基于平均比容人工神经网络辨识的绝热毛细管简化模型

采用分相集中参数的建模思路，提出将平均比容的权重因子作为两相区简化的特征参数，并获得该特参数的无量纲影响参数，采用人工神经网络方法建立特征参数与其影响参数之间的非线性映射，人工神经网络的学习样本采用工质Ｒ１２，检验样本包括Ｒ１２、Ｒ２２、Ｒ１３４ａ和Ｒ６００ａ等多种工质。

General Simple Model for Adiabatic Capillary Tubes Based on the Identification of Average Specific Volume Using Artificial Neural Network

A lumped parameter model simplified from a general distributed parameter model was established. The weighted factor of average specific volume in two phase region was presented as a characteristic variable. The correlative factors of the characteristic variable were analyzed and transformed into dimensionless form. The nonlinear mapping between the characteristic variable and its correlative factors was built by a forward neural network. The learning samples were generated with the working fluid R12. The checking samples covers the working fluids R12, R22, R134a and R600a. In the common range of refrigeration and air conditioning working conditions, the new model was compared with the general distributed parameter model. The average deviation falls into 0.3%, while the maximum deviations are not greater than 5% except R600a being used as the working fluid. The computation speed of the new model is one order of magnitude higher than that of the distributed parameter's one. Therefore, the new simple model for adiabatic capillary tube is of good theoretical precision and generalization. It is suitable for simulation and optimization of refrigeration systems.