Thermodynamic properties of refrigerants using artificial neural networks

• Artificial Neural Networks (ANNs) can be used to predict thermodynamic properties of refrigerants.
• ANNs can replace Equations of State with acceptable performance.
• ANNs could be used to program devices to achieve real-time optimization.
• Refrigeration process can be optimized using ANNs.

The application of Artificial Neural Networks (ANNs) for prediction of thermodynamic properties of refrigerants in vapor–liquid equilibrium is the scope of this article. It is very important to find new ways to calculate thermodynamic properties of new refrigerants to simplify equipment operation and design. ANNs are capable of learning the complex relationships between input and output data, therefore they can be a good replacement of the commonly used Equations of State (EoS) for thermodynamic properties prediction. In this work multilayer perceptron ANNs with back-propagation algorithm were employed to obtain accurate thermodynamic properties prediction models. No EoS were needed so far. ANNs show their ability to accurately predict properties of refrigerants opening a promissory way to process optimization and construction of intelligent devices, impacting in both cost and energy savings.