论文部分内容阅读
将人工神经网络(ANN)应用于非连续螺旋折流板换热器的壳程换热和流阻分析。中试试验研究了具有3个螺旋角和2种管型的换热器。作为人工神经网络最常用的一种类型,将多层感知器神经网络(MLP)应用于本研究,使用一定的实验数据进行网络训练及预测。应用遗传算法(GA)对MLP的初始权值和阈值进行优化,预测结果精确。通过比较不同网络结构的预测误差来选择最适宜的网络结构为9-7-5-2。和关联结果比较可知MLP-GA网络对于换热器性能预测更加适合。此外,当使用MLP-GA方法在训练数据范围以外对壳程换热系数和压降进行预测时,网络预测结果和实验结果吻合程度也较高。因此,MLP-GA混合算法能够用来预测螺旋折流板管壳式换热器的传热和水力学性能。
Application of Artificial Neural Network (ANN) to Shell Side Heat Exchange and Flow Resistance Analysis of Discontinuous Helical Baffled Heat Exchanger. Pilot test of the three helix angle and two kinds of tube heat exchanger. As the most commonly used type of artificial neural network, MLN (Multilayer Perceptron Neural Network) is applied in this study, and some experimental data are used to train and predict the network. The genetic algorithm (GA) is used to optimize the initial weight and threshold of MLP, and the prediction result is accurate. By comparing the prediction errors of different network structures, the most suitable network structure is selected as 9-7-5-2. Compared with the correlation results, MLP-GA network is more suitable for heat exchanger performance prediction. In addition, when using the MLP-GA method to predict the shell side heat transfer coefficient and pressure drop outside the training data range, the network prediction results are in good agreement with the experimental results. Therefore, the MLP-GA hybrid algorithm can be used to predict the heat transfer and hydraulics of spiral baffled shell and tube heat exchangers.