CN107179503A - The method of Wind turbines intelligent fault diagnosis early warning based on random forest - Google Patents

The method of Wind turbines intelligent fault diagnosis early warning based on random forest Download PDF

Info

Publication number
CN107179503A
CN107179503A CN201710266789.9A CN201710266789A CN107179503A CN 107179503 A CN107179503 A CN 107179503A CN 201710266789 A CN201710266789 A CN 201710266789A CN 107179503 A CN107179503 A CN 107179503A
Authority
CN
China
Prior art keywords
wind turbines
model
data
fault diagnosis
random forest
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.)
Granted
Application number
CN201710266789.9A
Other languages
Chinese (zh)
Other versions
CN107179503B (en
Inventor
程宏亮
刘宏
白朝旭
饶思维
张会
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Merrill Lynch Data Technology Ltd By Share Ltd
Original Assignee
Merrill Lynch Data Technology Ltd By Share Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Merrill Lynch Data Technology Ltd By Share Ltd filed Critical Merrill Lynch Data Technology Ltd By Share Ltd
Priority to CN201710266789.9A priority Critical patent/CN107179503B/en
Publication of CN107179503A publication Critical patent/CN107179503A/en
Application granted granted Critical
Publication of CN107179503B publication Critical patent/CN107179503B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/34—Testing dynamo-electric machines
    • G01R31/343—Testing dynamo-electric machines in operation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Testing And Monitoring For Control Systems (AREA)
  • Wind Motors (AREA)

Abstract

The invention discloses a kind of method of the Wind turbines intelligent fault diagnosis early warning based on random forest, including：The historical data for extracting Wind turbines state is used as sample data；Exploratory analysis and pretreatment are carried out to the sample data；The Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, is carried out analyzing and evaluation model according to model result；Real-time diagnosis is carried out to Wind turbines equipment using the model after assay, if diagnostic result is abnormal, the model will send warning information.The present invention uses random forests algorithm, it is considered to the overall feature of index, the problem of this method not only can solve single index decision device state, and can consider hiding knowledge correlation between numerous indexs, and comprehensive judgement is made to output result.

Description

The method of Wind turbines intelligent fault diagnosis early warning based on random forest
Technical field
It is that the Wind turbines intelligent fault based on random forest is examined more specifically the present invention relates to technical field of electric power The method of disconnected early warning.
Background technology
Wind-power electricity generation is one of closest commercialized renewable energy technologies generally acknowledged in the world.Protection is emphasized current Under environment, the background of sustainable development, consumption of fossil fuels, the wind-power electricity generation of non-environmental-pollution are not considered as the energy most cleaned Source utilizes form.In past 10 years, because annual average rate of increase is close to 28%, wind-power electricity generation turns into and increased in the world most Fast regenerative resource.
With the fast development of wind energy and putting into operation for large-scale wind power unit, and because most of units' installation is inclined Remote area, the factor such as load is unstable, many Wind turbines of China all occur in that operation troubles, directly affects wind-power electricity generation Security and economy.To keep the long-term stability development of wind-powered electricity generation, strengthen the competitiveness of it and traditional energy, it is necessary to improve constantly Wind power generation efficiency, reduction wind power equipment maintenance cost and operation cost, promote the economic interests of enterprise to maximize.
Traditional method for diagnosing faults, although employ certain scientific and technological means, but the sum of maintenance personal's technology The factor weight that experience is accounted for is substantially the experience for first relying on maintenance personal to the positioning of failure, is then being subject to section than larger Learn diagnosis of technique method to be accurately positioned, the place of failure problems is found, to a certain extent in the presence of artificial subjective errors With the shortcoming of extension unit maintenance time；
Also, because wind power plant their location is far away, relatively more severe, it is necessary to attendant's length plus natural environment Time value is kept, and as wind-powered electricity generation station quantity increases, the waste of personnel's capital is will result in a certain extent.
The content of the invention
Instant invention overcomes traditional manual method in trouble-shooting time length, deviation present on fault diagnosis and monitoring The problem of big and personnel waste.Therefore, the invention provides a kind of Wind turbines intelligent fault diagnosis based on random forest The method of early warning.
To achieve these goals, the present invention provides following technical scheme：
The method of Wind turbines intelligent fault diagnosis early warning based on random forest, including：
The historical data for extracting Wind turbines state is used as sample data；To the sample data carry out exploratory analysis and Pretreatment；
The Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, is analyzed simultaneously according to model result Evaluation model；
Real-time diagnosis is carried out to Wind turbines equipment using the model after assay, should if diagnostic result is abnormal Model will send warning information.
Preferably, the historical data of Wind turbines state is extracted as sample data, including：
By analyzing Wind turbines most common failure, Wind turbines fault indices system is built；
Wind turbines status history data conduct is extracted from Wind turbines automated system or equipment background control system Sample data.
Preferably, exploratory analysis and pretreatment are carried out to the sample data, including：
Exploratory analysis is carried out to the sample data, refers to reject aging equipment, equipment of just going into operation, tentatively examines wind-powered electricity generation closely The feature situation of unit normally with abnormality；
The sample data is pre-processed, including：Data cleansing, missing values processing, Data Discretization, attitude layer With data conversion.
Preferably, the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, including：
Pretreated sample data is divided into training sample and test sample, Random Forest model expert's sample is built Collection；
Wind turbines intelligent fault diagnosis Early-warning Model is built using sample set.
Preferably, analyze and evaluate Wind turbines intelligent fault diagnosis Early-warning Model, including：
To predict the outcome carry out analyzing and diagnosing of the model on test set,
If prediction is all correct, illustrate that the diagnosis effect of model is more satisfactory；
If diagnostic result has error, further Optimized model.
Preferably, real-time diagnosis is carried out to Wind turbines equipment using the model trained, including：
Online data is accessed from Wind turbines automated system or equipment background control system；
The attribute in Wind turbines fault indices system is selected from online data as input attribute；
Online data is pre-processed, the model that the online data of pretreatment is accessed after assay, real-time pair sets Standby data are monitored diagnosis.
Preferably, in addition to the more historical failures of Wind turbines and normal data are collected, periodically model are trained, Upgrade in time model.
Compared with prior art, beneficial effects of the present invention：
There is generation problem the present invention be directed to Wind turbines failure frequency, construct Wind turbines intelligent fault diagnosis early warning mould Type, study is trained using random forests algorithm to Wind turbines historical data, excavates influence Wind turbines failure and occurs Each factor feature mode, the threshold value for each index that breaks down is determined, so that according to equipment real time execution to Wind turbines On-line fault diagnosis early warning is carried out, Wind turbines maintenance cost is effectively reduced, the utilization ratio of Wind turbines is improved；
The present invention uses random forests algorithm, it is considered to which the overall feature of index, this method can not only solve single finger The problem of marking decision device state, and hiding knowledge correlation between numerous indexs can be considered, output result is done Go out comprehensive judgement.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are only this The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is the method for the Wind turbines intelligent fault diagnosis early warning provided in an embodiment of the present invention based on random forest Flow chart；
The schematic diagram for the decision tree that Fig. 2 provides for real-time example of the invention；
The schematic diagram for the random forest that Fig. 3 provides for real-time example of the invention；
Fig. 4 is in the method for the Wind turbines intelligent fault diagnosis early warning based on random forest that example is provided in real time of the invention Build the flow chart of the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest；
The graph of a relation of the false determination ratio and decision tree number of the random forest that Fig. 5 provides for real-time example of the invention；
Fig. 6 is in the method for the Wind turbines intelligent fault diagnosis early warning based on random forest that example is provided in real time of the invention Wind turbines intelligent fault diagnosis Early-warning Model based on random forest realizes the flow chart of real-time diagnosis.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made Embodiment, belongs to the scope of protection of the invention.
Embodiment 1：
The method of Wind turbines intelligent fault diagnosis early warning based on random forest, including：
Wind turbines fault indices system is built, Wind turbines status history data is extracted and is used as sample data；
Wind turbines state history sample data to extraction carries out Data Mining and pretreatment；
The Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, is analyzed simultaneously according to model result Evaluation model；
Real-time diagnosis is realized using the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest of structure；
The more historical failures of Wind turbines and normal data are collected, periodically model is trained, upgrade in time model.
It is preferred that, the structure Wind turbines fault indices system extracts Wind turbines status history data and is used as sample Data, refer to：
By analyzing Wind turbines most common failure, Wind turbines fault indices system is built；
Part Wind turbines state is selectively extracted from Wind turbines automated system, equipment background control system to go through History data.
It is preferred that, the Wind turbines state history sample data of described pair of extraction carries out Data Mining and pretreatment, including：
Exploratory analysis is carried out to sample, aging equipment, equipment of just going into operation is rejected, preliminary close examination Wind turbines normally with it is different The feature situation of normal state；
Sample is pre-processed, including data cleansing, missing values processing, Data Discretization, attitude layer and data become Change.
It is preferred that, the Wind turbines intelligent fault diagnosis Early-warning Model of the structure based on random forest, according to model knot Fruit analyze and evaluation model, including：
Build Random Forest model expert's sample set；
Build Wind turbines intelligent fault diagnosis Early-warning Model；
Model result is analyzed and evaluated, Wind turbines failure intelligence is further optimized according to analysis and the result evaluated Early-warning Model can be diagnosed.
It is preferred that, it is described to be realized in fact using the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest built When diagnose, including：
According to the Wind turbines intelligent fault diagnosis model (after the completion of training) of structure, by Wind turbines equipment-related data It is linked into model, is monitored diagnosis to device data in real time, once find that diagnostic result is " abnormal ", just to wind turbine Group sends warning information.
It is preferred that, the collection more historical failures of Wind turbines and normal data are periodically trained to model, and When more new model, including：
Collect the more historical failures of Wind turbines and normal data so that model covers the spy of all situations as far as possible Levy, and can regular one month to model training once, upgrade in time model, improve constantly model accuracy.
The invention discloses the method for the Wind turbines intelligent fault diagnosis early warning based on random forest, including：For wind Group of motors failure frequency has generation problem, Wind turbines intelligent fault diagnosis Early-warning Model is constructed, using random forests algorithm pair Wind turbines historical data is trained study, excavates the feature mode for each factor that influence Wind turbines failure occurs, really The threshold value for each index that breaks down is made, so that on-line fault diagnosis early warning is carried out to Wind turbines according to equipment real time execution, Wind turbines maintenance cost is effectively reduced, the utilization ratio of Wind turbines is improved.
It is that diversified, conventional research often focuses on list due to constituting running of wind generating set abnormal state The threshold range of individual index determines whether Wind turbines send warning information, and this method has certain defect, right herein Some key indexs (often direct decision device state) do not do independent research, but use random forests algorithm, it is considered to index Overall feature, the problem of this method not only can solve single index decision device state, and crowd can be considered Knowledge correlation is hidden between multi objective, comprehensive judgement is made to output result.
Embodiment 2：
Referring to Fig. 1, being examined it illustrates the Wind turbines intelligent fault provided in an embodiment of the present invention based on random forest The flow chart of the method for disconnected early warning, can include：
S101：Wind turbines fault indices system is built, Wind turbines status data is extracted and is used as sample data.
By analyzing Wind turbines most common failure, 13 attributes (gear_temp, gear_rate are chosen Rotor_vol ..., motor_temp_sd, power_mean, is_running) as input attribute, just whether Wind turbines Often operation builds Wind turbines fault indices system as output attribute；
Part Wind turbines state is selectively extracted from Wind turbines automated system, equipment background control system to go through History data are used as sample data.It may be noted that selectively extraction part Wind turbines status history data is due in this step The achievement data that some need in equipment background control system may be imperfect, the selection data integrity so we should try one's best Compare high data as sample data.
S102：Wind turbines state history sample data to extraction carries out Data Mining and pretreatment.
Exploratory analysis is carried out to sample, aging equipment, equipment of just going into operation is rejected, preliminary close examination Wind turbines normally with it is different The feature situation of normal state；
Sample is pre-processed, including data cleansing, missing values processing, Data Discretization, attitude layer and data become Change.
S103：The Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, is carried out according to model result Analyze and evaluation model.
Build Random Forest model expert's sample set；
Build Wind turbines intelligent fault diagnosis Early-warning Model；
Model result is analyzed and evaluated, Wind turbines failure intelligence is further optimized according to analysis and the result evaluated Early-warning Model can be diagnosed.
S104：Realized and examined in real time using the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest of structure It is disconnected.
According to the Wind turbines intelligent fault diagnosis model (after the completion of training) of structure, by Wind turbines equipment-related data It is linked into model, is monitored diagnosis to device data in real time, once find that diagnostic result is " abnormal ", just to wind turbine Group sends warning information.
S105：The more historical failures of Wind turbines and normal data are collected, periodically model is trained, upgraded in time Model.
Collect the more historical failures of Wind turbines and normal data so that model covers the spy of all situations as far as possible Levy, and can regular one month to model training once, upgrade in time model, improve constantly model accuracy.
In the present embodiment, optionally, the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built.As schemed Shown in 6, including：
Random Forest model expert's sample set is built first, including：
S201：Parameter setting：Random forests algorithm chooses 500 trees, using stratified sampling by S102 pretreatments Sample is divided into training sample and test sample accounting is (0.8,0.2).
For stratified sampling in S201, it is necessary to which what is illustrated is：
Stratified sampling is to combine Scientific grouping method with sampling, is specifically that will totally be divided into several homogeneities Layer, then random sampling or mechanical sampling in each layer, stratified sampling ensure that extracted sample has enough representativenesses.
S202：500 sample sets of generation are concentrated from S201 training sample using the random sampling pattern that can be put back to, often The number of samples of individual sample set is identical with the number of samples of S201 training sample set, in theory 500 sample set coverings Original sample concentrates 2/3 data instance, and the data not included are referred to as the outer data (Out-Of-Bag, OOB) of bag, and the outer data of bag can For as test data, estimation can be very good to assess the classifying quality of the assembled classifier in random forests algorithm.
For being carried out in S202 to data after the random sampling that can put back to, sample set covers original sample concentration in theory 2/3 instance data is, it is necessary to which what is illustrated is：
Assuming that carrying out having the stochastical sampling put back to a number of samples for N sample set D, n times are extracted, each sample is not The probability being extracted is (1-1/N)N, when N is sufficiently large (1-1/N)N1/e ≈ 0.368 will be converged on, therefore it is considered that sample is sub Collection covers the instance data of original sample 2/3.
Then Wind turbines intelligent fault diagnosis Early-warning Model is built, including：
S203：500 decision trees are grown using 500 self-service sample sets of generation.Here in each node of each tree On, it is random from 13 features respectively to select m (m<=13) individual feature, generally takes in Practical ProjectOften It is secondary to select a feature to carry out branch according to certain rule (information gain) from this randomly selected m feature, until this Tree fully growth, cut operator is not done therebetween.
For the comentropy told in S203 with information gain, it is necessary to which explanation is：
Comentropy：The uncertainty (confusion degree) of information is illustrated, entropy is bigger, and information is more chaotic, be more difficult to prediction, then should The information content that index is provided is smaller, and the weight of the index is smaller, more inessential.For categorizing system, classification C is variable, it Possible value is C1,C2,…,Cn, and the probability that each classification occurs is P (C1),P(C2),…,P(Cn), therefore n is exactly The sum of classification.Now the entropy of categorizing system can be just expressed as：
Information gain is for feature one by one, exactly to see a feature t, system have it and do not have it when It is respectively how many to wait information content, and both differences are exactly the information content that this feature is brought to system, i.e. gain.System contains feature Information content is exactly formula above when t, the information content of system when what it was represented is comprising all features.
InfoGain=H (Y)-H (Y | X)
In categorizing system, the selection of attribute and the division of decision tree are selected according to information gain, for root Node and the maximum attribute variable of child node selection information gain, then using recursive method build whole decision tree and with Machine forest.
S204：Test sample collection is predicted according to the 500 of above-mentioned generation decision trees, the test knot of comprehensive each tree Fruit determines final result according to certain voting mechanism.The good of random forests algorithm utilizes randomness (including random generation Subsample collection, randomly chooses subcharacter), the correlation between each tree is minimized, overall classification performance is improved, together When, because the generation time of each tree is very short, and forest can realize parallelization, and the classification speed of random forest is very It hurry up.Assuming that random forest grader { hi(x,θi, i=1 ..., N) }, the class label of classification results is exactly by each decision tree hi (x,θi) and probability averagely obtain, for test case x, prediction class label cp, then,
Wherein, argmaxcRepresent that finding parameter c, N with maximum scores represents the number of decision tree in random forest, I (*) represents performance function,Classification results of the decision tree to C classifications are represented,Represent decision tree hiLeaf node number Mesh, WiRepresent the weight of i-th tree in random forest.
Model result is as follows：
The model output result of table 1
As can be seen from the above table, whether normally run for Wind turbines, it is 2.7%, explanation to wrap outer data error rate OOB The overall classifying quality of model is ideal.Fig. 5 be random forest OOB false determination ratios and decision tree number, as can be seen from Figure 5 with Machine forest false determination ratio constantly reduces with the increase of decision tree number, finally converges to a less definite value.Mark in Fig. 5 Number negative sample error rate is represented for 1 dotted line, the dotted line marked as 2 represents positive sample error rate, and solid black lines represent overall mistake Rate.
Finally, model result is analyzed and evaluated, further optimization is decided whether according to analysis and the result evaluated Wind turbines intelligent fault diagnosis Early-warning Model, including：
S205：To predict the outcome carry out analyzing and diagnosing, Model Diagnosis interpretation of result of the model on test set：
Table 2：Test data diagnostic result
Diagnostic analysis is carried out by 8 datas to test set, prediction is all correct, illustrates that the diagnosis effect of model compares Ideal, without further optimization Wind turbines intelligent fault diagnosis Early-warning Model.If the undesirable feelings of the diagnosis effect of model Under condition, can by adjust decision tree number, tree depth capacity, the parameter such as measure information mode and feature selection approach Setting come further Optimized model.It can learn：According to test result, Isosorbide-5-Nitrae, 6,8 Wind turbines are normally run, and the 2nd, 3,5,7 Wind turbines send warning information, and should start corresponding prediction scheme measure at once, prevent bigger safety Accident and economic loss.
In above-described embodiment, optionally, the Wind turbines intelligent fault diagnosis early warning based on random forest of structure is utilized Model realization real-time diagnosis.As shown in fig. 6, including：
S301：Online data is accessed from Wind turbines automated system, equipment background control system.
S302：Selected from online data in Wind turbines fault indices system 13 attributes (
Gear_temp, gear_rate, rotor_vol ..., motor_temp_sd, power_mean, is_running) It is used as input attribute.
S303：Online data is pre-processed, including data cleansing, missing values processing, Data Discretization, attitude layer With data conversion etc..
S304：Pretreated online data is linked into model, diagnosis is monitored to device data in real time, once It was found that diagnostic result is " abnormal ", just Wind turbines are sent with warning information, staff can be directed to parameters at once Take corresponding shutdown measure and alternative.
The foregoing description of the disclosed embodiments, enables those skilled in the art to realize or using the present invention.To this A variety of modifications of a little embodiments will be apparent for a person skilled in the art, and generic principles defined herein can Without departing from the spirit or scope of the present invention, to realize in other embodiments.Therefore, the present invention will not be limited It is formed on the embodiments shown herein, and is to fit to consistent with features of novelty with principles disclosed herein most wide Scope.
Above content is to combine specific preferred embodiment further description made for the present invention, it is impossible to assert The specific implementation of the present invention is confined to these explanations.For general technical staff of the technical field of the invention, On the premise of not departing from present inventive concept, some simple deduction or replace can also be made, should all be considered as belonging to the present invention's Protection domain.

Claims (7)

1. the method for the Wind turbines intelligent fault diagnosis early warning based on random forest, it is characterised in that including：
The historical data for extracting Wind turbines state is used as sample data；Exploratory analysis and pre- place are carried out to the sample data Reason；
The Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, is analyzed and is evaluated according to model result Model；
Real-time diagnosis is carried out to Wind turbines equipment using the model after assay, if diagnostic result is abnormal, the model Warning information will be sent.
2. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, the historical data of Wind turbines state is extracted as sample data, including：
By analyzing Wind turbines most common failure, Wind turbines fault indices system is built；
Wind turbines status history data is extracted from Wind turbines automated system or equipment background control system and is used as sample Data.
3. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, exploratory analysis and pretreatment are carried out to the sample data, including：
Exploratory analysis is carried out to the sample data, refers to reject aging equipment, equipment of just going into operation, tentatively examines Wind turbines closely Normally with the feature situation of abnormality；
The sample data is pre-processed, including：Data cleansing, missing values processing, Data Discretization, attitude layer sum According to conversion.
4. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, the Wind turbines intelligent fault diagnosis Early-warning Model based on random forest is built, including：
Pretreated sample data is divided into training sample and test sample, Random Forest model expert's sample set is built；
Wind turbines intelligent fault diagnosis Early-warning Model is built using sample set.
5. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, analyze and evaluate Wind turbines intelligent fault diagnosis Early-warning Model, including：
To predict the outcome carry out analyzing and diagnosing of the model on test set,
If prediction is all correct, illustrate that the diagnosis effect of model is more satisfactory；
If diagnostic result has error, further Optimized model.
6. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, real-time diagnosis is carried out to Wind turbines equipment using the model trained, including：
Online data is accessed from Wind turbines automated system or equipment background control system；
The attribute in Wind turbines fault indices system is selected from online data as input attribute；
Online data is pre-processed, the model that the online data of pretreatment is accessed after assay, in real time to number of devices According to being monitored diagnosis.
7. the method for the Wind turbines intelligent fault diagnosis early warning as claimed in claim 1 based on random forest, its feature exists In, in addition to the more historical failures of Wind turbines and normal data are collected, periodically model is trained, upgrade in time mould Type.
CN201710266789.9A 2017-04-21 2017-04-21 Wind turbine generator fault intelligent diagnosis and early warning method based on random forest Active CN107179503B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710266789.9A CN107179503B (en) 2017-04-21 2017-04-21 Wind turbine generator fault intelligent diagnosis and early warning method based on random forest

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710266789.9A CN107179503B (en) 2017-04-21 2017-04-21 Wind turbine generator fault intelligent diagnosis and early warning method based on random forest

Publications (2)

Publication Number Publication Date
CN107179503A true CN107179503A (en) 2017-09-19
CN107179503B CN107179503B (en) 2020-07-07

Family

ID=59832129

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710266789.9A Active CN107179503B (en) 2017-04-21 2017-04-21 Wind turbine generator fault intelligent diagnosis and early warning method based on random forest

Country Status (1)

Country Link
CN (1) CN107179503B (en)

Cited By (42)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107782442A (en) * 2017-10-24 2018-03-09 华北电力大学（保定） Transformer multiple features parameter selection method based on big data and random forest
CN107958395A (en) * 2017-12-13 2018-04-24 美林数据技术股份有限公司 A kind of recognition methods of electric system abnormal user
CN107967485A (en) * 2017-11-16 2018-04-27 国网冀北电力有限公司电力科学研究院 Electro-metering equipment fault analysis method and device
CN108011367A (en) * 2017-12-04 2018-05-08 贵州电网有限责任公司电力科学研究院 A kind of Characteristics of Electric Load method for digging based on depth decision Tree algorithms
CN108020785A (en) * 2017-12-14 2018-05-11 海安常州大学高新技术研发中心 A kind of electrical fault forecasting system and data managing method based on micromainframe
CN108535618A (en) * 2018-07-11 2018-09-14 云南电网有限责任公司电力科学研究院 A kind of GIS method for detecting insulation defect
CN109283378A (en) * 2018-08-30 2019-01-29 番禺珠江钢管（珠海）有限公司 A kind of rotating arc welding is seamed into shape parameter detection method, system, device and medium
CN109297689A (en) * 2018-09-11 2019-02-01 中国水利水电科学研究院 A kind of large-scale hydraulic intelligent diagnosing method introducing weight factor
CN109522599A (en) * 2018-10-16 2019-03-26 国电南瑞科技股份有限公司 Transmission line of electricity catastrophic failure method for early warning caused by a kind of typhoon
CN109543210A (en) * 2018-09-28 2019-03-29 国电电力宁夏新能源开发有限公司 A kind of Wind turbines failure prediction system based on machine learning algorithm platform
CN109544035A (en) * 2018-12-12 2019-03-29 上海理工大学 Electric energy efficiency analysis and ranking method based on random forest
CN109800127A (en) * 2019-01-03 2019-05-24 众安信息技术服务有限公司 A kind of system fault diagnosis intelligence O&M method and system based on machine learning
CN109840666A (en) * 2017-11-29 2019-06-04 中国电力科学研究院有限公司 A kind of model building method and system for predicting that the following Wind turbines break down
CN109855879A (en) * 2019-01-26 2019-06-07 厦门华夏国际电力发展有限公司 A kind of steam turbine servo mechanism On-line Fault Detection method for early warning and system
CN109887122A (en) * 2019-02-19 2019-06-14 南京洛普股份有限公司 A kind of ship lock functional failure of electromechanical long-distance intelligent diagnostic method and system
CN109936113A (en) * 2019-03-29 2019-06-25 国网浙江省电力有限公司 A kind of protection act intelligent diagnosing method and system based on random forests algorithm
CN110007232A (en) * 2019-05-23 2019-07-12 广东工业大学 A kind of prediction technique and relevant apparatus of squirrel cage asynchronous motor operational efficiency
CN110188834A (en) * 2019-06-04 2019-08-30 广东电网有限责任公司 A kind of method for diagnosing faults of power telecom network, device and equipment
CN110378427A (en) * 2019-07-23 2019-10-25 上海电气风电集团有限公司 Fault detection method, system, equipment and the medium of the blade root bolt of wind electricity blade
CN110657072A (en) * 2018-06-29 2020-01-07 中车株洲电力机车研究所有限公司 Wind power fault maintenance method and system based on structured knowledge base
CN110968069A (en) * 2018-09-28 2020-04-07 新疆金风科技股份有限公司 Fault prediction method of wind generating set, corresponding device and electronic equipment
CN111047215A (en) * 2019-12-09 2020-04-21 中国兵器科学研究院 Random forest based field replaceable unit classification and classification determination method
CN111080466A (en) * 2019-12-03 2020-04-28 广东工业大学 Calender calendering quality early warning system based on big data
CN111260201A (en) * 2020-01-13 2020-06-09 北京科技大学 Variable importance analysis method based on hierarchical random forest
CN111310786A (en) * 2020-01-15 2020-06-19 青岛海信网络科技股份有限公司 Traffic detector abnormity diagnosis method and device based on random forest classifier
CN111611146A (en) * 2020-06-18 2020-09-01 南方电网科学研究院有限责任公司 Micro-service fault prediction method and device
CN111724184A (en) * 2019-03-20 2020-09-29 北京嘀嘀无限科技发展有限公司 Transformation probability prediction method and device
CN112308146A (en) * 2020-11-02 2021-02-02 国网福建省电力有限公司 Distribution transformer fault identification method based on operation characteristics
CN112398218A (en) * 2020-09-28 2021-02-23 国网山东省电力公司冠县供电公司 Power grid safety and stability analysis device and method
CN112446144A (en) * 2020-11-17 2021-03-05 哈工大机器人（合肥）国际创新研究院 Fault diagnosis method and device for large-scale rotating machine set
CN112561176A (en) * 2020-12-21 2021-03-26 深圳供电局有限公司 Early warning method for online running state of electric power metering device
CN112578794A (en) * 2020-12-12 2021-03-30 云南昆船智能装备有限公司 AGV fault detection method based on machine learning, storage medium and system
CN112765878A (en) * 2021-01-11 2021-05-07 龙源（北京）风电工程技术有限公司 Intelligent diagnosis method and device for abnormity of generator bearing of wind turbine generator
CN113107785A (en) * 2021-05-12 2021-07-13 浙江浙能技术研究院有限公司 Real-time monitoring method and device for power performance abnormity of wind turbine generator
CN113869578A (en) * 2021-09-26 2021-12-31 合肥通用机械研究院有限公司 Intelligent prediction and diagnosis method for salt content of crude oil after removal of electric desalting system of atmospheric and vacuum distillation unit
CN114483479A (en) * 2021-12-23 2022-05-13 中国大唐集团新能源科学技术研究院有限公司 Fan high-temperature capacity reduction state evaluation method based on random forest
US20220291084A1 (en) * 2021-03-12 2022-09-15 General Electric Renovables Espana, S.L. System and method for detecting actual slip in a coupling of a rotary shaft
CN115079626A (en) * 2022-07-21 2022-09-20 东方电气风电股份有限公司 Early warning method and system for potential operation risk of wind generating set component
WO2023015484A1 (en) * 2021-08-11 2023-02-16 Siemens Aktiengesellschaft Method and apparatus for training a model
CN116664096A (en) * 2022-08-23 2023-08-29 国家电投集团科学技术研究院有限公司 Wind power bolt data processing method and device based on federal learning
WO2023182944A1 (en) * 2022-03-25 2023-09-28 Nanyang Technological University Alternator monitoring systems and methods
CN117436569A (en) * 2023-09-18 2024-01-23 华能核能技术研究院有限公司 Nuclear power equipment fault prediction and intelligent calibration method and system based on random forest

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104747368A (en) * 2015-01-27 2015-07-01 风脉（武汉）可再生能源技术有限责任公司 Method and system for optimizing power of wind turbine generator
CN105069476A (en) * 2015-08-10 2015-11-18 国网宁夏电力公司 Method for identifying abnormal wind power data based on two-stage integration learning

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104747368A (en) * 2015-01-27 2015-07-01 风脉（武汉）可再生能源技术有限责任公司 Method and system for optimizing power of wind turbine generator
CN105069476A (en) * 2015-08-10 2015-11-18 国网宁夏电力公司 Method for identifying abnormal wind power data based on two-stage integration learning

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
温延立: "风电机组监测数据可视分析研究", 《东北电力大学硕士学位论文》 *

Cited By (58)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107782442A (en) * 2017-10-24 2018-03-09 华北电力大学（保定） Transformer multiple features parameter selection method based on big data and random forest
CN107967485A (en) * 2017-11-16 2018-04-27 国网冀北电力有限公司电力科学研究院 Electro-metering equipment fault analysis method and device
CN107967485B (en) * 2017-11-16 2024-05-14 国网冀北电力有限公司电力科学研究院 Fault analysis method and device for electricity metering equipment
CN109840666A (en) * 2017-11-29 2019-06-04 中国电力科学研究院有限公司 A kind of model building method and system for predicting that the following Wind turbines break down
CN109840666B (en) * 2017-11-29 2020-12-18 中国电力科学研究院有限公司 Model construction method and system for predicting future wind turbine generator failure
CN108011367A (en) * 2017-12-04 2018-05-08 贵州电网有限责任公司电力科学研究院 A kind of Characteristics of Electric Load method for digging based on depth decision Tree algorithms
CN108011367B (en) * 2017-12-04 2020-12-18 贵州电网有限责任公司电力科学研究院 Power load characteristic mining method based on depth decision tree algorithm
CN107958395A (en) * 2017-12-13 2018-04-24 美林数据技术股份有限公司 A kind of recognition methods of electric system abnormal user
CN107958395B (en) * 2017-12-13 2021-11-26 美林数据技术股份有限公司 Method for identifying abnormal users of power system
CN108020785A (en) * 2017-12-14 2018-05-11 海安常州大学高新技术研发中心 A kind of electrical fault forecasting system and data managing method based on micromainframe
CN110657072A (en) * 2018-06-29 2020-01-07 中车株洲电力机车研究所有限公司 Wind power fault maintenance method and system based on structured knowledge base
CN108535618A (en) * 2018-07-11 2018-09-14 云南电网有限责任公司电力科学研究院 A kind of GIS method for detecting insulation defect
CN109283378A (en) * 2018-08-30 2019-01-29 番禺珠江钢管（珠海）有限公司 A kind of rotating arc welding is seamed into shape parameter detection method, system, device and medium
CN109297689A (en) * 2018-09-11 2019-02-01 中国水利水电科学研究院 A kind of large-scale hydraulic intelligent diagnosing method introducing weight factor
CN109543210A (en) * 2018-09-28 2019-03-29 国电电力宁夏新能源开发有限公司 A kind of Wind turbines failure prediction system based on machine learning algorithm platform
CN110968069B (en) * 2018-09-28 2022-01-25 新疆金风科技股份有限公司 Fault prediction method of wind generating set, corresponding device and electronic equipment
CN110968069A (en) * 2018-09-28 2020-04-07 新疆金风科技股份有限公司 Fault prediction method of wind generating set, corresponding device and electronic equipment
CN109522599A (en) * 2018-10-16 2019-03-26 国电南瑞科技股份有限公司 Transmission line of electricity catastrophic failure method for early warning caused by a kind of typhoon
CN109544035A (en) * 2018-12-12 2019-03-29 上海理工大学 Electric energy efficiency analysis and ranking method based on random forest
CN109800127A (en) * 2019-01-03 2019-05-24 众安信息技术服务有限公司 A kind of system fault diagnosis intelligence O&M method and system based on machine learning
CN109855879A (en) * 2019-01-26 2019-06-07 厦门华夏国际电力发展有限公司 A kind of steam turbine servo mechanism On-line Fault Detection method for early warning and system
CN109887122A (en) * 2019-02-19 2019-06-14 南京洛普股份有限公司 A kind of ship lock functional failure of electromechanical long-distance intelligent diagnostic method and system
CN111724184A (en) * 2019-03-20 2020-09-29 北京嘀嘀无限科技发展有限公司 Transformation probability prediction method and device
CN109936113B (en) * 2019-03-29 2021-06-01 国网浙江省电力有限公司 Protection action intelligent diagnosis method and system based on random forest algorithm
CN109936113A (en) * 2019-03-29 2019-06-25 国网浙江省电力有限公司 A kind of protection act intelligent diagnosing method and system based on random forests algorithm
CN110007232A (en) * 2019-05-23 2019-07-12 广东工业大学 A kind of prediction technique and relevant apparatus of squirrel cage asynchronous motor operational efficiency
CN110007232B (en) * 2019-05-23 2021-09-03 广东工业大学 Method and related device for predicting running efficiency of squirrel-cage asynchronous motor
CN110188834A (en) * 2019-06-04 2019-08-30 广东电网有限责任公司 A kind of method for diagnosing faults of power telecom network, device and equipment
CN110378427A (en) * 2019-07-23 2019-10-25 上海电气风电集团有限公司 Fault detection method, system, equipment and the medium of the blade root bolt of wind electricity blade
CN111080466A (en) * 2019-12-03 2020-04-28 广东工业大学 Calender calendering quality early warning system based on big data
CN111047215A (en) * 2019-12-09 2020-04-21 中国兵器科学研究院 Random forest based field replaceable unit classification and classification determination method
CN111047215B (en) * 2019-12-09 2023-06-23 中国兵器科学研究院 Method for determining classification of field replaceable units based on random forest
CN111260201A (en) * 2020-01-13 2020-06-09 北京科技大学 Variable importance analysis method based on hierarchical random forest
CN111260201B (en) * 2020-01-13 2023-04-28 北京科技大学 Variable importance analysis method based on layered random forest
CN111310786B (en) * 2020-01-15 2023-07-25 青岛海信网络科技股份有限公司 Traffic detector abnormality diagnosis method and device based on random forest classifier
CN111310786A (en) * 2020-01-15 2020-06-19 青岛海信网络科技股份有限公司 Traffic detector abnormity diagnosis method and device based on random forest classifier
CN111611146A (en) * 2020-06-18 2020-09-01 南方电网科学研究院有限责任公司 Micro-service fault prediction method and device
CN111611146B (en) * 2020-06-18 2023-05-16 南方电网科学研究院有限责任公司 Micro-service fault prediction method and device
CN112398218A (en) * 2020-09-28 2021-02-23 国网山东省电力公司冠县供电公司 Power grid safety and stability analysis device and method
CN112308146A (en) * 2020-11-02 2021-02-02 国网福建省电力有限公司 Distribution transformer fault identification method based on operation characteristics
CN112446144A (en) * 2020-11-17 2021-03-05 哈工大机器人（合肥）国际创新研究院 Fault diagnosis method and device for large-scale rotating machine set
CN112578794B (en) * 2020-12-12 2023-09-01 云南昆船智能装备有限公司 AGV fault detection method, storage medium and system based on machine learning
CN112578794A (en) * 2020-12-12 2021-03-30 云南昆船智能装备有限公司 AGV fault detection method based on machine learning, storage medium and system
CN112561176A (en) * 2020-12-21 2021-03-26 深圳供电局有限公司 Early warning method for online running state of electric power metering device
CN112765878A (en) * 2021-01-11 2021-05-07 龙源（北京）风电工程技术有限公司 Intelligent diagnosis method and device for abnormity of generator bearing of wind turbine generator
US20220291084A1 (en) * 2021-03-12 2022-09-15 General Electric Renovables Espana, S.L. System and method for detecting actual slip in a coupling of a rotary shaft
US11774324B2 (en) * 2021-03-12 2023-10-03 General Electric Renovables Espana, S.L. System and method for detecting actual slip in a coupling of a rotary shaft
CN113107785B (en) * 2021-05-12 2022-05-31 浙江浙能技术研究院有限公司 Real-time monitoring method and device for power performance abnormity of wind turbine generator
CN113107785A (en) * 2021-05-12 2021-07-13 浙江浙能技术研究院有限公司 Real-time monitoring method and device for power performance abnormity of wind turbine generator
WO2023015484A1 (en) * 2021-08-11 2023-02-16 Siemens Aktiengesellschaft Method and apparatus for training a model
CN113869578A (en) * 2021-09-26 2021-12-31 合肥通用机械研究院有限公司 Intelligent prediction and diagnosis method for salt content of crude oil after removal of electric desalting system of atmospheric and vacuum distillation unit
CN114483479B (en) * 2021-12-23 2023-09-12 大唐可再生能源试验研究院有限公司 Fan Gao Wenjiang capacity state assessment method based on random forest
CN114483479A (en) * 2021-12-23 2022-05-13 中国大唐集团新能源科学技术研究院有限公司 Fan high-temperature capacity reduction state evaluation method based on random forest
WO2023182944A1 (en) * 2022-03-25 2023-09-28 Nanyang Technological University Alternator monitoring systems and methods
CN115079626A (en) * 2022-07-21 2022-09-20 东方电气风电股份有限公司 Early warning method and system for potential operation risk of wind generating set component
CN116664096A (en) * 2022-08-23 2023-08-29 国家电投集团科学技术研究院有限公司 Wind power bolt data processing method and device based on federal learning
CN116664096B (en) * 2022-08-23 2024-02-13 国家电投集团科学技术研究院有限公司 Wind power bolt data processing method and device based on federal learning
CN117436569A (en) * 2023-09-18 2024-01-23 华能核能技术研究院有限公司 Nuclear power equipment fault prediction and intelligent calibration method and system based on random forest

Also Published As

Publication number Publication date
CN107179503B (en) 2020-07-07

Similar Documents

Publication Publication Date Title
CN107179503A (en) The method of Wind turbines intelligent fault diagnosis early warning based on random forest
Hsu et al. Wind turbine fault diagnosis and predictive maintenance through statistical process control and machine learning
CN110929918B (en) 10kV feeder fault prediction method based on CNN and LightGBM
CN109657789A (en) Gear case of blower failure trend prediction method based on wavelet neural network
CN106124982A (en) Automatic expert&#39;s resultant fault diagnostic system of a kind of Wind turbines and diagnostic method
CN110674189B (en) Method for monitoring secondary state and positioning fault of intelligent substation
CN108876163A (en) The transient rotor angle stability fast evaluation method of comprehensive causality analysis and machine learning
CN105628425A (en) Rotation machinery early stage fault diagnosis method based on heredity annealing optimization multi-core support vector machine
CN109501834A (en) A kind of point machine failure prediction method and device
CN104952000A (en) Wind turbine operating state fuzzy synthetic evaluation method based on Markov chain
CN109492790A (en) Wind turbines health control method based on neural network and data mining
CN110245795A (en) A kind of method and system for the simulation of offshore wind farm O&amp;M Simulation on Decision
CN106485594A (en) A kind of main distribution integration incident response decision method
CN107705054A (en) Meet the new energy grid-connected power remote measuring and diagnosing platform and method of complex data
CN110297207A (en) Method for diagnosing faults, system and the electronic device of intelligent electric meter
CN113738595A (en) Method and system for monitoring state of blade of wind generating set
CN106709607A (en) Intelligent operation and maintenance monitoring system of distributed photovoltaic power station
CN106768933A (en) A kind of wind power generation unit blade method for diagnosing faults based on genetic algorithm
CN116308304A (en) New energy intelligent operation and maintenance method and system based on meta learning concept drift detection
CN113641486B (en) Intelligent turnout fault diagnosis method based on edge computing network architecture
Wang et al. Tracking and predicting technological knowledge interactions between artificial intelligence and wind power: Multimethod patent analysis
KR20220089853A (en) Method for Failure prediction and prognostics and health management of renewable energy generation facilities using machine learning technology
CN114320773B (en) Wind turbine generator system fault early warning method based on power curve analysis and neural network
CN115563885A (en) Energy conversion system for energy storage battery of wind power plant
CN114358160A (en) Data anomaly detection method in power system

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
GR01 Patent grant
GR01 Patent grant
CP02 Change in the address of a patent holder
CP02 Change in the address of a patent holder

Address after: 710075 6th floor, national e-commerce demonstration base, No.528, tianguba Road, software new town, high tech Zone, Xi'an City, Shaanxi Province

Patentee after: MERITDATA TECHNOLOGY CO.,LTD.

Address before: 710000 C501, Hanyun Pavilion, Xi'an Software Park, No.68, Keji 2nd Road, high tech Zone, Xi'an City, Shaanxi Province

Patentee before: MERITDATA TECHNOLOGY CO.,LTD.