Human Communication Technology

9Real Time Hazardous Gas Classification and Management System Using Artificial Neural Networks

R. Anitha*, S. Anusooya, V. Jean Shilpa and Mohamed Hishaam

B.S. Abdur Rahman Crescent Institute of Science & Technology, India

Abstract

Generally, peoples working in coal mines, gas industries and sewage cleaning are prone to high health risk. The gases released in coal mines, oil & gas industries and sewage areas are to be monitored and the different gas concentration that has been released out have to be identified. Hence, a gas detection system is developed to detect hazardous gases. Hazardous gases are harmful to humans. A metal-oxide gas sensors arrays of 4 different MQ series gas sensors are used to recognize non-combustible gas such as ammonia as well as combustible gaseous like LPG, ethanol and to a specific concentration of the respective gas in the environment. The development of systems involved 3 steps i.e. Data Set preparation, Artificial Neural Network Model Creation and Training the network using the data set. An Artificial electronic nose system is developed to classify and to measure the concentration of single or mixed gas in the environment. The electronic nose composed of two hidden layered neural networks with backpropagation has been developed to classify the gas and regression model-based neural network to find the concentration. An industrial standard IoT device, National Instruments-Compact Reconfigurable Input-Output (NI-CRIO) is used to process the incoming ...