中文摘要：

多联机VRF系统是目前应用较为广泛的空调系统之一,但目前该系统仿真方法主要集中于单一仿真器的建模研究。为了在有限的计算成本下实现同时模拟制冷系统特性与空调系统运行性能,提出了基于部件神经网络建立VRF系统仿真模型,并通过与Energy Plus负荷计算模块的数据交换,实现了可以在建筑物实时负荷变化基础上进行VRF系统运行策略研究的协同仿真方法,从而为该类系统控制及运行策略的分析研究提供了新的途径。

英文摘要：

Multi-split variable-refrigerant-flow （VRF） air-conditioning system has been widely used today. The simulation methods of VRF system were mainly developed using single simulator. To simultaneously conduct the refrigerant system modeling and the air-conditioning system simulation, a new co-simulation method was developed. The new method integrated neural networks for refrigerant component modeling with EnergyPlus for building dynamic load calculation. It carried out a new approach for VRF system control and operating strategy study based on the building dynamic load profile.