Diagnosis and Detection Method of Critical Equipment Failure Based on Electronic Nose Technology

Abstract

In the manufacturing process of modern space and aeronautics industry, chemical industry, industry and manufacturing industry, the safe and efficient operation of equipment has played an increasingly important role. In order to realize the early diagnosis of critical equipment failure, this paper uses the e-nose gas sensor technology to qualitatively and quantitatively analyze the odor volatilized by the oil generated by equipment failure and the odor emitted by the wire during the heating process. The experimental results show that the linear discriminant analysis and artificial neural network method can be applied to the e-nose gas sensor array to perform accurate qualitative separation of the odor. The curve change of the response stability amplitude is in accordance with the related law of the gas sensor output stability and concentration change. With the increase of the gas concentration, the larger the stability amplitude K, the smaller the response time constants T_P1 and T_P2, and the higher the confidence. By establishing a mixed gas response model, it is conductive to the separation of the gas sensor array signals of the mixed gas.