Switch cabinet monitoring and fault diagnosis method based on voiceprint recognition technology

The invention provides a switch cabinet monitoring and fault diagnosis method based on a voiceprint recognition technology, and relates to the technical field of switch cabinet monitoring and fault diagnosis, and the method comprises a terminal layer, a cloud platform layer and an edge layer. The terminal layer is provided with a voiceprint acquisition device and is used for realizing high-precision acquisition of operation voiceprints of monitoring equipment; the edge layer is used for carrying out data acquisition, cleaning and feature extraction on various equipment vibration and sound signals and is used for constructing a model; a voiceprint sensor is used for collecting sound signals of different components of the switch cabinet, voiceprint features of the different components under different working conditions are extracted to identify and position faults, meanwhile, signal collection of a vibration sensor is adopted, and a total voiceprint and vibration multi-mode feature spectrum of equipment under v