RBF与BP神经网络对逆流鼓泡板片蒸发式冷凝器传热性能预测对比

Comparison of RBF and BP Neural Network Onheat Transfer Performance Prediction of Evaporative Condenser with Countercurrent Bubbling Plate

鼓泡板片蒸发式冷凝器是一种较为新型的蒸发式冷凝器,按喷淋水与空气流向不同分为逆流、横流、混流等形式,其热质传递过程较为复杂。依据逆流鼓泡板片性能测试实验,以板间距、喷淋水流量、截面风速、空气进口干湿球温度、板片壁面平均温度为输入变量,以复合传热系数为输出变量,通过RBF和BP两种神经网络模型,对鼓泡板片外空气侧的复合传热系数进行预测。比较这两种神经网络模型预测值和真实值的主要评价指标(决定系数R2、平均相对误差MRE、均方根误差RMSE),结果表明RBF神经网络预测性能优于BP神经网络,更适合用于鼓泡板片蒸发式冷凝器的传热性能预测,将有助于对鼓泡板片蒸发式冷凝器的研究和产品开发。

Bubbling plate evaporative condenser is a relatively new type of evaporative condenser. According to the different flow directions of spray water and air, it can be divided into counter-current, cross-flow and mixed flow forms, and the heat and mass transfer process is relatively complex. Countercurrent bubble plate based on performance test experiment, plate spacing, spray water flow, cross section of dry wet bulb temperature, wind speed, the air inlet plate wall average temperature as input variables, with compound heat transfer coefficient for the output variable, by two RBF and BP neural network model, the bubble plate predict compound heat transfer coefficient of the outside air side. The main evaluation indexes (determination coefficient R2, average relative error MRE, root mean square error RMSE) of the predicted and real values of the two neural net-work models were compared. The results show that RBF neural network has better prediction performance than BP neural network, and it is more suitable for the prediction of heat transfer performance of evaporative condenser with bubbling-plate plate. It will be helpful to the research and product development of evaporative bubbling plate condenser.