Arduino-Based Intelligent Optical Gas Leakage Detection System for Industrial Safety Using Machine Learning
Аннотация
Following this paper, the design, implementation, and evaluation of an intelligent optical gas leakage detection system based on machine learning and an Arduino platform to enhance the detection accuracy and minimize the false alarms in industry setups have been reported. The system proposed combines an optical gas sensor array, Arduino microcontroller to acquire a signal, edge pre-processing and a lightweight machine learning classifier (tested: Random Forest and SVM) to classify the target hazardous gases, common interferents and environmental noise. The models are trained and tested on a dataset gathered under controlled laboratory experiments and industrial-like settings. Experimental findings indicate high detection accuracy, quick response time and low power consumption that can be used in distributed industrial applications. The system is inexpensive, modular and can provide real time alerts through serial, visual (LED) and networked (optional Wi-Fi/LoRa) channels.