电厂燃气轮机热参数监测过程中的方差分析

Application of mean square deviation analysis in the research of the monitoring standards of thermal parameters for gas turbine plant

通过分析上海闸电燃机发电厂监测数据,发现一年中大气温度和功率的变化使得热参数监测标准难以选取。为了找出与大气温度变化、功率变化无关的热参数监测标准,本文从方差入手,采取时间序列模型方差均值方式对运行数据进行了分析。结果表明,用时间序列模型求出的方差基准线不随大气温度变化而变化,采用满负荷工况监测可以显示出故障时性能变化最大的数据。在有故障出现时,通过时间序列模型求得的各监测数据的方差将会超过警戒线而报警,从而可以起到监测作用。

When analyzing the monitoring data of Shanghai Zhadian Gas Turbine Power Plant, we found it hard to define the monitoring standardsline of thermal parameters as the atmospheric temperature and power load varied during the whole year. In order to find out the standardsline, which was independent of atmospheric temperature and power load, time series variant level models were applied in this research. The analytic result led to a confirmed monitoring standard that was not influenced by any change of atmospheric temperature and full load monitoring could show up the performace deterioration furthest.lf some faults happened,consequently the variant line would exceed the alarm monitoring standardsline. Therefore it served the purpose of inspection and monitoring.