论文部分内容阅读
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 ,并获得该特征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 R1 2 ,检验样本包括 R1 2、R2 2、R1 34a和 R60 0 a等多种工质 .在常用制冷空调工况范围内 ,该简化模型与分布参数模型相比 ,平均偏差为 0 .3%,只有以 R60 0 a为工质时的最大偏差超过 5%,计算速度提高 1个数量级
Using the modeling principle of phase-separation concentration parameters, the weight factor of average specific volume is proposed as a simplified characteristic parameter of two-phase region, and the dimensionless influence parameters of the characteristic parameters are obtained.An artificial neural network method is used to establish the characteristic parameters and their influence parameters Between the nonlinear mapping, artificial neural network learning samples using the working fluid R1 2, the test sample includes R1 2, R2 2, R1 34a and R60 0 a and other working fluid in the common refrigeration and air conditioning operating conditions, the simplified The average deviation of model and distributed parameter model is 0.3%. Only when the maximum deviation of R60 0 a is more than 5%, the calculation speed is increased by 1 order of magnitude