A new approach for the globally optimal design of air coolers is presented. First, by coupling the air cooler geometric options with commercially available fans, the traditional mismatch emerging from designing the air cooler first and selecting a commercial fan later is removed. Second, the method departs from the traditional LMTD and ε-NTU methods, adopting a model composed of a differential–algebraic system of equations (DAE system) for the air cooler simulation, which is discretized to consider properties variable with temperature. The resultant optimization problem is solved using Set Trimming and Smart Enumeration, which can identify the global optimum through the simulation of only a small fraction of the search space. The comparison of the simulation results with different commercial software provided similar outputs. A comparison with design results obtained using LMTD solutions indicates that the LMTD models may overdesign/underdesign the unit

中文翻译：

---

关于空气冷却器与真实风扇耦合设计优化的新全局最优方法


提出了一种空气冷却器全局优化设计的新方法。首先，通过将空气冷却器的几何选项与市售风扇耦合，消除了先设计空气冷却器，然后再选择市售风扇的传统不匹配。其次，该方法与传统的 LMTD 和 ε-NTU 方法不同，采用由微分代数方程组 （DAE system） 组成的模型进行空气冷却器仿真，该模型被离散化以考虑随温度变化的属性。结果优化问题是使用 Set Trimming 和 Smart Enumeration 解决的，它们可以通过仅模拟一小部分搜索空间来识别全局最优值。将仿真结果与不同的商业软件进行比较，提供了相似的输出。与使用 LMTD 解决方案获得的设计结果进行比较表明，LMTD 模型可能会过度设计/欠估单元