CN-122014653-A - Quantitative vibration monitoring and fault early warning method and related device for thermal power fan

CN122014653ACN 122014653 ACN122014653 ACN 122014653ACN-122014653-A

Abstract

The invention discloses a thermal power fan vibration quantitative monitoring and fault early warning method and a related device, which comprise the steps of obtaining vibration and operation parameters of a thermal power fan, preprocessing the vibration and operation parameters of the thermal power fan to obtain processed data, extracting characteristics of the preprocessed data, screening the extracted characteristics, constructing a characteristic matrix by utilizing the screened characteristics, inputting the characteristic matrix into a trained random forest quantitative model, and carrying out fault early warning according to an output result of the random forest quantitative model.

Inventors

KANG SHILIN
ZHENG XIANGHUA
QIU HUA
LIU SHUPENG
ZHOU YAOCHONG
WANG BAOYUN
KANG WEI
KONG LINGHUI
LI BAOGEN
FAN YANHONG

Assignees

西安热工研究院有限公司
华能云南滇东能源有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260121

Claims (10)

1. The method for quantitatively monitoring vibration and pre-warning faults of the thermal power fan is characterized by comprising the following steps of: Obtaining vibration and operation parameters of a thermal power fan; Preprocessing the vibration and operation parameters of the thermal power fan to obtain processed data; Extracting features of the preprocessed data, screening the extracted features, and constructing a feature matrix by using the screened features; And inputting the feature matrix into the trained random forest quantitative model, and performing fault early warning according to the output result of the random forest quantitative model.
2. The method for quantitatively monitoring vibration and pre-warning faults of a thermal power fan according to claim 1, wherein vibration and operation parameters of the thermal power fan comprise acceleration signals of front and rear bearings in a bearing seat, radial vibration speed signals of a coupling, circumferential vibration displacement signals at an outlet of a shell, fan rotating speed, unit load, inlet and outlet wind pressure and wind quantity, medium temperature and bearing temperature.
3. The method for quantitatively monitoring vibration and pre-warning faults of a thermal power fan according to claim 1, wherein the process of preprocessing vibration and operation parameters of the thermal power fan is as follows: and carrying out baseline correction, noise removal and normalization processing on the vibration and operation parameters of the thermal power fan.
4. The method for quantitatively monitoring vibration and pre-warning faults of a thermal power fan according to claim 1, wherein the process of extracting features from the preprocessed data, screening the extracted features and constructing a feature matrix by using the screened features is as follows: and extracting features of the preprocessed data, screening the extracted features by a feature importance assessment mechanism of a random forest algorithm, and constructing a feature matrix by using the screened features.
5. The method for quantitatively monitoring vibration and pre-warning faults of a thermal power fan according to claim 1, wherein the extracted features comprise time domain features, frequency domain features and operation parameter features.
6. The method for quantitatively monitoring vibration and early warning faults of a thermal power fan according to claim 1, wherein the output result of the random forest quantitative model comprises a vibration intensity quantitative value, a fault type probability and a fault severity coefficient.
7. The method for quantitatively monitoring vibration and pre-warning faults of a thermal power fan according to claim 6, wherein the process of pre-warning faults according to the output result of the random forest quantitative model is as follows: The vibration intensity quantitative value is more than or equal to 4, or the fault severity coefficient is 0.3-0.5, triggering a prompting alarm and pushing the prompting alarm to an operation and maintenance terminal; The vibration intensity quantitative value is more than or equal to 6, or the fault severity coefficient is 0.5-0.8, an important alarm is triggered, and the linked power plant DCS system displays fault information; and if the vibration intensity quantitative value is more than or equal to 8 or the fault severity coefficient is more than or equal to 0.8, triggering an emergency alarm and suggesting the unit to reduce load or stop for maintenance.
8. The utility model provides a thermal power fan vibration ration monitoring and trouble early warning system which characterized in that includes: the acquisition module is used for acquiring vibration and operation parameters of the thermal power fan; the pretreatment module is used for carrying out pretreatment on vibration and operation parameters of the thermal power fan to obtain data after treatment; The extraction module is used for extracting the characteristics of the preprocessed data, screening the extracted characteristics and constructing a characteristic matrix by utilizing the screened characteristics; And the early warning module is used for inputting the feature matrix into the trained random forest quantitative model and carrying out fault early warning according to the output result of the random forest quantitative model.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the thermal power fan vibration quantitative monitoring and fault warning method according to any one of claims 1-7.
10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of the thermal power fan vibration quantitative monitoring and fault warning method according to any one of claims 1 to 7.

Description

Quantitative vibration monitoring and fault early warning method and related device for thermal power fan Technical Field The invention belongs to the technical field of state monitoring of thermal power equipment, and relates to a thermal power fan vibration quantitative monitoring and fault early warning method and a related device. Background The thermal power fan (comprising a draught fan, a blower, a primary fan and the like) is key auxiliary equipment of the thermal power plant, and the operation stability of the thermal power fan directly influences the safe and economic operation of the unit. Vibration is a core index reflecting the running state of the fan, and in the long-term running process of the fan, the fan is easy to cause abnormal vibration due to the problems of unbalanced rotor, non-centering of a coupling, abrasion of a bearing, dust accumulation and corrosion of a blade and the like, and if the vibration is not monitored and early-warned in time, serious consequences such as equipment shutdown, component damage and even unplanned shutdown of a unit can be caused. The existing thermal power fan vibration monitoring method mainly comprises a traditional spectrum analysis, a threshold judgment method and a simple machine learning method: 1. The traditional spectrum analysis method (such as FFT analysis) identifies faults by extracting frequency characteristics of vibration signals, but is greatly influenced by fan operation condition fluctuation (such as load change and medium parameter fluctuation), quantitative evaluation of fault severity is difficult to realize, and identification sensitivity to early slight faults is insufficient; 2. the threshold judgment method triggers an alarm based on a preset vibration amplitude threshold, can only realize qualitative judgment, cannot reflect the fault development trend, and has poor generalization capability due to the dependence of threshold setting on experience; 3. Although the existing simple machine learning method (such as a single decision tree and a support vector machine) tries to quantitatively analyze, the problems of inaccurate feature screening, weak anti-interference capability of a model, low processing efficiency of high-dimensional monitoring data and the like exist, the Mean Square Error (MSE) of quantitative prediction is large, the decision coefficient (R2) is more lower than 0.85, and the high-precision monitoring requirement of a thermal power fan is difficult to meet. Disclosure of Invention The invention aims to overcome the defects of the prior art, and provides a thermal power fan vibration quantitative monitoring and fault early warning method and a related device. In order to achieve the purpose, the invention discloses a thermal power fan vibration quantitative monitoring and fault early warning method, which comprises the following steps: Obtaining vibration and operation parameters of a thermal power fan; Preprocessing the vibration and operation parameters of the thermal power fan to obtain processed data; Extracting features of the preprocessed data, screening the extracted features, and constructing a feature matrix by using the screened features; And inputting the feature matrix into the trained random forest quantitative model, and performing fault early warning according to the output result of the random forest quantitative model. Further, the vibration and operation parameters of the thermal power fan comprise acceleration signals of front and rear bearings in a bearing seat, radial vibration speed signals of a coupler, circumferential vibration displacement signals at an outlet of a shell, fan rotating speed, unit load, inlet and outlet wind pressure and wind quantity, medium temperature and bearing temperature. Further, the process of preprocessing the vibration and operation parameters of the thermal power fan is as follows: and carrying out baseline correction, noise removal and normalization processing on the vibration and operation parameters of the thermal power fan. Further, the process of extracting features from the preprocessed data, screening the extracted features, and constructing a feature matrix by using the screened features includes: and extracting features of the preprocessed data, screening the extracted features by a feature importance assessment mechanism of a random forest algorithm, and constructing a feature matrix by using the screened features. Further, the extracted features include time domain features, frequency domain features, and operating parameter features. Further, the output result of the random forest quantitative model comprises a vibration intensity quantitative value, a fault type probability and a fault severity coefficient. Further, the fault early warning process according to the output result of the random forest quantitative model comprises the following steps: The vibration intensity quantitative value is more than or equal to 4, or the fault severity coeffici