题，从而丰富和发展基于BRB的建模理论与方法。.2) 在工程上，应用所提方法解决惯性平台的在线故障预测和最优维护问题，为提高武器装备可靠性和部队战斗力服务。.基于该基金的研究成果，已经在科学出版社出版专著2部；发表论文46篇，其中以第一作者或通信作者在IEEE Trans. on Systems, Man and Cybernetics: Systems、IEEE Trans. on Fuzzy Systems和自动化学报等国内外知名期刊上发表论文15篇，包括IEEE Trans期刊论文5篇；获得中国电子学会科技进步二等奖1项（排名第二）、陕西省第十三届自然科学优秀学术论文二等奖1项（排名第一）、中国自动化学会高等教育教学成果二等奖1项（排名第二）、军队科技进步二等奖2项（分别排名第四和第六）和2013年吉林省高校科研成果一等奖1项（排名第二）；培养博士研究生5名，硕士研究生4名。已经超额完成了预期的研究任务, 后续研究工作再一次受到了2017年度国家自然科学基金面上项目和国家自然科学基金重点项目的资助。