基于专家知识和模糊推理的大科学装置故障诊断方法

Fault Diagnosis Method of Big Scientific Device Based on Expert Knowledge and Fuzzy Reasoning

摘要:

针对大科学装置技术综合、结构复杂、系统庞大，在故障诊断方面面临的故障机理不清楚，难以建立精确的数学模型；诊断信息不完整、不精确，难以进行确定性推理；诊断数据受限，无法实现数据驱动等诸多问题。提出了基于专家知识和模糊推理相结合的故障诊断方法和模式匹配算法，通过模糊因子的引入和基于数据库的模糊诊断知识可视化建模方法的使用,解决了故障诊断环节的诸多不确定性问题，形成了面向用户的模糊专家系统故障诊断基础平台，并在某大型激光驱动装置测试验证平台中得到初步应用，实现了电气驱动及控制系统故障的智能诊断。

Abstract:

A fault diagnosis method and pattern matching algorithm based on combination of expert knowledge and fuzzy reasoning is proposed.,Aiming at many problems faced by large scientific installations in fault diagnosis, such as it is difficult to establish an accurate mathematical model because of the failure mechanism is not clear, the diagnostic information is incomplete and inaccurate, so it is difficult to carry out deterministic inference and Diagnostic data limitations do not enable data-driven self-learning .The method solves many Uncertainty issues in the fault diagnosis process, and forms User-oriented fuzzy expert system fault diagnosis basic platform by ntroducing fuzzy factors in diagnostic knowledge and the use of database-based fuzzy knowledge modeling methods,which Has been initially applied in a large-scale laser drive device test platform and Achieved intelligent diagnosis of system failure.