CN-122014541-A - Wind turbine generator monitoring method, device and monitoring system

CN122014541ACN 122014541 ACN122014541 ACN 122014541ACN-122014541-A

Abstract

The application discloses a wind turbine generator monitoring method, a device and a monitoring system, wherein an edge computing device receives sensor data sent by a multichannel optical fiber demodulator, performs multi-physical-field coupling load inversion on the sensor data based on a preset wind turbine generator dynamics reduced-order model to obtain a load inversion result, performs feature extraction on the sensor data to obtain feature information, performs wind turbine generator fault prediction based on a prediction model and the feature information to obtain a fault prediction result, performs decoupling calculation on the load inversion result based on a preset fault transfer matrix under the condition that at least one of the load inversion result and the fault prediction result is abnormal, determines fault source information, determines first data according to the sensor data, the load inversion result and the fault source information, and sends the first data to a front-end platform so that the front-end platform performs three-dimensional visual rendering and state cloud map mapping presentation on the first data to complete real-time visual display of the running state of the wind turbine generator.

Inventors

HAN XUMING
XUE WEILI
QIAO JIE
XU XINXING
ZHANG JIN

Assignees

安徽籍田科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260416

Claims (10)

1. The wind turbine generator monitoring method is characterized by being applied to an edge computing device, wherein the edge computing device is connected with a multichannel optical fiber demodulator of a wind turbine generator, the multichannel optical fiber demodulator is connected with optical fiber sensors arranged at the root parts of blades, a tower barrel flange and an anchor cable anchoring area of the wind turbine generator, and the method comprises the following steps: receiving sensor data sent by the multichannel optical fiber demodulator, wherein the sensor data comprises a blade dynamic strain signal of the root of the blade, a tower cylinder flange displacement signal of the tower cylinder flange and an anchor rope dynamic stress signal of the anchor rope anchoring area; Performing multi-physical-field coupling load inversion on the sensor data based on a preset wind turbine generator dynamics reduced order model to obtain a load inversion result; Extracting the characteristics of the sensor data to obtain characteristic information; carrying out wind turbine generator fault prediction based on the prediction model and the characteristic information to obtain a fault prediction result; under the condition that at least one of the load inversion result and the fault prediction result is abnormal, decoupling calculation is carried out on the load inversion result based on a preset fault transmission matrix, and fault source information is determined; and determining first data according to the sensor data, the load inversion result and the fault source information, and sending the first data to a front-end platform so that the front-end platform performs three-dimensional visual rendering and state cloud image mapping presentation on the first data to complete real-time visual presentation of the running state of the wind turbine generator.
2. The wind turbine monitoring method of claim 1, further comprising: acquiring an optical power fluctuation value of the multichannel optical fiber demodulator in a data transmission process; Calibrating the dynamic strain signal of the blade based on the optical power fluctuation value to obtain a calibrated blade strain signal; And determining the sensor data based on the calibrated blade strain signal, the tower flange displacement signal and the anchor cable dynamic stress signal.
3. The wind turbine monitoring method according to claim 2, wherein the performing the multi-physical-field coupling load inversion on the sensor data based on the preset wind turbine dynamics reduced order model to obtain a load inversion result includes: Based on a preset wind turbine generator system dynamics reduced order model, carrying out conversion of real physical loads on the calibrated blade strain signals, tower cylinder flange displacement signals and anchor cable dynamic stress signals in the sensor data respectively to obtain real-time stress loads, tower cylinder flange bending moment loads and anchor cable tension loads of the blades; And determining the load inversion result according to the real-time stress load of the blade, the bending moment load of the tower flange and the tension load of the anchor cable.
4. The wind turbine monitoring method according to claim 1, wherein the feature extracting the sensor data to obtain feature information includes: performing wavelet transformation on the sensor data to obtain a multi-scale time-frequency decomposition signal; Extracting signal peak value, spectrum energy and frequency band amplitude of the multi-scale time-frequency decomposition signal; and determining the characteristic information according to the peak value of the signal, the spectrum energy and the frequency band amplitude.
5. The wind turbine monitoring method according to claim 1, wherein the predicting wind turbine faults based on the prediction model and the feature information to obtain fault prediction results comprises: Based on the prediction model and the characteristic information, predicting the abnormal condition of the wind turbine generator in a preset time period in the future to obtain an abnormal prediction probability value, wherein the abnormal prediction probability value comprises a resonance event, a load exceeding standard, flange clearance abnormality and a probability value of each abnormal condition of anchor cable loosening, the resonance event represents the abnormal condition of the wind turbine generator resonance caused by rotation of a blade, the load exceeding standard represents the abnormal condition of the blade, a tower and the anchor cable, the flange clearance abnormality represents the abnormal condition that the flange clearance does not accord with a preset clearance range, and the anchor cable loosening represents the abnormal condition of anchor cable failure caused by insufficient tension; and determining the fault prediction result based on the abnormal prediction probability value.
6. The wind turbine monitoring method according to claim 1, wherein the performing decoupling calculation on the load inversion result based on a preset fault transmission matrix, and determining fault source information, includes: Inputting the load inversion result into a preset fault transmission matrix for decoupling calculation to obtain the contribution degree duty ratio of blade load abnormality, tower barrel flange looseness and anchor cable looseness to the current abnormality; And determining fault source information according to the contribution ratio.
7. The wind turbine monitoring method of claim 1, further comprising: Determining that the load inversion result is abnormal under the condition that any load in the load inversion result is larger than a corresponding preset load threshold value; And under the condition that any one of the fault prediction probability values is larger than a corresponding preset probability threshold value, determining that the fault prediction result is abnormal.
8. The wind turbine monitoring method of claim 1, further comprising: Under the condition that the load inversion result and the fault prediction result are not abnormal, determining second data according to the real-time azimuth angle signal of the blade encoder of the wind turbine and the characteristic information; And sending the second data to a front-end platform for data display.
9. An edge computing device comprising a processor, a memory storing executable instructions for the processor, which when executed by the processor, implement a wind turbine monitoring method according to any one of claims 1 to 8.
10. The wind turbine generator monitoring system is characterized by comprising an optical fiber sensor, a multichannel optical fiber demodulator, an edge computing device and a front-end platform, wherein the optical fiber sensor is arranged at the root of a blade of a wind turbine generator, a tower flange and an anchor cable anchoring area, the multichannel optical fiber demodulator is connected with the optical fiber sensor, and the edge computing device is connected with the multichannel optical fiber demodulator; the optical fiber sensor is used for collecting dynamic strain data of the blade, displacement data of the tower barrel flange and dynamic stress data of the anchor cable respectively; the multichannel optical fiber demodulator is used for demodulating the data acquired by each optical fiber sensor to obtain sensor data, and transmitting the sensor data to the edge computing device; The edge computing device is used for receiving sensor data sent by the multichannel optical fiber demodulator, carrying out multi-physical field coupling load inversion on the sensor data based on a preset wind turbine generator dynamics reduced order model to obtain a load inversion result, carrying out feature extraction on the sensor data to obtain feature information, carrying out wind turbine generator fault prediction based on a prediction model and the feature information to obtain a fault prediction result, carrying out decoupling computation on the load inversion result based on a preset fault transfer matrix under the condition that at least one of the load inversion result and the fault prediction result is abnormal, determining fault source information, determining first data according to the sensor data, the load inversion result and the fault source information, and sending the first data to the front-end platform, wherein the sensor data comprises a blade dynamic strain signal of a blade root, a tower flange displacement signal of a tower flange and a dynamic stress signal of an anchor cable of the anchor cable anchoring area; The front-end platform is used for performing three-dimensional visual rendering and state cloud mapping presentation on the first data to complete real-time visual display of the running state of the wind turbine.

Description

Wind turbine generator monitoring method, device and monitoring system Technical Field The application relates to the technical field of wind turbine generator system state monitoring, in particular to a wind turbine generator system monitoring method, device and system. Background Wind power generation has been rapidly developed in recent years as an important component of renewable energy sources worldwide. Along with the continuous increase of the single-machine capacity of the wind turbine generator (8 MW-16 MW at present), the length of the blades exceeds 100 meters, the height of the tower barrel breaks through 160 meters, and the running load and fatigue damage risk of each component of the wind turbine generator are also increased. The method has the advantages that accidents such as blade breakage, tower flange loosening, anchor cable failure and the like frequently occur, huge economic loss (single maintenance cost is over million yuan) is caused, and the safety of personnel in a wind power plant is threatened, so that a high-reliability and high-precision monitoring system for the states of key parts of the wind power generator is established, early fault early warning and intelligent diagnosis are realized, and the method becomes urgent requirements for industry development. At present, the state monitoring of the wind turbine generator mainly adopts a resistance strain gauge as a sensing element, so that the problems of poor electromagnetic interference resistance, short service life, serious temperature drift and the like exist, the problems of dynamic monitoring of a rotating part and coupling analysis of multiple physical fields cannot be solved by a general optical fiber monitoring scheme, the contradiction between real-time performance and diagnosis depth exists in a traditional data transmission mode, and the high-precision, long-service life and intelligent monitoring requirements of the large wind turbine generator cannot be met. Disclosure of Invention The application provides a wind turbine generator system monitoring method, a device and a monitoring system, which can realize the integrated processing of multi-component data of the wind turbine generator system, give consideration to the real-time performance of data processing and fault diagnosis depth, and provide a full-flow intelligent solution for monitoring the state of the wind turbine generator system. In order to achieve the above object, the present application provides the following technical solutions: In a first aspect, an embodiment of the present application provides a wind turbine generator monitoring method, applied to an edge computing device, where the edge computing device is connected with a multichannel optical fiber demodulator of the wind turbine generator, and the multichannel optical fiber demodulator is connected with an optical fiber sensor disposed at a root of a blade, a flange of a tower and an anchor cable anchoring area of the wind turbine generator, and the method includes: Receiving sensor data sent by a multichannel optical fiber demodulator, wherein the sensor data comprises a blade dynamic strain signal of the root of a blade, a tower cylinder flange displacement signal of a tower cylinder flange and an anchor rope dynamic stress signal of an anchor rope anchoring area; Carrying out multi-physical-field coupling load inversion on sensor data based on a preset wind turbine generator dynamics reduced order model to obtain a load inversion result; extracting characteristics of the sensor data to obtain characteristic information; carrying out wind turbine generator fault prediction based on the prediction model and the characteristic information to obtain a fault prediction result; under the condition that at least one of the load inversion result and the fault prediction result is abnormal, decoupling calculation is carried out on the load inversion result based on a preset fault transmission matrix, and fault source information is determined; And determining first data according to the sensor data, the load inversion result and the fault source information, and sending the first data to a front-end platform so that the front-end platform performs three-dimensional visual rendering and state cloud mapping presentation on the first data to complete real-time visual display of the running state of the wind turbine. In some embodiments, the method further comprises: Acquiring an optical power fluctuation value of the multichannel optical fiber demodulator in a data transmission process; calibrating the dynamic strain signal of the blade based on the optical power fluctuation value to obtain a calibrated blade strain signal; and determining sensor data based on the calibrated blade strain signals, tower barrel flange displacement signals and anchor cable dynamic stress signals. In some embodiments, performing multi-physical-field coupling load inversion on sensor data based on a preset wind turbine dyn