University Of Tasmania

File(s) under permanent embargo

Investigations on optimal discharge pressure in CO2 heat pumps using the GMDH and PSO-BP type neural network—Part A: Theoretical modeling

journal contribution
posted on 2023-05-20, 06:16 authored by Yin, X, Cao, F, Wang, J, Li, M, Xiaolin WangXiaolin Wang
Discharge pressure is an important factor that heavily affects the system COP in the transcritical CO2 heat pump. In most cases, it is commonly confirmed by the empirical correlations or calculated by the mathematical model according to a single operation condition, thus leading to the prediction error or lengthy time. In this paper, a novel model using the statistical method known as the group method of data handling-type (GMDH) and PSO-BP-type (Particle-Swarm-Optimization and Back-Propagation) neural network was developed to predict the optimal discharge pressure. The relevance of all the parameters to the optimal discharge pressure was investigated orderly. Results showed that the new model had the highest accuracy compared to the current correlations. The relative error was around 1.6% while the error of traditional methods ranged from 11.1% to 44.9%. Therefore, the CO2 heat pump could work better in the optimal COP operation condition with the novel statistical model.


Publication title

International Journal of Refrigeration








School of Engineering


Elsevier Sci Ltd

Place of publication

The Boulevard, Langford Lane, Kidlington, Oxford, England, Oxon, Ox5 1Gb

Rights statement

Copyright 2019 Elsevier Ltd and IIR

Repository Status

  • Restricted

Socio-economic Objectives

Residential energy efficiency