CN-121977873-A - Noise source separation and fault diagnosis method and device, electronic equipment and storage medium

CN121977873ACN 121977873 ACN121977873 ACN 121977873ACN-121977873-A

Abstract

The application discloses a noise source separation and fault diagnosis method, a device, electronic equipment and a storage medium, which relate to the technical field of coal mills and comprise the steps of acquiring mechanical motion parameters of the coal mill in real time; the method comprises the steps of generating a reference frequency spectrum used for representing the motion characteristics of mechanical parts based on the mechanical motion parameters, collecting multipath noise signals in the running process of the coal mill, carrying out blind source separation processing on the multipath noise signals based on the reference frequency spectrum to extract target fault noise signals, carrying out characteristic extraction on the target fault noise signals, and inputting a fault diagnosis model to obtain a diagnosis result containing fault types and/or fault positions. By the method, the device and the system, the technical effects of improving the separation accuracy of fault noise signals, comprehensively capturing the operation noise information of the coal mill and accurately realizing fault type identification and position location are achieved.

Inventors

HAN YUDONG
LI ZHENDONG
PENG JIANHUA
GAO LING
WANG WENJUN
Zhou Tiechen
LI LEI
LIU QIANG
SU XICHEN
LIU DINGPO

Assignees

国能民权热电有限公司
河南省锅炉压力容器检验技术科学研究院
西安西热锅炉环保工程有限公司

Dates

Publication Date: 20260505
Application Date: 20260114

Claims (10)

1. A noise source separation and fault diagnosis method, comprising: acquiring mechanical motion parameters of a coal mill in real time; Generating a reference spectrum for characterizing the motion characteristics of the mechanical component based on the mechanical motion parameters; collecting multipath noise signals in the running process of the coal mill; performing blind source separation processing on the multipath noise signals based on the reference frequency spectrum to extract a target fault noise signal; And extracting the characteristics of the target fault noise signals, and inputting a fault diagnosis model to obtain a diagnosis result containing the fault type and/or the fault position.
2. The noise source separation and fault diagnosis method according to claim 1, wherein the generating a reference spectrum for characterizing a motion characteristic of a mechanical component based on the mechanical motion parameter comprises: calculating the basic frequency of mechanical contact or impact according to the structural parameters and the mechanical motion parameters of the coal mill; dynamically correcting the basic frequency according to the operation time length and/or the load parameter of the coal mill; and synthesizing the reference frequency spectrum based on the corrected fundamental frequency and harmonic components thereof.
3. The method for noise source separation and fault diagnosis according to claim 1, wherein the collecting the multipath noise signals during the operation of the coal mill comprises: and arranging a plurality of acoustic sensors in different circumferential directions of the coal mill shell so as to synchronously collect the multipath noise signals, wherein probes of the acoustic sensors face the inner side of the shell and form a preset inclination angle.
4. The noise source separation and fault diagnosis method according to claim 1, wherein the blind source separation processing is performed on the multipath noise signals based on the reference spectrum to extract a target fault noise signal, comprising: Constructing an objective function taking the reference frequency spectrum as a constraint condition, wherein the objective function comprises an approximation term for prompting a separation signal to approximate the reference frequency spectrum and a regularization term for constraining the sparsity of a decoupling matrix; Solving the objective function through an iterative optimization algorithm to obtain a decoupling matrix for separating the objective fault noise signal from the mixed noise signal; And performing decoupling operation on the multipath noise signals by using the decoupling matrix, so as to separate out the target fault noise signals.
5. The noise source separation and fault diagnosis method according to claim 1, wherein the feature extraction of the target fault noise signal and inputting a fault diagnosis model to obtain a diagnosis result including a fault type and/or a fault location, comprises: Extracting spectral kurtosis characteristics and cepstrum entropy characteristics from the frequency spectrum of the target fault noise signal; normalizing the extracted spectral kurtosis characteristic and the cepstrum entropy characteristic to form a characteristic vector; The feature vectors are input to a pre-trained fault diagnosis model, from which probabilities corresponding to at least one preset fault type, and/or identification information indicating the location of the fault occurrence, are output.
6. The noise source separation and fault diagnosis method according to claim 1, further comprising: and outputting the diagnosis result and the corresponding early warning information to a local human-computer interaction interface and/or a remote monitoring system.
7. A noise source separation and fault diagnosis apparatus, comprising: The acquisition module is configured to acquire mechanical motion parameters of the coal mill in real time; a generation module configured to generate a reference spectrum for characterizing a motion feature of a mechanical component based on the mechanical motion parameter; The acquisition module is configured to acquire multipath noise signals in the running process of the coal mill; The extraction module is configured to perform blind source separation processing on the multipath noise signals based on the reference frequency spectrum so as to extract target fault noise signals; and the input module is configured to perform feature extraction on the target fault noise signal and input a fault diagnosis model to obtain a diagnosis result containing a fault type and/or a fault position.
8. An electronic device, comprising: At least one processor; and a memory communicatively coupled to the at least one processor; Wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the noise source separation and fault diagnosis method of any one of claims 1-6.
9. A non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the noise source separation and fault diagnosis method according to any one of claims 1-6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the noise source separation and fault diagnosis method according to any one of claims 1-6.

Description

Noise source separation and fault diagnosis method and device, electronic equipment and storage medium Technical Field The application relates to the technical field of coal mills, in particular to a noise source separation and fault diagnosis method, a device, electronic equipment and a storage medium. Background The coal mill is core equipment in the production of industries such as electric power, chemical industry, and the like, and the running state of the coal mill directly influences production efficiency and safety, and a fault diagnosis technology is crucial to guaranteeing stable operation of the coal mill. In the related coal mill fault diagnosis technology, a single-point or limited-point arrangement mode is adopted for noise collection, global noise distribution characteristics of the coal mill are difficult to fully cover, so that fault related noise signals are incomplete to capture, a fixed weight blind source separation algorithm without mechanical motion reference guidance is often adopted in a noise source separation link, the noise source separation link is easily influenced by irrelevant signals such as environmental interference and airflow noise, the fault signal separation precision is low, target fault noise is difficult to effectively extract, the fault diagnosis is judged based on single characteristics, the diagnosis accuracy is insufficient, fault position information cannot be accurately output, and inconvenience is brought to follow-up maintenance work. These problems seriously affect the reliability and practicality of the fault diagnosis of the coal mill. Disclosure of Invention The application provides a noise source separation and fault diagnosis method, a device, electronic equipment and a storage medium. The method can solve the problems that in the related technology, blind source separation precision is low, information is omitted due to single noise acquisition, and fault type and position cannot be simultaneously determined due to lack of mechanical motion characteristic reference. According to a first aspect of the present application, there is provided a noise source separation and fault diagnosis method, including: acquiring mechanical motion parameters of a coal mill in real time; Generating a reference spectrum for characterizing the motion characteristics of the mechanical component based on the mechanical motion parameters; collecting multipath noise signals in the running process of the coal mill; performing blind source separation processing on the multipath noise signals based on the reference frequency spectrum to extract a target fault noise signal; And extracting the characteristics of the target fault noise signals, and inputting a fault diagnosis model to obtain a diagnosis result containing the fault type and/or the fault position. According to a second aspect of the present application, there is provided a noise source separation and fault diagnosis apparatus comprising: The acquisition module is configured to acquire mechanical motion parameters of the coal mill in real time; a generation module configured to generate a reference spectrum for characterizing a motion feature of a mechanical component based on the mechanical motion parameter; The acquisition module is configured to acquire multipath noise signals in the running process of the coal mill; The extraction module is configured to perform blind source separation processing on the multipath noise signals based on the reference frequency spectrum so as to extract target fault noise signals; and the input module is configured to perform feature extraction on the target fault noise signal and input a fault diagnosis model to obtain a diagnosis result containing a fault type and/or a fault position. According to a third aspect of the present application, there is provided an electronic device comprising: At least one processor; and a memory communicatively coupled to the at least one processor; Wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the noise source separation and fault diagnosis method of the first aspect described above. According to a fourth aspect of the present application, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the noise source separation and fault diagnosis method of the foregoing first aspect. According to a fifth aspect of the present application there is provided a computer program product comprising a computer program which, when executed by a processor, implements the noise source separation and fault diagnosis method of the first aspect described above. The application provides a noise source separation and fault diagnosis method, a device, electronic equipment and a storage medium, which comprise the steps of acquiring mechanical motion parameters of a coal mill in real time; the method comprises the st