电站锅炉运行故障实时诊断专家系统

Expert System for Real Time Failure Analysis of Power Boilers

目前火力发电站运行故障的诊断主要依赖于运行人员对现测数据的分析,诊断的正确与否很大程度上取决于运行人员的素质。所以对电站这一复杂的系统很有必要进行故障的自动化诊断,专家系统的应用是近年来的一个发展方向。本文介绍一个基于深、浅两层知识的锅炉运行故障诊断专家系统,深知识包括锅炉系统的结构知识、物理模型和故障模型。浅知识为运行专家的经验知识。由于知识全面,专家系统的诊断能力比传统的基于经验的专家系统强。系统采用面向对象的知识表达,便于知识的更新和移植。该系统采用PDC—prolog编程。目前的诊断对象为锅炉的汽水系统和燃烧风烟系统。

The application of Expert System is a trend in automatic diagnoses of power plant faults in recent years. A boiler faults diagnosis Expert System which is based on deep and shallow two-tier knowledge is introduced in this paper. Because it has a robust knowledge-base, this system has stronger diagnostic ability compared with traditional Expert Systems which are only based on experience knowledge. This system uses object-oriented knowledge expression, so its knowlegde base is easy to be updated and transplanted. The Expert System is coded in PDC-prolog language.