Equipment and method for detecting abnormal sound of device

A technology for equipment abnormality and sound detection, which is applied in speech analysis, instruments, alarms, etc., can solve the problems of not eliminating the difference in magnitude of sound samples, unable to distinguish abnormal sounds, and audio features are not prominent, etc., to achieve clustering Effects with remarkable effect, low misconvergence rate, and excellent audio characteristics

Pending Publication Date: 2020-04-21
SHENZHEN JIANGXING INTELLIGENCE INC
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Existing similar research proposes to predict abnormal sounds based on MFCC algorithm and SVM algorithm. This research has flaws in data preprocessing and extraction of audio signal features, and does not eliminate the numerical magnitude difference between sound samples in the acqui

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Equipment and method for detecting abnormal sound of device
  • Equipment and method for detecting abnormal sound of device
  • Equipment and method for detecting abnormal sound of device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0047] This embodiment provides a specific structure of equipment for abnormal sound detection of equipment, such as figure 1 As shown in -3, a device for detecting abnormal sound of equipment includes a device main body 1, and the device main body 1 includes a front-end data collector and an edge computing container;

[0048] The front-end data collector is a device that can collect audio features and compress, transmit and store data signals. The device includes an audio collection module 2, an audio service module 3 and a storage module 4;

[0049] The edge computing container is a device that can acquire data and has computing and analysis capabilities. The device includes an arm development board 5 , a network communication module 6 , an alarm module 7 and a power supply module 8 .

[0050] Such as figure 2 and 3 As shown, the audio acquisition module 2 adopts a pickup device, which can realize high-fidelity recording operation in outdoor and long-distance environments. ...

Embodiment 2

[0058] This embodiment provides a specific use method and detection steps of a device and method for abnormal sound detection of equipment, such as Figure 4 shown, follow the steps below:

[0059] (1) Start the abnormal sound detection equipment;

[0060] (2) Abnormal sound detection equipment does not need to be close to the detected equipment, and is placed near the end of the detected equipment to trigger the starting end of the audio acquisition module 2, and the audio acquisition module 2 collects the audio data of the detected equipment in real time;

[0061] (3) adopt audio service module 3 to receive the audio data of audio collection module 2, carry out digital signal compression and transmission;

[0062] (4) Transfer the compressed digital signal to the arm development board 5 in the edge computing container; for the modeling of the application scene and noise recognition, enter the training program, the arm development board 5 trains the normal equipment operatio...

Embodiment 3

[0067] This embodiment provides real-time collection of audio data of the operating equipment through the front-end data collector, slices the audio data, analyzes the characteristics of the sliced ​​audio data and compares the characteristics of the normal operation of the equipment in the experimental data, and accurately identifies the failure of the operating equipment Specific steps when the abnormal sound, such as Figure 5 -7 shows,

[0068] After step S3, the audio data of the operating equipment is collected in real time by the front-end data collector, and the audio data is sliced, and in step S4, the arm development board 5 is used to process the sliced ​​data; as Figure 5-7 As shown, the specific steps of the above operation are as follows:

[0069] A. Obtain and integrate the operating sound information of the equipment to be detected, and establish an audio database.

[0070] B. Read the data in the audio database, including dividing the training-test set, sli...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

Equipment for detecting the abnormal sound of equipment device comprises an equipment body, and is characterized in that the equipment body comprises a front-end data collector and an edge computing container; the front-end data collector is a device capable of collecting audio characteristics and compressing, transmitting and storing data signals. The technical point is characterized in that after data slicing processing is carried out in audio data processing, each obtained feature vector matrix of audio features extracted through an MFCC method is straightened and subjected to dimension reduction processing, the extracted audio features are better, and the clustering effect is more remarkable; by adopting the clustering algorithm of the self-organizing feature mapping neural network (PSO-SOM) based on particle swarm optimization, the label of the data does not need to be known; according to the method, the class with large data volume is classified as the normal operation sound by default in a segment of long equipment operation sound data, the SO-SOM algorithm is superior to the SOM basic algorithm in clustering effect evaluation indexes (AVQ, UTE and LWDI), and the obtained network is good in quality and low in misclustering rate.

Description

technical field [0001] The invention belongs to the technical field of sound detection, in particular to a device and a method for detecting abnormal sound of equipment. Background technique [0002] Audio monitoring technology is one of the most important components in monitoring applications. The key to realizing intelligent audio monitoring is to automatically detect abnormal sounds from ambient background sounds. To identify abnormal sounds, it is necessary to distinguish abnormal sounds from multiple background sounds superimposed on the equipment operating environment, and effectively detect the target scene in real time. The real-time intelligent audio monitoring greatly reduces the manpower, material resources, and financial resources required for equipment detection. [0003] At present, abnormal sound detection has been applied to security monitoring, industrial production, medical treatment and other fields. The purpose of real-time alarm can be achieved through a...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G10L25/51G10L25/72G10L25/24G10L25/30G08B21/18
CPCG08B21/18G10L25/24G10L25/30G10L25/51G10L25/72
Inventor 王尧陈哲樊小毅张聪宋丹阳庞海天
Owner SHENZHEN JIANGXING INTELLIGENCE INC
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap