Abstract
A Fast and Robust Gas Recognition Algorithm Based on Hybrid Convolutional and Recurrent Neural Network
Highlights
Gas recognition is greatly needed in a wide range of internet of things (IoT) applications, including petrochemical [1], [2], electric power [3], natural gas industry [4], indoor gaseous pollutants detection [5], [6] and gas metering systems along gas pipes [7]
Limited by the selectivity of individual gas sensing device in the first layer of the smart gas grid system of a classical IoT system [7], gas sensors are always implemented in an array with signal processing circuitry and recognition algorithm to form a bionic system, namely ‘‘electronic
DATASET DESCRIPTION To evaluate the performance of our proposed model for fast gas classification within a limited response time, we conducted extensive experiments based on a public dataset and the multi-gas testing system is detailed in [23]
Summary
Gas recognition is greatly needed in a wide range of internet of things (IoT) applications, including petrochemical [1], [2], electric power [3], natural gas industry [4], indoor gaseous pollutants detection [5], [6] and gas metering systems along gas pipes [7]. The overall gas recognition process is greatly retarded, due to the inherent slow reaction between the odor molecule and the surface of the gas sensing element, which typically takes tens of seconds to minutes before the response reaches a saturation status [11]. It is well-known that fast recognition of the combustion and toxic gases are critical for real-life applications [12]. By taking full use of the initial part of the response curve, the needed volume of the target gas can be significantly reduced, which is highly beneficial for various applications with very limited sources of target
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
