Research on Neural Network Prediction of Optimal Exhaust Pressure of Heat Pump System

The widespread application of heat pump technology presents significant growth opportunities for the heat pump industry. As the demand for high temperature heat pumps in industry rises, compound high temperature heat pumps and transcritical heat pump systems are gaining increasing attention. Due to their critical operating characteristics, heat pump systems currently possess the optimal exhaust pressure in terms of energy efficiency. This paper uses experimental data as the basis for the calculation and fitting of the optimal pressure, and employs an artificial neuron network scheme to analyze the correlation between the influence of the system state parameters and the optimal exhaust pressure, in order to better solve the problem of maintaining the system in a high energy-efficient operating condition under variable operating conditions. The impact of genetic and particle swarm algorithms on the outcomes is also incorporated to the artificial neuron network. The experimental outcomes demonstrate that the particle swarm optimised neural network has superior generalisation properties.