Exergy approach in decision-based design of absorption refrigeration system using artificial intelligence and simulink

Exergy approach is essential for the design and possible operation of the absorption system with energy loss minimization. Operational limitations of absorption refrigeration system are due to Gibbs phase rules-base. Optimized operational decision rules are extracted from machine learning algorithm C4.5 which is suitable for different exergy evaluation methods and reference conditions. This study investigates proposed and investigated classification models for the same which are based on different thermodynamic features and operational design data. ANN model is used to predict the thermodynamic properties of the working fluids using improved thermodynamics properties of data patterns. Mathematical expressions are formulated from validated artificial neural network for predicting specific enthalpy and entropy of water–lithium bromide solution. Exergy design data are estimated on 356 thermodynamic design data pattern of absorption refrigeration system by MATLAB Simulators. Pearson’s correlation heatmap is used for extracting 14 thermodynamic features. 94.38% is the highest performance observed with 12 feature class six classification model.