Synthesizing a multi-level refrigeration system by considering the optimization of pressure levels results in a complex Mixed Integer Nonlinear Programming (MINLP) problem. Existing studies mostly adopted fixed pressure levels or optimized them in series of fixed alternative levels. In this study, a more comprehensive methodology was developed for the simultaneous synthesis of a refrigeration system and its heat exchanger network, wherein the pressure/temperature levels were optimized as continuous variables. A series configuration of connected sub-coolers and expansion valves was used in our superstructure to reduce the inlet temperatures of expansion valves, thus reduce the energy consumption. A Generalized Disjunctive Programming model was used and reformulated as a MINLP problem. Our methodology was validated by three case studies of an ethylene refrigeration system, with total shaft work saving of compressors reached 7.17%. Besides, refrigerant from refrigeration system, together with cold process streams, was verified to suitable cooling capacity for sub-coolers.

通过考虑压力水平的优化来合成多级制冷系统会导致复杂的混合整数非线性规划（MINLP）问题。现有研究大多采用固定压力水平或以一系列固定替代水平对其进行优化。在这项研究中，开发了一种用于制冷系统及其热交换器网络的同时综合的更全面的方法，其中压力/温度水平作为连续变量进行了优化。我们的上层建筑采用了一系列连接的过冷器和膨胀阀的配置，以降低膨胀阀的入口温度，从而降低能耗。使用了广义析取规划模型，并将其重新构造为MINLP问题。我们的方法通过乙烯制冷系统的三个案例研究得到验证，压缩机的总轴功节省达到7.17％。此外，还验证了来自制冷系统的制冷剂以及冷流程物流，是否适合过冷器。