论文部分内容阅读
为了提高往复式压缩机故障诊断效率和质量 ,研究了贝叶斯不确定性推理的基本理论和原理 ,分析了贝叶斯不确定性推理的特点和用于设备故障诊断的优势。探讨了几种贝叶斯网络先验概率的获取途径。通过实际分析确定出了往复式压缩机贝叶斯诊断网络的先验概率 ,为利用不确定性信息和不确定关系进行推理提供了条件。研究了往复式压缩机贝叶斯推理算法、算法步骤 ,并在 Windows环境下用 VisualBasic语言开发的系统 ,建立了往复式压缩机贝叶斯诊断网络系统。
In order to improve the efficiency and quality of reciprocating compressor fault diagnosis, the basic theory and principle of Bayesian uncertainty reasoning are studied. The characteristics of Bayesian uncertainty reasoning and the advantages of equipment failure diagnosis are analyzed. Several ways to obtain the prior probability of Bayesian networks are discussed. The prior probability of the Bayesian diagnosis network of the reciprocating compressor is determined through the actual analysis, which provides the conditions for the reasoning using the uncertainty information and the indeterminate relationship. Reciprocating compressor Bayesian inference algorithm, algorithm steps and the system developed by Visual Basic language under Windows environment are studied, and a reciprocating compressor Bayesian diagnosis network system is established.