CN-121978223-A - Voiceprint diagnosis method and device for boiler water wall coking

CN121978223ACN 121978223 ACN121978223 ACN 121978223ACN-121978223-A

Abstract

The application relates to the technical field of boiler coking diagnosis, in particular to a voiceprint diagnosis method and device for boiler water wall coking, wherein the method comprises the steps of collecting a voiceprint signal in a boiler; and extracting at least one voiceprint signal characteristic according to the boiler internal voiceprint data, and diagnosing the coking position and degree of the boiler water-cooled wall of the boiler according to the at least one voiceprint signal characteristic. Therefore, the problem that in the related art, the boiler water wall coking condition cannot be tracked comprehensively and in real time due to the fact that the existing boiler water wall coking detection technology is limited by factors such as equipment service life and safety is solved.

Inventors

ZHANG HAI
ZHANG HANLIN
ZHANG YIHANG
ZHANG YANG
LV JUNFU

Assignees

清华大学

Dates

Publication Date: 20260505
Application Date: 20251231

Claims (10)

1. A voiceprint diagnosis method for boiler water wall coking is characterized by comprising the following steps: Collecting a voiceprint signal in a boiler; generating boiler internal voiceprint data according to the boiler internal voiceprint signals; And extracting at least one voiceprint signal characteristic according to the voiceprint data in the boiler, and diagnosing the coking position and degree of the boiler water wall of the boiler according to the at least one voiceprint signal characteristic.
2. The method of claim 1, wherein said extracting at least one voiceprint signal feature from the in-boiler voiceprint data comprises: And inputting the voiceprint data in the boiler into a pre-trained voiceprint diagnosis model to output the at least one voiceprint signal characteristic generated by the real-time coking condition of the water-cooled wall in the boiler.
3. The method as recited in claim 1, further comprising: Acquiring the current running condition of the boiler; and matching the diagnosis measures of the boiler according to the current running condition, the coking position and degree of the water-cooled wall of the boiler, wherein the diagnosis measures comprise at least one action of soot blowing, furnace blowing and checking.
4. The method of claim 1, further comprising, prior to extracting at least one of the voiceprint signal features from the in-boiler voiceprint data: And preprocessing the boiler internal voiceprint data to extract the boiler internal voiceprint data meeting preset use conditions based on a time domain, a frequency domain and a time-frequency domain.
5. The method of claim 1, wherein diagnosing the boiler water wall coking location and extent of the boiler based on the at least one voiceprint signal feature comprises: Acquiring the actual type and/or working condition parameters of the boiler; And determining the coking position and degree of the boiler water wall by combining the actual type and/or working condition parameters and the at least one voiceprint signal characteristic.
6. The method as recited in claim 1, further comprising: predicting the coking trend of the boiler according to the coking position and degree of the water-cooled wall of the boiler; and generating a reminding signal of the boiler according to the coking trend.
7. A voiceprint diagnostic device for boiler water wall coking, comprising: the acquisition module is used for acquiring voiceprint signals in the boiler; The generation module is used for generating boiler internal voiceprint data according to the boiler internal voiceprint signals; And the diagnosis module is used for extracting at least one voiceprint signal characteristic according to the voiceprint data in the boiler and diagnosing the coking position and degree of the boiler water wall of the boiler according to the at least one voiceprint signal characteristic.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, the processor executing the program to implement the voiceprint diagnostic method of boiler water wall coking of any one of claims 1-6.
9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the program is executed by a processor for implementing the voiceprint diagnostic method of boiler water wall coking of any one of claims 1-6.
10. A computer program product comprising a computer program, wherein the computer program is executed for implementing a voiceprint diagnostic method of boiler water wall coking according to any one of claims 1-6.

Description

Voiceprint diagnosis method and device for boiler water wall coking Technical Field The application relates to the technical field of boiler coking diagnosis, in particular to a voiceprint diagnosis method and device for boiler water wall coking. Background Coal electricity is used as a basic guarantee and system regulation power supply and plays an important role in an electric power energy structure. In the running process of the coal motor group, the water-cooling wall coking of the boiler severely restricts the safe, stable and economic running of the group, and the water-cooling wall coking can influence the heat exchange between high-temperature flue gas in the boiler and working medium in the pipe, so that the efficiency of the boiler is reduced, and the problems of local heating surface overtemperature and the like can be caused. In the related technology, in the aspects of boiler water wall coking protection measures, the existing main means are still to improve the combustion condition in the boiler by means of operation adjustment, fuel blending combustion and the like, if necessary, the boiler water wall coking is timely cleared by matching with soot blowing equipment, in order to clearly and effectively adjust the boiler operation state or coke clearing in the boiler, the position and degree of the boiler water wall coking are required to be timely detected, the existing boiler water wall coking detection is often directly observed depending on a fire observation hole, and partial invasive detection equipment can enter the boiler through the fire observation hole to acquire data, such as laser ranging, high-temperature camera shooting and the like. However, the detection method is limited by factors such as equipment life, safety and the like, and cannot comprehensively and real-timely track the coking condition of the water-cooled wall of the boiler. In recent years, the acoustic wave temperature measurement technology is widely focused, and researchers also use the technology to measure the temperature distribution in the furnace so as to predict the coking condition of the water-cooled wall, but the coking problem of the water-cooled wall is not only related to the temperature, and the accuracy of a coking detection result obtained by the indirect measurement mode is yet to be examined. Disclosure of Invention The application provides a voiceprint diagnosis method and device for boiler water wall coking, which are used for solving the problems that in the related art, the boiler water wall coking condition cannot be tracked comprehensively and in real time due to the fact that the existing boiler water wall coking detection technology is limited by the factors of equipment service life, safety and the like. An embodiment of the first aspect of the application provides a voiceprint diagnosis method for coking of a water-cooled wall of a boiler, which comprises the following steps of collecting a voiceprint signal in the boiler, generating voiceprint data in the boiler according to the voiceprint signal in the boiler, extracting at least one voiceprint signal characteristic according to the voiceprint data in the boiler, and diagnosing the coking position and degree of the water-cooled wall of the boiler according to the at least one voiceprint signal characteristic. Through the technical means, the embodiment of the application can realize accurate diagnosis on the coking position and degree of the water-cooled wall by collecting the voiceprint signal in the boiler, generating the structured voiceprint data and extracting the characteristic parameters, so that the traditional mode relying on manual inspection or hysteresis parameter monitoring can be abandoned, the acoustic abnormal signal in early coking stage can be captured by means of the correlation analysis of the voiceprint characteristics and the coking state, the timeliness and the accuracy of coking fault diagnosis are greatly improved, the potential safety hazards of overheat and tube explosion of a heated surface caused by coking missing judgment and misjudgment are effectively avoided, and real-time and reliable technical guarantee is provided for safe and efficient operation of the boiler. Optionally, in one embodiment of the application, the extracting at least one voiceprint signal feature according to the in-boiler voiceprint data comprises inputting the in-boiler voiceprint data into a pre-trained voiceprint diagnostic model to output the at least one voiceprint signal feature generated by real-time coking conditions of the water wall in the boiler. By means of the technical means, the embodiment of the application can extract the voiceprint signal characteristics corresponding to the water wall coking by inputting the voiceprint data in the boiler into the pre-trained voiceprint diagnosis model, thereby replacing the complicated process of manually extracting the characteristics in the prior art, accurately capturing the ch