基于人工神经网络的滴膜共存冷凝传热模型的研究

INVESTIGATION OF DROPWISE AND FILMWISE COEXISTING CONDENSATION HEAT TRANSFER PREDICTION MODEL USING ARTIFICIAL NEURAL NETWORKS

为了研究影响滴膜共存冷凝传热特性的因素,如滴膜区间面积比、滴膜相对位置、表面分割方式,表面过冷度等对冷凝传热的特性共同作用,本文应用人工神经网络技术,建立表面分割数、滴膜区面积比、凝液环数、表面过冷度与强化传热比之间的综合评价预测模型。结果表明,基于Matlab语言的人工神经网络模型具有较高的预测准确率及泛化能力,能够很好的评价和预测不同条件下的冷凝传热特性。

The purpose of this paper is to study the effect on heat transfer characteristics in dropwise and filmwise coexisting Condensation, such as the area ratio, different relative positions and the surface division patterns of the dropwise-filmwise region, the surface subcooling and the interaction of these factors. The prediction modei corresponding to the relative factors of heat transfer enhancement in-cluding surface division numbers, area ratio of dropwise and filmwise regions, condensate ring numbers, and ...