Improved genetic algorithm-based prediction of a CO2 micro-channel gas-cooler against experimental data in automobile air conditioning system. (October 2019)

• Record Type: Journal Article
• Authors: Yang, Jingye; Yu, Binbin; Chen, Jiangping
• Is Part Of: International journal of refrigeration. Volume 106(2019)
• Page Start: 517
• Page End: 525
• Publication Date: 2019-10
• Subjects: CO2 -- Micro-channel gas cooler -- Distributed parameter model -- Genetic algorithm -- Interpolation method -- Heat transfer performance; CO2 -- Refroidisseur de gaz à micro-canaux -- Modèle à paramètre distribué -- Algorithme génétique -- Procédé d'interpolation -- Performance du transfert de chaleur; Refrigeration and refrigerating machinery -- Periodicals; 621.56
• DOI: 10.1016/j.ijrefrig.2019.05.017
• Languages: English
• ISSNs: 0140-7007

Abstract

Highlights: Experimental investigation of the micro-channel gas cooler was implemented. Performance prediction of a micro-channel gas cooler is conducted based on an improved genetic algorithm. Improvements in Tabular Taylor series expansion (TTSE) method are proposed. Abstract: For the prevailing CO2 micro-channel gas-cooler (MCGC) mathematical modeling research, a big challenge for an efficient numerical model is the trade-off between fast calculation and high-precision considering the specific properties of CO2 . The main purpose of this work is to propose a regression method with multiple variables for CO2 MCGC model based on the combination of Distributed Parameter Model (DPM) and Genetic algorithm (GA). The GA based prediction model of MCGC is established, which takes correlation coefficients and mean squared error into consideration. A test rig of CO2 MCGC is developed, and the experimental data used to validate the MCGC model are collected. The fixed-point iteration algorithm is utilized to reduce the complexity of simulation to accelerate the iteration speed. Moreover, an improved Taylor series expansion-based method for fast calculation of working fluid properties was proposed. The model precision was analyzed by comparing the prediction results with conducted experimental data. The improved CO2 -to-air MCGC model enables to predict refrigerant outlet temperature and pressure drop within a maximum deviation of 1.2 °C and 2 kPa against the experimental data.