摘要
凝汽器真空系统故障诊断是火电厂安全运行的一个重要环节。本文通过从凝汽器传热机理和DCS采集的数据分析影响真空的主要因素,运用多元线性回归的方法来判断影响凝汽器真空的相关性大的因素,建立DCS采集的实时数据的参数值与凝汽器真空之间的关系模型。运用BP神经网络建立真空模型,通过凝汽器真空应达值与检测值对比,判断此时的真空运行状态是否合理。通过对凝汽器真空的软测量,为凝汽器的故障诊断提出理论依据。
The fault diagnosis of condenser vacuum system is the significant process in safe operation of the power plant. The method of multivariate linear regression was used to judge the correlation factors which effect vacuum of condenser through analyzing the main factors which affect vacuum of condenser based on the heat- transfer mechanism and data form DCS. Then, the relational model between condenser vacuum value and real- time parameter value of DCS was established. Finally, the rationality of vacuum operation state was estimated through the contrast of target value and assessment value of condenser vacuum based on the vacuum model using BP neural network. The theoretical basis for the fault diagnosis of condenser was proposed by the soft measurement of condenser vacuum.
出处
《东北电力大学学报》
2012年第3期54-58,共5页
Journal of Northeast Electric Power University
