课题组李彦澎博士的研究论文Performance evaluation and optimization of the cascade refrigeration system based on the digital twin model被期刊《Applied Thermal Engineering》收录，论文的创新点在于提出了一种复叠制冷系统的数字孪生模型，有效提高预测精度7.42%~23.62%，节能率高达13.1%。论文摘要如下：

Cascade refrigeration systems (CRSs) have become the priority choice in low-temperature applications, causing severe energy waste due to the lack of practical control strategies. Currently, refrigeration simulations are mainly based on the simplified thermodynamic or Artificial Intelligence (Al) model. However, the former is insufficiently accurate, while the latter is of unclear thermodynamic significance. By connecting the physical and digital space, the digital twin technology provides exciting opportunities for combining thermodynamic and AI methods. In this paper, an integrated digital twin model for the CRS is innovatively proposed based on the modified semi-empirical model of the twin-screw refrigeration compressor. As a major update, the identification and update method of characteristic parameters is developed based on real-time
operating data to ensure performance consistency between digital and physical twins. By comparing with the simplified model, results show that the digital twin model can improve prediction accuracy by 7.42%-23.62%. Founded on the proposed model, the thermodynamic performance of the CRS is evaluated. To facilitate on-site applications, the definition of flat intermediate temperature is given, which is fitted to a control correlation with several easily observed variables as the inputs. Taking a food factory as a case study, simulation results reveal that the optimum intermediate temperature should be significantly lower than the original, and the coefficient of performance (COP)has an average improvement potential of 9.1%.By deploying the optimum strategy under the on-site condition, a superior energy-saving rate of up to 13.1% is attained according to test reports.