基于智能事例推理技术的凝汽器在线监测和故障诊断系统  

Condenser On-line Status Monitoring and Fault Diagnosis System Based on Intelligent Case-based Reasoning Technique

作　　者：邱凤翔[1] 徐治皋[1] 李旭辉[2] 司风琪[1] 乔宗良[1] 张晓[1] 

机构地区：[1]东南大学能源与环境学院,南京210096 [2]大唐淮南洛河发电厂,淮南232000

出　　处：《动力工程》2009年第6期554-558,共5页Power Engineering

摘　　要：提出了基于知识库和事例推理技术相结合的双背压凝汽器状态监测与故障诊断的方法.通过对凝汽器常见故障特点以及凝汽器主要功能的分析,结合现场经验,建立了以凝汽器真空故障为最高层次的"征兆-原因"层次化结构模型,并由此建立了层次化结构知识库,用于新事例初始化故障智能定位.提出了节点搜索法的事例检索新方法,缩短了事例平均检索时间.介绍了凝汽器在线状态监测与故障诊断系统的体系结构和工作流程.通过某600 MW机组凝汽器现场运行数据和凝汽器故障诊断实例,验证了该方法的有效性,表明系统具有较强的实用价值.Based on knowledge base and case-based reasoning technique, the method for double pressure condenser status monitoring and fault diagnosis was proposed. By analyzing the characteristics of the condenser common faults and main functions, combining with the practical experience, a model with ＇signreasons＇ hierarchical structure was built, which taking the condenser vacuum faults as the top. A hierarchical structure of knowledge base which was used for new fault initial intelligent location was established. Node-Searching method as a new case searching method was given, which can shorten the average retrieval time of the case. System structure and work flow of condenser on-line status monitoring and fault diagnosis was introduced. Condenser running data and the practical condenser faults from a 600 MW unit have proved the validity of this method. Results show that this system has high practical value.

关 键 词：凝汽器 真空 事例推理 节点搜索法 故障智能定位 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]