An integral identification method of characteristic parameters and optimization of parallel connection heat transfer systems based on the power flow method

作者：Zhang, Meng-Qi; Chen, Qun^*; Shao, Wei; Chen, Xi; Hao, Jun-Hong

来源：Applied Thermal Engineering, 2018, 143: 1057-1067.

DOI：10.1016/j.applthermaleng.2018.08.016

• Heat transfer system
• Integral identification and optimization
• Power flow method
• Characteristic parameters
• Thermal resistance

Optimization of heat transfer systems benefits energy conservation, but the conventional fixed node-temperature operation strategy hardly offers the optimal performance under changing working conditions. This paper utilizes the power flow method to construct the integral heat transfer model and combines the overall flow resistance model to set up the integral constraints of heat transfer systems with independent variables. On this basis, we develop an integral identification method to identify the free degree of systems and the characteristic parameters of each component, and propose the corresponding optimization method, which is universal and has been applied to many kinds of heat transfer systems. For the convenience and economy of experiment, a parallel connected counter-flow heat exchanger network is studied to validate the identification method and the optimization method. The experimental results show that the identified characteristic parameters by the newly proposed method have enough accuracy for practical use; while the optimization method offers the optimal operating parameters with the least power consumption under given conditions. When the operating frequency of the pump in a cold-water loop deviates the optimal one of 10%, the total pumping power consumption of the system increases up to 18.2%. Meanwhile, a bigger deviation will lead to a larger pumping power consumption. Besides, the node temperatures in the optimal cases vary with the heat loads, which illustrates the limitation of fixed node-temperature operation strategy.

• 出版日期2018-10
• 单位清华大学