基于改进DMI与故障树以及IFR的故障诊断专家系统设计

摘要

为了达到减少压制观瞄装置的检修时间，提高部队维修人员工作效率的目的，设计了一种压制观瞄装置故障诊断专家系统。采用基于粗糙集理论的改进DMI(改进可区别矩阵)属性约简算法和建立故障树的方式来实现专家系统的知识获取，解决专家系统知识获取的瓶颈问题；提出一种改进正向推理(IFR)算法来设计专家系统的推理机，使得诊断推理更为高效。通过对比实验来验证此故障诊断专家系统的实际应用效果，从而体其现诊断推理的可靠性好、效率高的优点。%In order to reduce the maintenance time and improve maintenance personnel’s working efifciency, an fault diagnosis expert system for the suppress sighting device is designed. The improved DMI attribute reduction algorithm based on rough set theory and fault tree is used to achieve knowledge acquisition of the expert system, the improved forward reasoning algorithm is put forward to design reasoning machine of the expert system. The practical application effect of this fault diagnosis expert system is veriifed by experimental data, it is compared with manual maintenance, so as to indicate the advantage of high reliability and efifciency while in the process of diagnostic reasoning.