摘要
针对现有多传感器集成氢气检测系统不能对自身工作状态进行诊断的缺点,对基于知识的故障诊断专家系统进行了分析,提出了一种基于知识管理和知识主动推送的故障诊断方法。对提出的用于多传感器集成氢气检测系统故障诊断的主要层次结构进行了研究。根据专家系统的基本原理介绍了基于知识推送的专家系统的总体结构。分析了系统故障模式,讨论了故障诊断单元知识库的设计方法。建立了知识主动推送的推理机构模型,给出了其推理方法和步骤。最后,给出了基于知识推送的多传感器集成氢气检测系统故障诊断单元的设计方法。实验结果表明:本方法的故障诊断准确率可达到97%以上,验证了该方法的有效性。提出的方法能够主动将知识在适当的时候传递给决策者,提高了故障诊断的速度和准确度。
As multi-sensor integration hydrogen detection systems can not diagnose the working state by itself, an expert system of fault diagnosis based on knowledge was analyzed and a fault diagnosis method based on knowledge management and knowledge pushing was proposed. The hierarchical structure of the fault diagnosis method used in the multi-sensor integration hydrogen detection system was investigated. On the basis of the principle of expert systems, the texture of expert system based on knowledge pushing actively was presented. After the fault modes were analyzed, a knowledge base fault diagnosis design was discussed. Then the method and step of inference engine based on knowledge pushing actively were researched by its established model. Finally, the fault diagnosis unit for multi-sensor integration hydrogen detection system based on knowledge pushing was designed. Experimental results indicate that the fault diagnosis accuracy rate of the proposed method reaches above 97%, which verifies the effectiveness of the method. It can transfer initiatively the knowledge to the decision-maker at opportune moment, and improve the speed and accuracy rate of fault diagnosis.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2015年第6期1742-1748,共7页
Optics and Precision Engineering
基金
国家自然科学基金资助项目(No.61473095)
青年科学基金资助项目(No.61201306)
关键词
氢气检测系统
多传感器集成
故障诊断
知识推送
专家系统
hydrogen detection system
multi sensor integration
fault diagnosis
knowledge pushing
expert system
