基于对向传播神经网络的汽轮机排汽焓预测计算

Predicted Calculation of Steam Turbine Exhaust Enthalpy Based on Counter Propagation Neural Network

为实现机组经济性能在线诊断,将对向传播神经网络方法引入汽轮机排汽焓预测计算。该预测方法很好地建立了汽轮机排汽焓特性与相关运行参数之间的复杂关系模型,并以某电厂300MW汽轮机组末级抽汽及排汽焓值为例进行了在线计算,实例结果表明:该方法能够准确地预测汽轮机排汽焓值,同时具有训练速度快、结构简单、精度高等特点,是一种行之有效的预测方法。

In order to diagnose the economic performance of unit online, a new algorithm to predict the exhaust enthalpy in the steam turbine online based on Counter Propagation Neural Network is introduced in this paper. The predicting method can establish the complicated relation model between the steam turbine exhaust enthalpy and the relative operating parameters. The enthalpy of both the last stage extraction steam and the exhaust is online calculated for a 300MW set. The example shows that this method can accurately predict the steam turbine exhaust enthalpy and has the training quick, simple and accurate features. This is an effective and feasible predicting method.