基于Elman神经网络的汽轮机排汽焓在线预测计算  

Online Predicting and Calculating of Steam Turbine Exhaust Enthalpy Based on Elman Neural Network

作　　者：宫唤春[1] 吴义虎[1] 

机构地区：[1]长沙理工大学能源与动力工程学院,长沙410076

出　　处：《热力透平》2007年第4期241-244,共4页Thermal Turbine

基　　金：湖南省自然科学基金杰出青年项目(01jzy2102);湖南科技大学机械设备健康维护省重点实验室开放基金资助项目

摘　　要：为实现机组经济性能在线诊断,将Elman神经网络方法引入汽轮机排汽焓在线预测计算。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,实例表明:该方法能够准确地在线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 realize online diagnosis for the economic performance of unit, a new algorithm of online predicting for the exhaust enthalpy in the steam turbine based on Elman neural network is introduced in this paper. This online predicting method establishes the complicated relation model between the steam turbine exhaust enthalpy and the relative operating parameters. The online calculation for enthalpy of the last stage steam extraction and the exhaust of a 300MW set is made and the example shows that this method can accurately predict the steam turbine exhaust enthalpy and the model is simple,＇accurate and convergence, proving to be an effective and feasible predicting method.

关 键 词：汽轮机 排汽焓 ELMAN神经网络 在线预测 

分 类 号：TK226[动力工程及工程热物理—动力机械及工程]