Eventos Anais de eventos
COBEM 2023
27th International Congress of Mechanical Engineering
INTELLIGENT REGRESSION MODELING FOR PERFORMANCE PREDICTION OF A VAPOR COMPRESSION REFRIGERATION PROTOTYPE USING MACHINE LEARNING TECHNIQUES
Submission Author:
Evandro Pedro Alves de Mendonça , PE , Brazil
Co-Authors:
Paulo Silva, Evandro Pedro Alves de Mendonça, Sérgio Franco, Alvaro Antonio Ochoa Villa, Gustavo de Novaes Pires Leite, JOSÉ ÂNGELO PEIXOTO DA COSTA, kilvio ferraz
Presenter: Felipe Roque de Albuquerque Neto
doi://10.26678/ABCM.COBEM2023.COB2023-0986
Abstract
Vapor compression refrigeration systems are essential for cooling and refrigeration in various settings, but they require maintenance and optimization for optimal performance. Analyzing data on their behavior can provide valuable insights for improving efficiency and lifespan. This study focuses on analyzing the energy performance of a refrigeration prototype using the R404A fluid and tube-in-tube evaporator. The prototype consists of two circuits: one with the R404A fluid refrigeration system and the other with ethylene glycol and thermal load provided by electrical resistances. The system incorporates an integrated supervisory system for data acquisition and control. The collected data includes temperatures, pressures, power, voltage, current, and consumed energy. The study aims to predict the system's performance using intelligent regression models based on linear regression, decision tree, random forest, and artificial neural network techniques. The dataset is divided into training, validation, and test sets, and performance parameters such as determination coefficient, root mean square error and mean absolute error are used to evaluate the models. The results show that the proposed approach accurately estimates the transient behavior of the refrigeration system and can be used to optimize its performance, making it a valuable tool for dynamic analysis of vapor compression equipment operating at full or partial loads.
Keywords
vapor-compression refrigeration, machine learning, Regression Modeling, Transient Behavior, Performance Prediction
