Gas classification and fault diagnosis of the gas sensor in the gas monitoring system using neural networks

In this paper we proposed a method of fault diagnosis and gas classification for tin oxide gas sensors using resistance and sensitivity sets and ART2 NN (adaptive resonance theory 2 neural network) with uneven vigilance parameters. In this method two ART2 NN modules are used for gas classification and fault isolation. The sensor features for diagnosis were sensor resistance and gas sensitivity sets and the features were manipulated by ART2 NN modules. We diagnosed tin oxide gas sensors upon exposure to oil vapor, silicon vapor, and high humidity. The performances were finally evaluated with hydrogen sulfide (H₂S). Proposed method proves to be helpful to diagnose a fault and classify gas concentration in gas monitoring system.