【作者】 崔泽朋；

【导师】 刘吉臻； 曾德良；

【作者基本信息】 华北电力大学（河北） ， 控制理论与控制工程， 2007， 硕士

【摘要】 电厂生产过程复杂,设备故障存在不确定性,由传统故障诊断技术得到的诊断结果可信度不高。因此将信息融合技术应用到故障诊断中,可以提高诊断结果的可信度。本文将基于D-S证据理论的诊断方法用于煤粉细度的监测,针对基本概率赋值难以获取问题,介绍了基于典型样本的获取方法,并应用正态分布加以改进;提出了基于灰关联分析与D-S证据推理的分层信息融合故障诊断方法,避免了灰关联分析时的无量纲化问题,同时客观的获取基本概率赋值,通过在磨煤机故障诊断中应用,验证了方法的可行性。最后,介绍了基于SIS平台的制粉系统故障诊断软件系统设计方案。更多还原

【Abstract】 Because of the complexity of power plant and the uncertainty of equipment fault, the result which comes from the traditional fault diagnosis technology is not highly reliable. Information fusion technology is applied to fault diagnosis, which can improve reliability of the diagnostic result. The paper uses fault diagnosis method based on D-S evidence theory to monitor pulverized coal granularity. It is difficult to obtain the basic probability assignment, so the acquisition method based on typical sample is introduced and improved by normal distribution. The hierarchical information fusion fault diagnosis method based on gray correlation analysis and D-S evidence theory is presented in the paper. The method has avoided dimensionless processing of gray correlation analysis. At the same time, it can externally obtain basic probability assignment. The method is applied to fault diagnosis for coal pulverizer and its feasibility is validated. Finally, it introduces the design scheme of the software system for pulverizing system fault diagnosis based on SIS.更多还原