<![CDATA[Refrigerant leaks are one of the main causes of decreased cooling system efficiency and increased greenhouse gas emissions. This study aims to develop a low-cost sensor-based refrigerant leak early detection system and a machine learning classification algorithm to improve diagnostic accuracy in small-scale cooling systems. Data were obtained through pressure, temperature, electric current, and humidity measurements using analog-digital sensors such as MQ-135, DS18B20, and ACS712. The machine learning model was tested with the K-Nearest Neighbors (KNN), Random Forest, and Support Vector Machine (SVM) algorithms to classify normal system conditions, light leaks, and heavy leaks. The test results showed that the Random Forest model provided the highest accuracy of 96.7%, with a detection response time of <2 seconds. This system has proven to be efficient and economical, potentially applicable to household and small industrial cooling systems.]]>
Copyrights © 2026
