论文部分内容阅读
针对机器设备磨损故障中由于采样数据样本较少,从而故障不易诊断的情况,提出了一种基于极大熵原理的磨损状态判别标准的方法.首先利用信息论中的极大熵原理对机器设备的油液分析数据进行处理,得到油样监测数据的最优无偏估计.利用油样监测数据的无偏估计得到它的算术均值和方差,然后结合三线值法得到基于极大熵原理的磨损状态判别标准(正常值、警告值、危险值).最后以某石化企业厂鼓风机的监测为例,分析油样铁谱数据,建立故障状态判别标准,得到初步的判别结果,并与实际的大修结果进行比较,发现所建立的方法是有效的.
Aiming at the situation that the equipment wear failure is not easy to diagnose due to less sample data, a method based on the maximum entropy principle is proposed.Firstly, the maximum entropy principle in information theory is applied to the machine equipment The oil analysis data is processed to obtain the optimal unbiased estimation of the oil sample monitoring data.The arithmetic mean and variance of the oil sample data are obtained by unbiased estimation of the oil sample monitoring data and then the wear state based on the maximum entropy principle Finally, taking the monitoring of a blower of a petrochemical plant as an example, the data of oil samples were analyzed, and the criterion of fault status was established to get the preliminary results of the discrimination. The results were compared with the actual overhaul results The comparison shows that the established method is effective.