Power equipment health monitoring method based on voiceprint features

• Main Authors: HUANG WENLI, TONG XIN, CHEN YUE, GU HAISHENG, ZHANG YANG, HU YAN, GUAN SHAOQING
• Format: Patent
• Language: Chinese; English
• Published: 26.01.2024
• Subjects: ACOUSTICS; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

The invention relates to the technical field of power equipment health monitoring, and particularly discloses a power equipment health monitoring method based on voiceprint characteristics, and the method comprises the steps of audio data preprocessing, voiceprint characteristic information evaluation and health state grade feedback. The audio preprocessing feature value of the power equipment is judged, the audio data and the frame data of the power equipment are judged in sequence, factors interfering voiceprint features in the audio data can be reduced to a great extent, and meanwhile, the health state level of the power equipment is matched and health feedback is performed, so that the voiceprint feature matching accuracy is improved. The feature anomaly degree value of the power equipment is imported into the analysis model, so that data support can be provided for subsequent health feedback, the downtime and production loss of the power equipment can be reduced, and the manual inspection and maintenance