The traditional capillary radiant air conditioning system based on back propagation (BP) neural network optimization model of carbon dioxide refrigerant filling, using neural network algorithm to optimize the carbon dioxide refrigerant filling, there is a big error. Aiming at this problem, a carbon dioxide refrigerant charging and filling optimization model in capillary radiant air conditioning system based on genetic algorithm (GA) is constructed. Based on the idea of solving multiobjective optimization problem with genetic algorithm, an adiabatic capillary steady‐state distribution model is constructed to analyze the change of refrigerant flow rate and critical temperature in capillary tube. Corresponding data samples, the GA‐BP neural network model is calculated based on the sample data, and the optimization of carbon dioxide refrigerant filling capacity is realized through the interpolation and extrapolation test model of condenser heat transfer and refrigerant outlet temperature. The experimental results show that the model has a significant application in optimizing the charging capacity of carbon dioxide refrigerant. The error of optimizing the charging capacity of carbon dioxide is small, and the error range is between (−0.03, 0.03).

中文翻译：

---

基于遗传算法的毛细管辐射空调系统二氧化碳制冷剂充量优化

传统的毛细管辐射空调系统基于BP神经网络的二氧化碳制冷剂充注优化模型，采用神经网络算法对二氧化碳制冷剂充注进行优化，存在较大误差。针对该问题，建立了基于遗传算法的毛细管辐射空调系统二氧化碳制冷剂充装优化模型。基于遗传算法解决多目标优化问题的思想，建立了绝热毛细管稳态分布模型，分析了毛细管中制冷剂流量和临界温度的变化。对应的数据样本是根据样本数据计算出GA‐BP神经网络模型，通过冷凝器传热和制冷剂出口温度的内插和外推试验模型实现了二氧化碳制冷剂填充量的优化。实验结果表明，该模型在优化二氧化碳制冷剂充注量方面具有重要的应用价值。优化二氧化碳的充电容量的误差很小，误差范围在（-0.03，0.03）之间。