Open AccessSupervised machine learning applied to gas leak detection in air conditioner cooling system
Author(s) -
Sofia Moreira,
Vaibhav Shah,
Leonilde Varela,
Ana Paula Monteiro,
Goran D. Putnik
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1174/1/012008
Subject(s) - refrigerant , leakage (economics) , air conditioning , leak , leak detection , machine learning , computer science , artificial intelligence , process engineering , automotive engineering , engineering , environmental science , gas compressor , mechanical engineering , environmental engineering , economics , macroeconomics
This paper aims to present a concept test for an alternative refrigerant gas leak detection method, to be used in air conditioning manufacturing processes, in order to increase confidence in retaining products with gas leak in the factory, minimizing human interference in the test. To analyse the proposed solution, experimentation cycles were conducted, involving variables of industrial environment, product, and a thermographic camera with infrared technology, responsible for collecting the thermal image of the leak study area. Supervised machine learning method was used to train algorithms on temperature dataset to classify an area either as “Gas leakage” or “Normal”. The regression logistic algorithm had the best performance in the predictions, showing that it is possible to detect “Gas leakage” area in automatic decision-making in an industrial environmental.