This paper proposes a high-precision refrigerant charge diagnostic method based on active control， which effectively addresses the shortcomings of traditional passive diagnostic methods， such as low sensitivity to small deviations and susceptibility to interference from system operating conditions. The method actively anchors the core thermodynamic boundary conditions of the system and， based on the principle of mass conservation and heat transfer mechanisms， directionally converts variations in refrigerant charge into changes in liquid-phase retention within the condenser. The condenser outlet subcooling degree is thereby employed as the core characteristic parameter for quantitative diagnosis， and a standardized four-stage diagnostic procedure， including startup， coarse regulation， fine regulation， and measurement is established， with quantitative stability criteria. Cross-validation experiments covering six operating conditions and outdoor temperatures from 20 ℃ to 45 ℃ were conducted in an enthalpy difference laboratory using two multi-split units with nominal cooling capacities of 16 kW and 18 kW. The results demonstrate that， within the charge ratio range of 85%-105%， the subcooling degree exhibits a strictly monotonic increasing trend with respect to refrigerant charge. The subcooling difference between 100% and 90% charge levels remains above 6 °C across all test conditions， and this relationship remains stable across different unit models and outdoor temperatures. The diagnostic model constructed from this relationship can accurately identify refrigerant charge deviations as low as 5%， providing a reliable technical solution for online operational diagnostics and early fault warning in multi-split air-conditioning systems.
A one-dimensional real-fluid delayed equilibrium model （DEM） was developed to improve the prediction of mass flow rate， axial pressure， and nonequilibrium flashing in cryogenic throttling flows. The model was validated against the data from Venturis， converging-diverging nozzles， thick orifices， and a self-built orifice test rig. Compared with the homogeneous equilibrium model （HEM）， DEM reduced the root mean squared error of pressure （RMSE） at the Venturi pressure taps from 0.616 to 0.232 kPa. For the nozzle cases， the mean absolute percentage error （MAPE） values of axial pressure and mass flow rate were 5.10% and 6.94%， respectively， while the mass-flow MAPE and RMSE for thick orifices were 11.81% and 14.40%， respectively. The results show that nonequilibrium effects depend strongly on the throttling geometry and flow time scale and are more pronounced in sharp-edged short orifices than in smoothly varying nozzles. Further liquid-nitrogen experiments with 0.5-2.0 mm orifices showed that the mass flow rate increased with the inlet pressure， inlet subcooling， and orifice diameter， approximately following m˙∝d1.8. These results indicate that cryogenic throttling is jointly governed by the inlet thermodynamic state， geometric scale， and nonequilibrium phase change， and the proposed model can support the flow prediction， pressure evaluation， and design of cryogenic throttling components.
A field test was conducted in an office building in Chengdu to evaluate the performance of a fresh-air system integrating electrical energy storage， dew-point evaporative cooling， and mechanical refrigeration under humid summer conditions. A total of 6 589 stable data points over 11 working days were selected based on power， air velocity， operating duration， and air-state variations. The dew-point evaporative cooler reduced the fresh-air temperature from 29.60 ℃ to 23.03 ℃， corresponding to dew-point and wet-bulb effectiveness values of 71.06% and 100.34%， respectively. The enthalpy differences across the evaporative cooling and mechanical cooling sections were 3.20 and 27.39 kJ/kg， respectively. The precooling load fraction was （10.39 ± 3.31）%， and the system coefficient of performance （COP） increased when this ratio exceeded 9.81%. The daily mean COP ranged from 3.16 to 3.49， while its variation after precooling coupling was （0.82 ± 4.17）%. The 30 kW·h battery energy storage unit shifted 11.70 kW·h of load per day， yielding estimated savings of 4.59-5.74 CNY. The evaporative cooling section primarily provided sensible precooling， whereas mechanical refrigeration handled most of the enthalpy load and dehumidification. Adjusting the operation of the evaporative cooling section based on outdoor air conditions can enhance the system's adaptability to operating conditions.
Faced with the heat dissipation challenges posed by the rapid increase in the computing power of data centers， traditional air cooling can no longer meet their cooling demands because of its limited heat dissipation density. Meanwhile， full immersion liquid cooling is difficult to adopt widely owing to coolant cost， ease of maintenance， and material compatibility. This paper reviews the progress in the application of liquid-air coupled cooling systems in high-power-density data centers and highlights that utilizing cold-plate liquid cooling to efficiently capture the chip heat can significantly reduce the auxiliary air cooling load. It confirms that through the reconstruction of rack airflow fields and optimization of low-flow air cooling， combined with intelligent coordinated control， the power usage effectiveness of the system can be optimized to below 1.15. In the future， the integration of liquid and air cooling systems sharing the same source and waste heat recovery technologies would drive data centers toward zero-carbon goals. This review aims to provide technical guidance for the construction of next-generation data centers that balance cooling efficiency and economic benefits.
We creatively proposed a heat pump system （PT/RC-HP） supported by photothermal and radiative cooling to increase the utilization rate of solar equipment and improve the energy efficiency of heat pumps. Using TRNSYS-MATLAB， we jointly simulated the system to assess its heating performance and created mathematical models of photothermal and radiative cooling. The effects of PT-RC area and thermal storage tank volume on system heating performance were systematically investigated， and typical meteorological year data for Changzhou， Jiangsu Province， China， were used to evaluate its long-term heating-season performance. The simulation results indicate that the optimal system configuration comprises a photothermal and radiative cooling collector with an area of 30 m2 coupled with a 2 m³ thermal storage tank. Under this configuration， the system cumulatively collected 3 256.89 kW·h of heat， achieving an average COP of 3.43 for the heat pump. This represents a 10% COP improvement relative to that of conventional air-source heat pumps， corresponding to an energy saving of 160.45 kW·h. The proposed system effectively improves the operational conditions of heat pumps during the heating season， demonstrating both significant energy-saving potential and robust operational stability. Thus， it offers a promising technical pathway for the integration and application of clean energy in building heating systems.
