基于历史数据计算母管制机组各台汽轮机进汽流量的研究

Research on the Calculation of the Steam Rate of Turbines Based on the Historical Data in the Units With Common Steam Header System

为实现准确测量计算母管制运行机组各台汽轮机进汽流量,研究了利用机组运行的历史稳态数据进行数学建模,借助BP神经网络仿真计算母管制运行机组各台汽轮机的进汽流量的方法,并优化了神经网络隐含层神经元的数目。研究表明借助BP神经网络仿真计算母管制运行机组各台汽轮机进汽流量准确性较高,能满足工程要求,且借助厂级监控系统(SIS)平台,应用方便,解决了部分母管制运行机组各台汽轮机利用历史运行数据进行进汽流量计算的问题。

In order to calculate the steam rate of turbines connected to common steam headers, a model based on BP neural network was built and the number of neurons in the hidden layers was optimized. The research illustrates that the simulation of the model is characterized by good precision and accuracy, which could meet engineering standards, and it is convenient to use when it was installed on SIS platform. This mathematical method can solve the problem of how to calculate steam rate of turbines without flowmeter connected to common steam headers. © 2016 Chin. Soc. for Elec. Eng.