Two-phase ejectors are highly useful devices because they can recover and use the energy lost in the expansion process of the refrigeration cycle. However, in a small-sized household refrigeration cycle, a modified ejector cycle is required due to the low operating pressure and mass flow rate. In this study, a dual evaporator ejector cycle (DEEC), which is advantageous for household refrigeration, was applied. In the DEEC, unlike the conventional ejector cycle, the motive nozzle exit position (NXP) considerably affects the cycle performance. The objective of this study was to optimize the NXP of a DEEC for a household refrigeration cycle with low-pressure refrigerants via an artificial neural network (ANN) model. Using the developed model, the DEEC pressure lifting ratio and coefficient of performance (COP) were analyzed under various operating and ejector geometry conditions. Moreover, the optimal NXP of a small-sized household DEEC was proposed using the developed correlation to achieve maximum performance under different operating conditions. The COP of the DEEC with the optimized NXP is 2.3% and 8.4% higher than those of the DEEC with the conventional NXP and the baseline cycle, respectively. These results validate the ANN model used for optimization and serve as design guidelines for obtaining optimized NXPs and cycle performance with increased energy efficiency.

中文翻译：

---

使用人工神经网络优化改进的两相喷射器膨胀家用制冷循环中的动力喷嘴位置

两相喷射器是非常有用的装置，因为它们可以回收和利用制冷循环膨胀过程中损失的能量。然而，在小型家用制冷循环中，由于工作压力和质量流量较低，需要改进喷射器循环。在这项研究中，应用了有利于家庭制冷的双蒸发器喷射循环（DEEC）。在 DEEC 中，与传统的喷射器循环不同，动力喷嘴出口位置 (NXP) 显着影响循环性能。本研究的目的是通过人工神经网络 (ANN) 模型优化采用低压制冷剂的家用制冷循环 DEEC 的 NXP。使用开发的模型，在各种操作和喷射器几何条件下分析了 DEEC 压力提升比和性能系数 (COP)。此外，利用所开发的相关性提出了小型家用 DEEC 的最佳 NXP，以在不同操作条件下实现最大性能。采用优化 NXP 的 DEEC 的 COP 分别比采用传统 NXP 和基准循环的 DEEC 高 2.3% 和 8.4%。这些结果验证了用于优化的 ANN 模型，并可作为获得优化的 NXP 和循环性能以及提高能源效率的设计指南。