电站锅炉运行性能综合预测模型

采用人工神经网络提出了一种预测电站锅炉运行性能的综合模型方法。在对某台200MW电站燃煤锅炉进行多工况热态正交试验的基础上，采用BP网络建立了预测锅炉运行性能包括排烟温度、飞灰含碳以及氮氧化物（NOx）排放的综合模型。由于综合模型采用了锅炉排烟温度预测模型的输出为中间变量，并作为另外两个模型的输入，因而大大提高了整个模型的预测精度。数据测试和定量分析表明：该模型是有效的，具有较高的精度。该模型目前已经成功应用于现场实际的锅炉运行优化监督控制，有效地提高了锅炉的运行性能。

An Integrative Model for Predicting the Performance of Utility Boilers

An integrative modeling method,using artificial neural networks,for predicting the performance of boilers is being proposed.On the basis of orthogonal tests of a 200MW coal firing utility boiler,under various modes of operation,an integrative BP model has been built for predicting a boiler's performance, including the exhaust gas temperature,carbon content of the fly ash and NO_x emission.Because the boiler's exhaust gas temperature is taken as the intermediate variable,to be used as the input of two other models,the predictive precision of the whole model is markedly raised.Data logged tests and quantitative analyses show that the proposed model is effective and has a relatively high accuracy.It is already being successfully used in practical on site application,for monitoring optimized operation of boilers,and is thus effectively raising the boilers' operational performance.Figs 4,tables 6 and refs 9.