Online detection technology of transformer oil dissolved gas can be used to analyze insulation status of transformer effectively

and the gas sensor in online detection is crucial to this technology. Using the intersection sensitivity for semiconductor gas sensors

this paper presents a smart sensor integrating gas sensor array with artificial neural network technology. The author employs gas sensor array consisting of six semiconductor sensors and BP neural network to do pattern recognition. Lots of experiments prove that the smart sensor can improve resolution and sensitivity of online detection on six gases of H2

CO

CH4

C2H4

C2H2 and C2H6 .