Equipment health state intelligent detection method and system based on sound wave signals

Abstract

The invention belongs to the field of sound wave signal processing, and particularly relates to an equipment health state intelligent detection method and system based on sound wave signals. The method is based on an equipment fault sound wave signal identification technology of deep learning. Equipment sound is collected according to a unified voiceprint acquisition device and an acquisition scheme, equipment sound signals are preprocessed, a CNN model and an RNN model are trained, the CNN model and the RNN model are integrated, and the model is solidified and stored. During use, a to-be-tested signal is collected firstly, then the to-be-tested signal is input into the trained model for prediction, and a prediction result is displayed in real time. By using the method and the system provided by the invention, the adaptive capacity and the judgment accuracy of the model are enhanced, the result can be displayed in real time, and the fault detection efficiency is improved.

Description

technical field [0001] The invention belongs to the field of sound wave signal detection systems, and in particular relates to a method and system for intelligent detection of equipment health status based on sound wave signals. Background technique [0002] There is a niche research direction in the field of acoustic signals—sound classification. An Israeli start-up company called "3DSignals" has a patent on acoustic anomaly detection for equipment in the field of electrification, such as rotating equipment. These devices include, for example, electric motors, pumps, turbines, gearboxes and generators. [0003] In the deep learning domain model, the "black box" problem brought by the acoustic signal makes it difficult for us to determine and understand the reason why a certain segment of the acoustic signal is marked (normal, fault type); at the same time, the ubiquitous noise in production and life also It suppresses the feature signals that interfere with us, thus limit...

AI Technical Summary

Problems solved by technology

[0003] In the deep learning domain model, the "black box" problem brought by the acoustic signal makes it difficult for us to determine and understand the reason why a certain segment of the acoustic signal is marked (normal, fault type); at the same time, the ubiquitous noise in production and life also It suppresses the feature signals that interfere with us, thus limiting the ability of the model to obtain features from the sound wave signal; although there are currently some technologies that use deep learning for sound classification, the sound recognition ability for equipment failures is weak, and cannot be fast and accurate. Accurately identify equipment failures

In addition, in the field of equipment maintenance, at present, it mainly relies on manual inspection of equipment one by one, lack of effective rapid inspection and feedback mechanism, and the inspection efficiency is low

Images