Industrial equipment general fault detection method and system based on sound signals
A fault detection and industrial equipment technology, applied in neural learning methods, measurement devices, measurement of ultrasonic/sonic/infrasonic waves, etc., can solve problems such as low fault detection accuracy, achieve fast and accurate fault detection, easy deployment, and scalability. strong effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0056] Such as Figure 1-4 As shown, Embodiment 1 of the present disclosure provides a general fault detection method for industrial equipment based on sound signals, and the method involves two types of equipment: a sending end and a control processing end.
[0057] The sending end is composed of an acoustic signal acquisition device and a microcomputer with a Wi-Fi module. In the same network, there can be multiple sending ends.
[0058] The control processing end is a processing device that can run the python programming language. In the same network, the control processing end is unique.
[0059] Each sending end maintains a bidirectional path to the control processing end, the uplink path is used for transmission of collected audio signals, and the downlink path receives management configuration data from the control processing end.
[0060] figure 1 ① in ① is used for audio signal transmission and state transmission from the sending end to the control processing end, ...
Embodiment 2
[0139] Embodiment 2 of the present disclosure provides a general fault detection method for industrial equipment based on sound signals, including the following process:
[0140] Obtain audio data of industrial equipment to be identified;
[0141] Input the acquired audio data into the preset classification model to obtain the fault detection result;
[0142] Among them, in the preset classification model based on Deep-SVDD, the features of Mel frequency cepstrum coefficient and short-term zero-crossing rate are extracted, and the acquired features are encoded by BP neural network and the distance between the encoded vector and the center of the hypersphere is calculated. According to the distance, the detection score is obtained, and the fault detection result is obtained according to the comparison between the score and the preset threshold.
[0143] The detailed working method is the same as that provided in Embodiment 1, and will not be repeated here.
Embodiment 3
[0145] Embodiment 3 of the present disclosure provides a general fault detection system for industrial equipment based on sound signals, including:
[0146] The data acquisition module is configured to: acquire the audio data of the industrial equipment to be identified;
[0147] The fault detection module is configured to: input the acquired audio data into a preset classification model constructed based on Deep-SVDD to obtain a fault detection result;
[0148] Among them, in the preset classification model based on Deep-SVDD, the features of Mel frequency cepstrum coefficient and short-term zero-crossing rate are extracted, and the acquired features are encoded by BP neural network and the distance between the encoded vector and the center of the hypersphere is calculated. According to the distance, the detection score is obtained, and the fault detection result is obtained according to the comparison between the score and the preset threshold.
[0149] The working method of ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


