CN116733692A - Wind turbine generator system fault early warning method and system based on multi-model combination - Google Patents
Wind turbine generator system fault early warning method and system based on multi-model combination Download PDFInfo
- Publication number
- CN116733692A CN116733692A CN202310933648.3A CN202310933648A CN116733692A CN 116733692 A CN116733692 A CN 116733692A CN 202310933648 A CN202310933648 A CN 202310933648A CN 116733692 A CN116733692 A CN 116733692A
- Authority
- CN
- China
- Prior art keywords
- wind turbine
- fault
- turbine generator
- model
- early warning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000004804 winding Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 2
- 230000000087 stabilizing effect Effects 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 description 13
- 238000012544 monitoring process Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 6
- 230000000052 comparative effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/098—Distributed learning, e.g. federated learning
Abstract
The invention provides a wind turbine generator system fault early warning method and system based on multi-model combination, comprising the following steps: step 1, acquiring historical operation data of a wind turbine generator; step 2, correcting a preset fault judgment model by using historical operation data to obtain a fault scoring model corresponding to the wind turbine generator; step 3, inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into a fault scoring model to obtain a fault score; step 4, performing fault early warning on the wind turbine generator according to the obtained fault score; the invention can realize accurate scoring of the items corresponding to the data, further judge the operation condition of the machinery corresponding to the items, and play a role in early warning.
Description
Technical Field
The invention relates to the technical field of wind power, in particular to a wind turbine generator fault early warning method and system based on multi-model combination.
Background
In order to alleviate environmental problems caused by the use of fossil energy, wind energy, which is a clean energy source that has been developed, has been more commonly used in various fields. Among them, the use of wind energy for power generation is a field that is currently developing more rapidly. The wind generating set is an important device for realizing wind power generation;
the faults of key components such as a generator, a gear box, a low-speed shaft, a high-speed shaft, blades, an electric system, a yaw system and the like in the wind turbine generator set account for 25% of all faults, but account for 95% of all fault downtime, wherein the fault rate of the electric system is relatively high, and the average fault removal time of the gear box, the generator and a transmission chain is long, namely about 5-14 days. The fan numerous components transmit power together, so that any critical component failure will cause a shutdown or other associated component damage;
through retrieval, the invention patent with the Chinese patent number of CN111639110A discloses a wind turbine generator fault early warning method and device. The method comprises the following steps: collecting second-level real-time data of the wind turbine generator; dividing the wind turbine into different groups according to power characteristics in second-level real-time data of the wind turbine by parallel computation on a parallel computing system; storing the second-level real-time data into a distributed storage system according to the division result of the group; and identifying the outlier fans through monitoring the operation parameters of each group of wind turbine generators, so as to perform fault early warning. The wind turbine generator set fault early warning method and device provided by the invention utilize distributed storage and parallel calculation to mine fans and parameters of each fan group in the wind power plant, which are abnormal in performance under the operating condition, and alarm the abnormal wind turbine generator set in advance;
however, the method only can collect data and then early warn faults of the wind turbine, but the method lacks of eliminating abnormal data in the collected data, and also lacks of establishing a fault scoring model, and only through correlation among the data, accurate pre-judgment of faults of the wind turbine is difficult, so that a wind turbine fault early warning method based on multi-model combination is needed.
Disclosure of Invention
The invention aims to provide a wind turbine generator fault early warning method based on multi-model combination, which solves the defects existing in the prior art.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the invention provides a wind turbine generator system fault early warning method based on multi-model combination, which comprises the following steps:
step 1, acquiring historical operation data of a wind turbine generator;
step 2, correcting a preset fault judgment model by using historical operation data to obtain a fault scoring model corresponding to the wind turbine generator;
step 3, inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into a fault scoring model to obtain a fault score;
and 4, performing fault early warning on the wind turbine generator according to the obtained fault score.
Preferably, in step 1, the real-time operation data includes a temperature, a vibration frequency and a rotation speed inside the wind turbine.
Preferably, the temperature inside the wind turbine generator set includes the temperature of the gearbox, the temperature of the bearings, the temperature of the generator windings and the temperature inside the nacelle.
Preferably, the wind turbine generator rotational speed comprises an impeller rotational speed and a generator rotational speed.
Preferably, in step 2, the specific method of the preset fault judgment model is as follows:
acquiring an existing fault model;
splitting the obtained fault model to obtain data and corresponding faults;
and performing association learning according to the obtained data and the corresponding faults to obtain a fault judgment model.
Preferably, in step 2, the preset fault judgment model is corrected by using the historical operation data to obtain a fault scoring model corresponding to the wind turbine, and the specific method is as follows:
acquiring the working state of the corresponding part according to the historical operation data;
and correcting the fault judgment model according to the working state of the part and the convolutional neural network to obtain a fault scoring model.
Preferably, in step 4, fault early warning is performed on the wind turbine generator according to the obtained fault score, and the specific method is as follows:
when k is more than or equal to 75, judging that the project is in a stable state, and stabilizing the project machinery;
when the k is more than or equal to 75 and less than or equal to 60, judging that the project is in a fatigue state, and maintaining the project machinery when the service life of the project machinery is reduced;
when k is less than 60, the machine corresponding to the project is judged to be faulty and needs to be maintained in time.
Wind turbine generator system fault early warning system based on multimode combination includes:
the data acquisition unit is used for acquiring historical operation data of the wind turbine generator;
the model correction unit is used for correcting a preset fault judgment model by utilizing the historical operation data to obtain a fault scoring model corresponding to the wind turbine generator;
the score calculating unit is used for inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into the fault scoring model to obtain a fault score;
and the early warning unit is used for carrying out fault early warning on the wind turbine generator according to the obtained fault score.
Compared with the prior art, the invention has the beneficial effects that:
according to the wind turbine generator fault early warning method based on the multi-model combination, a large amount of wind turbine generator data are summarized, meanwhile, the working conditions of parts corresponding to the summarized data are corrected according to the use conditions of the parts, the generated fault judgment pair model is corrected, so that a wind turbine generator fault scoring model is obtained, accurate scoring of items corresponding to the data can be achieved through the wind turbine generator fault scoring model, the operation conditions of machinery corresponding to the items are judged, and an early warning effect is achieved.
Detailed Description
The present invention will be described in further detail below.
Examples
The invention provides a wind turbine generator system fault early warning method based on multi-model combination, which comprises the following steps:
firstly, collecting historical data of a wind turbine, wherein the collected data comprise the temperature, vibration frequency and rotating speed inside the wind turbine, and meanwhile, a monitoring camera is arranged outside the wind turbine, and the temperature inside the wind turbine comprises the temperature of a gear box, the temperature of a bearing, the temperature of a generator winding and the temperature inside a cabin;
the internal rotating speed of the wind turbine generator comprises the rotating speed of an impeller and the rotating speed of a generator;
step two, eliminating abnormal data in the acquired data to obtain eliminated data; the specific method for eliminating the abnormal data comprises the following steps:
and identifying the image shot by the monitoring camera, judging whether the abnormal data is caused by external factors or not according to the identified image, and then eliminating the abnormal data.
Thirdly, summarizing the existing fault model, splitting the existing fault model, and learning the association between data and faults so as to generate a fault judgment model;
summarizing a large amount of wind turbine generator data, meanwhile summarizing the working conditions of parts corresponding to the data, and correcting the fault judgment model according to the working conditions of the parts so as to obtain a wind turbine generator fault scoring model;
the specific correction method is as follows:
the wind turbine generator system fault scoring model adds a fault judgment model into a network in the building process, and then convolution operation is adopted to generate:
and (3) reactivating a function layer through a function generated by convolution operation, introducing a loss function, and judging faults generated by the parts according to the working conditions of the parts so as to further realize the generation of a fault scoring model of the wind turbine generator.
A monitoring camera is arranged outside the wind turbine generator to collect deflection angles of the blades, and at the same time, the monitoring camera shoots a main shaft of the wind turbine generator every 1s, and compares the main shaft with an initial main shaft image to judge whether the main shaft is bent or not;
fifthly, transmitting the deflection angle of the blade and the state information of the main shaft detected in the fourth step to a generated fault scoring model, and judging that the project is in a stable state and the project machinery works stably when k is more than or equal to 75;
when the k is more than or equal to 75 and less than or equal to 60, judging that the project is in a fatigue state, and maintaining the project machinery when the service life of the project machinery is reduced;
when k is less than 60, judging that the machinery corresponding to the project fails and needs to be maintained in time;
and judging the operation condition of the machine corresponding to the project through the obtained scores.
In this embodiment, before data are acquired, abnormal data generated due to external factors can be removed by identifying the external environment, a large number of wind turbine data are summarized, meanwhile, the working conditions of parts corresponding to the summarized data are summarized, the generated fault judgment is corrected for the model according to the use condition of the parts, so that a wind turbine fault scoring model is obtained, accurate scoring of items corresponding to the data can be achieved by the wind turbine fault scoring model, the operation condition of machinery corresponding to the items is judged, and an early warning effect is achieved.
Wind turbine generator system fault early warning system based on multimode combination includes:
the data acquisition unit is used for acquiring historical operation data of the wind turbine generator;
the model correction unit is used for correcting a preset fault judgment model by utilizing the historical operation data to obtain a fault scoring model corresponding to the wind turbine generator;
the score calculating unit is used for inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into the fault scoring model to obtain a fault score;
and the early warning unit is used for carrying out fault early warning on the wind turbine generator according to the obtained fault score.
Comparative example 1
The method comprises the steps that firstly, data of a wind turbine generator are collected in real time, the collected data comprise the temperature, vibration frequency and rotating speed inside the wind turbine generator, meanwhile, a monitoring camera is installed outside the wind turbine generator, and the temperature inside the wind turbine generator comprises the temperature of a gear box, the temperature of a bearing, the temperature of a generator winding and the temperature inside a cabin;
the internal rotating speed of the wind turbine generator comprises the rotating speed of an impeller and the rotating speed of a generator;
secondly, summarizing the existing fault model, splitting the existing fault model, learning the association between data and faults, so as to generate a fault judgment model, summarizing a large amount of wind turbine generator data, meanwhile summarizing the working condition of parts corresponding to the data, correcting the generated fault judgment model according to the use condition of the parts, and thus obtaining a wind turbine generator fault scoring model;
the wind turbine generator system fault scoring model adds a fault judgment model into a network in the building process, and then convolution operation is adopted to generate:
the function layer is re-activated through the function generated by convolution operation, a loss function is introduced, and meanwhile, the fault generated by the part pair is judged according to the use condition of the part, so that the generation of a wind turbine generator set fault scoring model is realized;
thirdly, arranging a monitoring camera outside the wind turbine generator to collect deflection angles of the blades, simultaneously shooting a main shaft of the wind turbine generator every 1s, comparing the main shaft with an initial main shaft image, and judging whether the main shaft is bent or not;
step four, transmitting the detected data to a generated fault scoring model, and judging that the project is in a stable state when k is more than or equal to 75, wherein the project machinery works stably;
when the k is more than or equal to 75 and less than or equal to 60, judging that the project is in a fatigue state, and maintaining the project machinery when the service life of the project machinery is reduced;
when k is less than 60, judging that the machinery corresponding to the project fails and needs to be maintained in time;
judging the operation condition of the machine corresponding to the project through the obtained scores;
in this embodiment, the abnormal data in the collected data is not removed, so that errors are likely to occur when the project corresponding machinery is scored through the wind turbine generator fault scoring model, the project corresponding machinery is not timely maintained, further faults are likely to occur in the later work of the project corresponding machinery, and the working efficiency of the wind turbine generator is reduced.
Comparative example 2
The method comprises the steps that firstly, data of a wind turbine generator are collected in real time, the collected data comprise the temperature, vibration frequency and rotating speed inside the wind turbine generator, meanwhile, a monitoring camera is installed outside the wind turbine generator, and the temperature inside the wind turbine generator comprises the temperature of a gear box, the temperature of a bearing, the temperature of a generator winding and the temperature inside a cabin;
the internal rotating speed of the wind turbine generator comprises the rotating speed of an impeller and the rotating speed of a generator;
the second step, reject the abnormal data in the data collected, discern the picture that the monitoring camera shoots first before rejecting the abnormal data, judge whether the abnormal data has external factor to cause through the picture that is discerned, and then reject the abnormal data;
a monitoring camera is arranged outside the wind turbine generator to collect deflection angles of the blades, and at the same time, the monitoring camera shoots a main shaft of the wind turbine generator every 1s, and compares the main shaft with an initial main shaft image to judge whether the main shaft is bent or not;
and fifthly, comparing the detected data with fault models in a database, so as to judge the fault reason of the wind turbine generator.
In this embodiment, the existing fault model is not decomposed and deeply learned, the working condition of the parts corresponding to the summarized data is not passed, and the generated fault judgment of the parts is not corrected according to the use condition of the parts, so that only the generated fault can be judged when the wind turbine generator is faulty, but the aged equipment cannot be maintained in time, and the early warning of the parts is not performed.
The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.
Claims (8)
1. The wind turbine generator system fault early warning method based on the multi-model combination is characterized by comprising the following steps of:
step 1, acquiring historical operation data of a wind turbine generator;
step 2, correcting a preset fault judgment model by using historical operation data to obtain a fault scoring model corresponding to the wind turbine generator;
step 3, inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into a fault scoring model to obtain a fault score;
and 4, performing fault early warning on the wind turbine generator according to the obtained fault score.
2. The method for early warning of wind turbine generator faults based on multi-model combination according to claim 1, wherein in step 1, the real-time operation data comprise temperature, vibration frequency and rotation speed inside the wind turbine generator.
3. The method for early warning of wind turbine generator faults based on multi-model combinations according to claim 2, wherein the temperature inside the wind turbine generator comprises the temperature of a gearbox, the temperature of a bearing, the temperature of a generator winding and the temperature inside a nacelle.
4. The method for early warning of wind turbine faults based on multi-model combination according to claim 2, wherein the wind turbine rotation speed comprises an impeller rotation speed and a generator rotation speed.
5. The wind turbine generator system fault early warning method based on multi-model combination according to claim 1, wherein in step 2, a preset fault judgment model is provided, and the specific method is as follows:
acquiring an existing fault model;
splitting the obtained fault model to obtain data and corresponding faults;
and performing association learning according to the obtained data and the corresponding faults to obtain a fault judgment model.
6. The method for early warning of wind turbine generator system faults based on multi-model combination according to claim 1, wherein in step 2, a preset fault judgment model is corrected by using historical operation data to obtain a fault scoring model corresponding to the wind turbine generator system, and the specific method comprises the following steps:
acquiring the working state of the corresponding part according to the historical operation data;
and correcting the fault judgment model according to the working state of the part and the convolutional neural network to obtain a fault scoring model.
7. The method for early warning the faults of the wind turbine generator based on the multi-model combination according to claim 1 is characterized in that in the step 4, the fault early warning is carried out on the wind turbine generator according to the obtained fault score, and the specific method is as follows:
when k is more than or equal to 75, judging that the project is in a stable state, and stabilizing the project machinery;
when the k is more than or equal to 75 and less than or equal to 60, judging that the project is in a fatigue state, and maintaining the project machinery when the service life of the project machinery is reduced;
when k is less than 60, the machine corresponding to the project is judged to be faulty and needs to be maintained in time.
8. Wind turbine generator system trouble early warning system based on multimode combination, its characterized in that includes:
the data acquisition unit is used for acquiring historical operation data of the wind turbine generator;
the model correction unit is used for correcting a preset fault judgment model by utilizing the historical operation data to obtain a fault scoring model corresponding to the wind turbine generator;
the score calculating unit is used for inputting the obtained deflection angle and main shaft state information of the real-time running blades of the wind turbine generator into the fault scoring model to obtain a fault score;
and the early warning unit is used for carrying out fault early warning on the wind turbine generator according to the obtained fault score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310933648.3A CN116733692A (en) | 2023-07-27 | 2023-07-27 | Wind turbine generator system fault early warning method and system based on multi-model combination |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310933648.3A CN116733692A (en) | 2023-07-27 | 2023-07-27 | Wind turbine generator system fault early warning method and system based on multi-model combination |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116733692A true CN116733692A (en) | 2023-09-12 |
Family
ID=87908191
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310933648.3A Pending CN116733692A (en) | 2023-07-27 | 2023-07-27 | Wind turbine generator system fault early warning method and system based on multi-model combination |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116733692A (en) |
-
2023
- 2023-07-27 CN CN202310933648.3A patent/CN116733692A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105134510A (en) | State monitoring and failure diagnosis method for wind generating set variable pitch system | |
| CN106126843A (en) | A kind of Bladed blower fan load processing system based on Matlab | |
| CN108072524B (en) | Wind turbine generator gearbox bearing fault early warning method | |
| CN101672723A (en) | Method and system for analyzing vibration and diagnosing failure for wind generating set | |
| CN111287911A (en) | Wind turbine fatigue load early warning method and system | |
| CN111198098A (en) | Wind power generator bearing fault prediction method based on artificial neural network | |
| CN112228290A (en) | Intelligent early warning method for faults of wind turbine variable pitch system | |
| CN110851469A (en) | Fan running state analysis method based on new energy centralized control system | |
| CN114323665A (en) | Method for detecting faults of fuel supply system of gas turbine | |
| CN116733692A (en) | Wind turbine generator system fault early warning method and system based on multi-model combination | |
| CN113759876A (en) | Wind turbine generator fault diagnosis method and system based on case reasoning | |
| CN112283047B (en) | Load monitoring method and system based on wind turbine clearance monitoring | |
| CN112696481A (en) | Intelligent diagnosis method and device for shaft temperature abnormity of wind turbine generator gearbox | |
| CN115829411A (en) | Method and system for evaluating operation state of offshore wind turbine generator | |
| CN110705581A (en) | Pitch bearing fault identification method based on improved hidden Markov model | |
| CN105649896A (en) | Intelligent monitoring operation control system for wind turbine unit and control method thereof | |
| CN113818997B (en) | Wind turbine generator wind direction automatic correction method based on big data analysis | |
| Ye et al. | Using SCADA data fusion by swarm intelligence for wind turbine condition monitoring | |
| CN111639110A (en) | Wind turbine generator fault early warning method and device | |
| CN217384748U (en) | Fault trend diagnosis system for wind power plant group | |
| Liu et al. | Fuzzy fmea of floating wind turbine based on related weights and topsis theory | |
| CN116292140B (en) | Monitoring control method and monitoring control device for wind driven generator | |
| CN113374654B (en) | Method for detecting blade cracking of wind generating set | |
| CN113323820B (en) | Backup emergency safety chain control method and system for wind power generator | |
| CN112922778B (en) | Yaw optimization method, system, equipment and storage medium for wind turbine generator |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |