神经网络在腐蚀污垢热阻预测中的应用

Application of Neural Network in Prediction Thermal Resistance of Corrosion Fouling

介绍了人工神经网络的理论基础、BP算法以及改进BP算法。并用改进的BP神经网络对换热设备表面形成的腐蚀污垢产生的热阻进行预测。结果表明:在相对误差15%之内,该模型能准确地预测换热设备表面腐蚀污垢产生的热阻。

This paper reviewed the basic principle of artificial neural network,BP algorithm and improved BP algorithm.Andin improved BP neural network we predicted the thermal resistance of corrosion fouling that was formed in heat exchanger surfaces.It indicated that during 15% of relative error,the prediction models could exactly predict thermal resistance ofcorrosion fouling thatwas formed in heat exchangersurfaces.