CN110334951B - Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator - Google Patents
Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator Download PDFInfo
- Publication number
- CN110334951B CN110334951B CN201910604545.6A CN201910604545A CN110334951B CN 110334951 B CN110334951 B CN 110334951B CN 201910604545 A CN201910604545 A CN 201910604545A CN 110334951 B CN110334951 B CN 110334951B
- Authority
- CN
- China
- Prior art keywords
- copula
- vine
- bayesian network
- network model
- characteristic variables
- 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.)
- Expired - Fee Related
Links
- 230000009467 reduction Effects 0.000 title claims abstract description 46
- 238000011156 evaluation Methods 0.000 title claims abstract description 34
- 241000039077 Copula Species 0.000 claims abstract description 37
- 238000012360 testing method Methods 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000001514 detection method Methods 0.000 claims description 20
- 238000005315 distribution function Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 5
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 2
- 230000000875 corresponding effect Effects 0.000 description 12
- 238000012423 maintenance Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 239000003921 oil Substances 0.000 description 7
- 238000009826 distribution Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000010219 correlation analysis Methods 0.000 description 4
- 238000010248 power generation Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000005251 gamma ray Effects 0.000 description 2
- 230000009347 mechanical transmission Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001364 causal effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 239000010687 lubricating oil Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Public Health (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Game Theory and Decision Science (AREA)
- Wind Motors (AREA)
Abstract
The invention discloses an intelligent evaluation method for a high-temperature capacity reduction state of a wind turbine generator, which comprises the following steps of: collecting multi-dimensional characteristic variables of the wind turbine generator, and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables; establishing a pair-Copula function according to the training data set to determine an optimal Copula function between every two multidimensional characteristic variables; obtaining correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function, constructing a relation tree based on the correlation coefficients, and generating a vine-Copula Bayesian network model by using the relation tree; and evaluating the test samples in the state point test sample set by using the vine-Copula Bayesian network model. The invention also discloses an intelligent evaluation system for the high-temperature capacity reduction state of the wind turbine generator. According to the method, the vine-Copula Bayesian network model is utilized to comprehensively consider the state parameters and the relativity of all dimensions of the characteristic variables of the wind turbine generator, and the occurrence probability of the high-temperature capacity reduction state of the wind turbine generator can be accurately evaluated.
Description
Technical Field
The invention relates to the technical field of wind power generation, in particular to a method and a system for intelligently evaluating a high-temperature capacity reduction state of a wind turbine generator.
Background
The wind turbine generator system has a complex working environment and is easy to be influenced by environmental factors as a mechanical transmission system, such as randomly changed wind speed and large temperature fluctuation range, so that various system parts can not operate under stable working conditions, the wind turbine generator system can operate in a sub-health state in certain time periods, the wind turbine generator system can be stopped, the output of the wind turbine generator system and the generated energy of the wind turbine generator system can be reduced, and the economic benefit of a wind turbine generator system operator can be influenced. In order to reduce the power generation loss of the wind turbine generator in the sub-health state, the states of the wind turbine generator need to be judged and evaluated, so that a more targeted operation maintenance method is formulated, and various faults caused by further deterioration of the states of the wind turbine generator are avoided.
Generally, the faults of the doubly-fed wind turbine are mostly generated on equipment components such as a gear box, a blade and an electrical system, and the cost of the components accounts for more than 80% of the cost of the wind turbine; and the gear box breaks down to cause the long-time shutdown of the wind turbine generator, and the maintenance time and the maintenance cost after the shutdown are also higher. Particularly, in summer, the output of the wind turbine generator is frequently reduced due to the fact that the oil temperature of the gear box is too high (a high-temperature capacity reduction state for short), and the high-temperature capacity reduction state frequently occurs to the wind turbine generator, so that the chemical performance of lubricating oil is reduced, the tooth surface is easily abraded and damaged, and the normal operation of the wind turbine generator is influenced.
At present, when a fan on site meets the condition, a power-over-limit and power-down operation instruction is generally issued to a unit through simple threshold judgment, but the unit is delayed in action when meeting high temperature due to the fact that temperature and temperature sensors have delay characteristics, and the influence of the high temperature of a gear box on the output of the unit is further expanded.
The existing wind generating set and a power control method thereof comprise the following steps: comparing the current operating environment temperature with a preset environment temperature range; judging whether the wind generating set is in a normal working state or not; when the current operating environment temperature is greater than the rated maximum value in the preset environment temperature range and less than the actual demand maximum value in the preset environment temperature range and the wind generating set is in a normal working state, performing capacity reduction control on the wind generating set according to the output power corresponding to the current operating environment temperature in the preset high-temperature capacity reduction control table; and when the current operating environment temperature is less than the rated maximum value with capacity increasing requirements in the preset environment temperature range and the wind generating set is in a normal working state, performing capacity increasing control on the wind generating set according to the output power corresponding to the current operating environment temperature in the preset low-temperature capacity increasing control table.
Therefore, the running state of the gear box of the unit is evaluated in real time according to the monitoring information of the unit, the sub-health state of the fan is known in advance, a reasonable running maintenance scheme is formulated, the output loss of the fan can be reduced, and meanwhile more serious component faults are avoided.
Most of existing methods for evaluating the performance of the unit utilize a feedforward neural network to establish a multi-parameter wind turbine state prediction model so as to evaluate the performance of the unit, but the neural network modeling needs to consume a large amount of training time, and the size of an input sample of the model affects the training precision of the model.
In addition, although the existing SCADA (Supervisory Control And Data Acquisition) system can provide a lot of unit state parameters, most of the existing SCADA systems adopt single-dimensional Data for analysis, or adopt multidimensional Data without comprehensively considering the correlation among the parameters, And cannot further utilize the coupling characteristics among the state parameters.
Moreover, few existing methods for evaluating the unit relate to the high-temperature capacity reduction state of the wind turbine.
Therefore, a fan state evaluation method is expected to be provided to solve the problems in the prior art by aiming at the defects that the traditional state estimation method cannot fully utilize the coupling characteristics among all state parameters and the modeling time is long.
Disclosure of Invention
Based on the technical problems in the background art, the invention provides a wind turbine generator set high-temperature capacity reduction state intelligent evaluation method and system, so that the coupling characteristics among the state parameters of each set can be completely utilized, and the technical problems in the prior art are solved.
In a first aspect, the intelligent evaluation method for the high-temperature capacity-reducing state of the fan, provided by the invention, comprises the following steps:
collecting multi-dimensional characteristic variables of the wind turbine generator, and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
establishing a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
obtaining correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function, constructing a relation tree based on the correlation coefficients, constructing a vine-Copula structure by using the relation tree, taking the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network, and generating a vine-Copula Bayesian network model;
calculating a conditional probability table of each multi-dimensional characteristic variable, and inputting the conditional probability table into a vine-Copula Bayesian network model to obtain the conditional probability table of the vine-Copula Bayesian network model;
detecting the test samples in the state point test sample set by using the conditional probability table of the vine-Copula Bayesian network model to obtain a state detection result and evaluating the state detection result;
the state point test sample set is formed after marking the high-temperature capacity reduction state points of the multi-dimensional characteristic variables.
In a second aspect, the invention provides an intelligent evaluation system for a high-temperature capacity reduction state of a wind turbine generator, which includes: the system comprises an acquisition module, a determination module, a vine-Copula Bayesian network model acquisition module and a probability table acquisition module;
the vine-Copula Bayesian network model generation module comprises a correlation coefficient acquisition unit, a relation tree acquisition unit, a vine-Copula structure construction unit, a generation unit and an evaluation unit;
the acquisition module is used for acquiring multi-dimensional characteristic variables of the wind turbine generator and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
the determining module establishes a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
the correlation coefficient obtaining unit obtains correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function;
the relation tree obtaining unit constructs a relation tree based on the correlation coefficient;
the vine-Copula structure building unit builds a vine-Copula structure by using the relation tree and takes the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network;
the generation unit generates a vine-Copula Bayesian network model according to the DAG network structure of the vine-Copula Bayesian network;
the probability table acquiring unit calculates a conditional probability table of each multi-dimensional characteristic variable, and inputs the conditional probability table into a vine-Copula Bayesian network model to obtain a conditional probability table of the vine-Copula Bayesian network model;
the vine-Copula Bayesian network model detects the test samples in the state point test sample set to obtain a state detection result; the evaluation unit evaluates the state detection result;
the state point test sample set is formed after marking the high-temperature capacity reduction state points of the multi-dimensional characteristic variables.
Compared with the prior art, the invention has the beneficial effects that:
according to the intelligent evaluation method for the high-temperature capacity reduction state of the wind turbine generator, effective correlation analysis is carried out on a vine-Copula Bayesian network model according to the distribution characteristics of each multi-dimensional characteristic variable per se from an SCADA (supervisory control and data acquisition) system with large data volume, and the vine-Copula Bayesian network model is established on the basis of a vine-Copula structure;
the vine-Copula Bayesian network model comprehensively considers state parameters and relativity of all dimensions of characteristic variables of the wind turbine generator, carries out prediction evaluation on the occurrence probability of the high-temperature capacity reduction state of the wind turbine generator, and evaluates the output result of the model through a cross entropy algorithm.
The intelligent evaluation method and system for the high-temperature capacity reduction state of the wind turbine generator set are close to the actual working condition of the wind turbine generator set, can provide reference for field operation and maintenance personnel, better deal with and handle the high-temperature capacity reduction state of the gear box, and reduce and avoid further larger loss.
Drawings
The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings, in which:
fig. 1 is a schematic flow chart of a method for intelligently evaluating a high-temperature capacity reduction state of a wind turbine generator according to an embodiment of the present invention.
Fig. 2 is a wind speed-power scatter diagram provided by an embodiment of the present invention.
FIG. 3 is a pair-Copula function image corresponding to the gearbox oil temperature-gearbox shaft temperature provided by the embodiment of the invention.
Fig. 4 is a tree diagram of multidimensional feature variables of each root node according to an embodiment of the present invention.
Fig. 5 is a diagram of a DAG network structure of a bayesian network according to an embodiment of the present invention.
Fig. 6 is a graph of the output results of the vine-Copula and bayesian network model test provided by the embodiment of the present invention.
Fig. 7 is a schematic structural diagram of an intelligent evaluation system for a high-temperature capacity reduction state of a wind turbine generator according to an embodiment of the present invention.
Detailed Description
In order to make the implementation objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be described in more detail below with reference to the accompanying drawings in the embodiments of the present invention. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are only some, but not all embodiments of the invention. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Because wind turbine generator system operational environment is complicated, as mechanical transmission system easily receive environmental factor to influence, like the wind speed of random variation and the temperature that fluctuation range is big for all kinds of system spare parts can not operate under stable operating mode, this just leads to the fan to be in sub-health state operation in some time quantum, though not necessarily can make the unit shut down, nevertheless can reduce the unit and exert oneself and generated energy, thereby influences fan operation enterprise economic benefits. In order to reduce the power generation loss of the wind turbine generator in the sub-health state, the implementation distinguishes and evaluates the high-temperature capacity reduction state of the wind turbine generator through the established vine-Copula bayes network model, so that a more targeted operation and maintenance scheme is made in time, various faults caused by further deterioration of the state of the wind turbine generator are avoided, and the intelligent evaluation method and the system for the high-temperature capacity reduction state of the wind turbine generator in the embodiment are described in detail below.
Intelligent evaluation method for high-temperature capacity reduction state of wind turbine generator
As an example, the following data are used in the present embodiment to check the effectiveness of the intelligent evaluation method for the high-temperature capacity reduction state of the wind turbine generator according to the present invention:
in the embodiment, data in an SCADA system collected by a certain wind power plant in the eastern part of inner Mongolia between 1 month and 1 month in 2018 and 12 month and 31 month in 2018 are adopted, the time resolution is 10min, and a wind power generation unit is a variable-pitch regulation three-blade horizontal-axis doubly-fed asynchronous generator with the rated power of 750 kW.
Fig. 1 is a schematic flow chart of a method for intelligently evaluating a high-temperature capacity reduction state of a wind turbine generator according to an embodiment of the present invention, and referring to fig. 1,
step 1: collecting multi-dimensional characteristic variables of the wind turbine generator, and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
preferably, the multidimensional characteristic variable of the wind turbine generator collected by the embodiment is derived from an SCADA (Supervisory Control And Data Acquisition) system of the wind farm.
Preferably, the most effectiveCarrying out data preprocessing on the multi-dimensional characteristic variables of the wind turbine generator by using a variance method in the optimal group; further, the edge distribution function U ═ U may be determined for each multidimensional feature variable by a kernel density estimation method1,U2,…,Um]And forming a training data set of the vine-Copula Bayesian network model based on the edge separating function.
Specifically, step 1 includes step 101, step 102, and step 103.
Step 101: preprocessing the collected multidimensional characteristic variables of the wind turbine generator to form preprocessed multidimensional characteristic variables;
referring to fig. 2, fig. 2 is a wind speed-power scatter diagram provided in the embodiment of the present invention, which is implemented by collecting multidimensional characteristic variables of a wind turbine at preset time intervals (for example, preset time intervals are 10min), and screening the collected multidimensional characteristic variables by using an optimal intra-group variance method to obtain preprocessed multidimensional characteristic variables, so that normal power points and abnormal power points can be screened out, and defect points and shutdown points can be deleted at the same time;
the multi-dimensional characteristic variables of the wind turbine generator set comprise wind speed, active power, gearbox oil temperature, gearbox shaft temperature, generator temperature, ambient temperature, cabin temperature, generator rotating speed, impeller rotating speed, wind direction angle, yaw angle, high-temperature capacity reduction filling data in abnormal condition complaints and the like.
Step 102: marking high-temperature capacity reduction state points in the preprocessed multidimensional characteristic variables, taking the marked preprocessed multidimensional characteristic variables as state sample data sets, and dividing the state sample data sets into state point training sample sets and test sample sets;
specifically, in this embodiment, the high-temperature capacity reduction filling data in the abnormal condition complaint in the preprocessed multidimensional characteristic variables may be labeled, the data of the preprocessed multidimensional characteristic variables at the labeled high-temperature capacity reduction state time may be taken out as a state sample dataset, and the state sample dataset may be divided into a state point training sample set and a test sample set.
Step 103: determining the set of state point training samplesThe edge distribution function U ═ U of the multidimensional characteristic variable in (a)1,U2,…,U1]2Taking the edge distribution function as training data of a vine-Copula Bayesian network model to form a training data set of the vine-Copula Bayesian network model; preferably, an edge distribution function U ═ U of the multidimensional feature variables (e.g., the 12 multidimensional feature variables in step 101) in the state point training sample set is determined by a kernel density estimation method1,U2,…,U12]。
Step 2: establishing a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
preferably, the optimal Copula function and the parameters of the optimal Copula function between two preprocessed multidimensional characteristic variables can be determined by an AIC (Akaike information criterion) criterion.
Specifically, step 2 includes step 201, step 202, and step 203.
Step 201: and performing pairwise relevance modeling on the preprocessed multidimensional characteristic variables by a pair-Copula method to obtain a pair-Copula function and parameters of the pairwise multidimensional characteristic variables.
Specifically, according to experience values, relevance modeling is carried out on pairwise multidimensional characteristic variables by using a pair-Copula method according to a training data set of a vine-Copula Bayesian network model, and relevance modeling is respectively carried out on the following 5 types of Copula functions, namely relevance modeling is carried out on the multidimensional characteristic variables by using Gaussian-Copula (Gaussian-Korea), t-Copula (t-Korea function), Gumbel-Copula (Ongbeier-Korea), Clayton-Copula (Cletton-Korea) and Frank-Copula (Frank-Korea) functions, so that pair-two pairwise multidimensional characteristic variables of pair-Copula functions and parameters are obtained.
By way of example, the formulas and parameter value ranges of the above-mentioned 5 types of Copula functions are shown in the following table. It should be noted that the parameter of the optimal Copula function in this embodiment is the value of the parameter ρ or θ in each Copula function in table 1 below.
TABLE 1
Step 202: determining an optimal Copula function between every two multidimensional characteristic variables through an AIC (advanced information center) criterion based on the pair-Copula functions and parameters of every two multidimensional characteristic variables;
wherein, the definition formula of the AIC criterion is as follows:
AIC＝2k-2ln L；(1)
wherein AIC is the red pool information content, k is the number of model parameters, and L is the maximum likelihood estimation. According to the definition, the smaller the value of AIC is, the higher the fitting degree of the constructed model is, and the more concise the value is.
Due to the limited space, the determination result of the AIC criterion of the pair-Copula function established by taking the wind speed, the active power, the oil temperature of the gearbox, the shaft temperature of the gearbox and the ambient temperature as main variables (i.e. the optimal Copula function type between two multidimensional characteristic variables) can be selected as shown in table 2 below:
TABLE 2
In Table 2 above, Frank is an abbreviation for Frank-Copula function, and t is an abbreviation for t-Copula function.
And step 3: obtaining correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function, constructing a relation tree based on the correlation coefficients, constructing a vine-Copula structure by using the relation tree, taking the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network, and generating a vine-Copula Bayesian network model;
specifically, step 3 includes step 301 and step 302. Step 303 and step 304, wherein the correlation coefficient comprises Kendall coefficient tau of each group of pair-Copula functionsi,jAnd Spearman coefficient
Step 301, correlation coefficient obtaining step: establishing a plurality of groups of pair-Copula functions according to the optimal Copula function, and obtaining Kendall coefficients tau of the groups of pair-Copula functionsi,jAnd Spearman coefficient
In this embodiment, according to the optimal Copula function obtained in step 202 (that is, the optimal Copula function used for determining each group of Pair-Copula), multiple groups of Pair-Copula functions may be established, and the Kendall coefficient τ of each group of Pair-Copula is obtainedi,jAnd Spearman coefficient
Referring to fig. 3, fig. 3 is a pair-Copula function image corresponding to two multidimensional characteristic variables of the gearbox oil temperature and the gearbox shaft temperature provided by the embodiment of the invention.
Wherein u and v are edge distributions of the ith dimension and the jth dimension characteristic variables respectively.
Step 302, selecting step: selecting taui,jAnd
the average of the multi-dimensional characteristic variables is not less than a predetermined value (for example, the predetermined value is 0.6), and the pair-Copula function corresponding to the selected multi-dimensional characteristic variable is used asIs the initial unit of the vine-Copula Bayesian model.
Specifically, by judging tau of each group of pair-Copula functionsi,jAnd
to select from the data with stronger correlation (i.e. with stronger correlation)
And τi,jAverage of which is not less than a predetermined value) as an initial unit of the vine-Copula model, in this embodiment, the initial unit includes 8 characteristic variables, such as wind speed, active power, gearbox oil temperature, gearbox shaft temperature, generator temperature, ambient temperature, cabin temperature, and high temperature and reduced capacity state of the fan, and the 8 characteristic variables constitute 12 groups of pair-Copula functions.
The correlation coefficient r of the pair-Copula function can be generally divided into three stages: | r | R<0.5 is low correlation; (ii) gamma ray not more than 0.6 ≤<0.8 is significantly correlated; gamma ray not less than 0.8 ≤<1 is high correlation; in the present embodiment, the expression (τ)k+ρs) And/2 is more than or equal to 0.6, selecting the multidimensional characteristic variable with strong correlation, finally retaining 12 groups of pair-Copula functions in the multidimensional characteristic variable, and establishing a Kendall coefficient matrix of the multidimensional characteristic variable.
Step 303, a relational tree construction step: for Kendall coefficient tau obtained in step 301i,jSumming to obtain Kendall coefficient sum
The Kendall coefficients are summed
U-th corresponding to the maximum value ofiMarking each characteristic variable as a first root node, and constructing Tree according to the pair-Copula function corresponding to the selected multidimensional characteristic variablei(relationship tree), the formula is as follows:
step 304: repeating the step 303 until all multidimensional characteristic variables are calculated to be the Tree of the root nodeiUsing said TreeiAnd constructing a vine-Copula structure, and taking the vine-Copula structure as a DAG (Directed Acyclic Graph) network structure of the vine-Copula Bayesian network to generate a vine-Copula Bayesian network model.
Specifically, fig. 4 is a Tree diagram of multidimensional feature variables of each root node according to the embodiment of the present invention, and referring to fig. 4, step 303 is repeatedly executed until all multidimensional feature variables are calculated as Tree of the root nodeiAnd according to said TreeiA vine-Copula structure is built, as can be seen from fig. 4, the generator temperature and the unit high-temperature capacity reduction state have no direct correlation, so that two groups of Directed connecting edges, i.e., U4 → U5 and U3 → U5, can be removed, the generator temperature is deleted from an initial unit, 7 characteristic variables, i.e., wind speed, active power, gearbox oil temperature, gearbox shaft temperature, ambient temperature, cabin temperature and fan high-temperature capacity reduction state, are left, the 7 characteristic variables form a basic unit of a vine-Copula bayes model, and a DAG (Directed Acyclic Graph) network structure of a bayes network shown in fig. 5 is constructed.
And 4, step 4: calculating a conditional probability table of each multi-dimensional characteristic variable, and inputting the conditional probability table into a vine-Copula Bayesian network model to obtain the conditional probability table of the vine-Copula Bayesian network model;
specifically, step 4 includes step 401 and step 402.
Step 401: obtaining a conditional probability table of each node based on a DAG network structure of the vine-Copula Bayesian network and through a maximum likelihood estimation algorithm;
in this embodiment, after determining the DAG network structure of the vine-Copula bayes network, the conditional probability table of each node may be calculated by using a Maximum Likelihood Estimation (MLE) algorithm, that is, sample data U ═ U in the DAG network structure of the bayes network obtained in step 304 is [ U ═ U1,U2,…,Um]The likelihood function can be written as:
wherein L is a likelihood function of a DAG network structure of the vine-Copula Bayesian network; m is the number of characteristic variables; xi is a father node, and Xj is a child node corresponding to the father node Xi; i and j are respectively the serial numbers corresponding to the father node and the child node; ui is the edge distribution corresponding to the current parent node i, and Pa (Xi) is the edge distribution probability of the parent node Xi.
Step 402: and inputting the Conditional Probability Table of each node into a vine-Copula Bayesian network model to obtain a Conditional Probability Table (CPT) of the Bayesian network model.
And 5: and detecting the test samples in the state point test sample set by using the conditional probability table of the vine-Copula Bayesian network model to obtain a state detection result, and evaluating the state detection result.
In this embodiment, for example, data of one week in 4 months in 2018 is used, and the size of a test sample in a state point test sample set is N1000, the test sample is input into a vine-Copula bayes network model, the test sample model is detected by using a bayesian inference algorithm to obtain a state detection result, and the state detection result is evaluated by using a cross entropy algorithm to form an evaluation result.
Specifically, step 5 includes step 501, step 502, and step 503.
Step 501: obtaining the prior probability of the test samples in the test sample set of the state points; wherein the test sample is derived from the test sample set of state points obtained in step 102;
step 502: inputting the prior probability into a vine-Copula bayes network model, carrying out bayes inference on the conditional probability table of the bayes network model obtained in the step 402, inspecting the test sample to obtain a state detection result, wherein the state detection result is shown in fig. 6 (in fig. 6, the abscissa is a time point, and the ordinate is the probability that the model prediction input example belongs to the real category)
Specifically, taking the variable of wind speed as an example, the process of obtaining the prior probability is described as follows: first, the wind speed in the training data set is divided into n state spaces, where n is 5 in this embodiment, and the wind speed is divided into [0,5 ] within 0-25m/s][5.001-10][10.001,15][15.001 20][20.001 25]Then looking up the number of the state interval that the wind speed falls into 0-5m/s from the training data samples (m in total), and recording as m0Then the prior probability that the wind speed falls within 0-5m/s can be recorded as m0And/ms, similarly, the prior probability that the wind speed falls into other state spaces can be obtained, so that the prior probability of the wind speed in n state intervals in the training data set is obtained, then the wind speed state interval to which the training sample belongs is corresponded according to the numerical value of the wind speed variable in the test sample, and the prior probability corresponding to the interval is the prior probability of the wind speed of the test sample.
Step 503: and evaluating the state detection result by using a cross entropy algorithm, wherein the calculation formula is as follows:
where Y is an output variable, X is an input variable, L is a loss function, and N is an input sample size (in this embodiment, N is 1000,) YiAs an input instance piTrue class of piPredicting for the model that the input instance belongs to the true class yiThe probability of (a) of (b) being,
the results of the number of state points obtained by equation (8) are shown in table 3 below:
TABLE 3
| Number of high temperature capacity-reducing state points (output result in model)>Number of dots 0.75) | logLoss |
| 167/189 | 0.2022 |
In the prior art, a cross entropy loss function of a probability prediction evaluation model for the number of high-temperature melting-down state points of the wind turbine generator is about 0.5, and a cross entropy loss function obtained by using a vine-Copula Bayesian network model in the embodiment is 0.2022, which shows that the vine-Copula Bayesian network model built in the embodiment has good evaluation precision, is more in line with actual conditions, and can accurately evaluate the high-temperature capacity-down state of the wind turbine generator, so that references are provided for operation and maintenance personnel on a wind power site, the high-temperature capacity-down state of the gear box can be better handled and treated, and further greater loss is reduced and avoided.
The vine-Copula in the embodiment is a correlation analysis model, so that the problem of feature loss or feature redundancy in the traditional correlation analysis can be solved, and more comprehensive feature variables in the wind turbine can be extracted;
the Bayesian network is an uncertain causal association model and can effectively merge and intensively express information of various channels by utilizing historical data probability information;
in the embodiment, the vine-Copula structure and the bayesian network structure are combined, the formed vine-Copula bayesian network model can perform relevance analysis on each state parameter of the wind turbine generator and the high-temperature capacity reduction state of the gear box, and extract the relevance characteristics among all multi-dimensional characteristic variables, so that the accurate relevance relation among all the multi-dimensional characteristic variables can be obtained, and the accuracy of the vine-Copula bayesian network model is improved.
Second, wind turbine generator system high temperature capacity reduction state intelligent evaluation system
Fig. 7 is a schematic structural diagram of an intelligent evaluation system for a high-temperature capacity reduction state of a wind turbine generator according to an embodiment of the present invention, and as shown in fig. 7, the intelligent evaluation system for a high-temperature capacity reduction state of a wind turbine generator according to the embodiment of the present invention includes: the system comprises an acquisition module, a determination module, a vine-Copula Bayesian network model acquisition module and a probability table acquisition module;
the vine-Copula Bayesian network model generation module comprises a correlation coefficient acquisition unit, a relation tree acquisition unit, a vine-Copula structure construction unit, a generation unit and an evaluation unit;
the acquisition module is used for acquiring multi-dimensional characteristic variables of the wind turbine generator and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
the determining module establishes a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
the correlation coefficient obtaining unit obtains correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function;
the relation tree obtaining unit constructs a relation tree based on the correlation coefficient;
the vine-Copula structure building unit builds a vine-Copula structure by using the relation tree and takes the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network;
the generation unit generates a vine-Copula Bayesian network model according to the DAG network structure of the vine-Copula Bayesian network;
the probability table acquiring unit calculates a conditional probability table of each multi-dimensional characteristic variable, and inputs the conditional probability table into a vine-Copula Bayesian network model to obtain a conditional probability table of the vine-Copula Bayesian network model;
the vine-Copula Bayesian network model detects the test samples in the state point test sample set to obtain a state detection result; the evaluation unit evaluates the state detection result;
the state point test sample set is formed after marking the high-temperature capacity reduction state points of the multi-dimensional characteristic variables.
The specific processing procedure of the evaluation system in this embodiment may refer to the specific processing procedure of the intelligent evaluation method for the high-temperature capacity-reducing state of the wind turbine generator provided in the above embodiment, which is not described herein again.
The invention has the beneficial effects that:
according to the intelligent evaluation method for the high-temperature capacity reduction state of the wind turbine generator, effective correlation analysis is carried out on a vine-Copula Bayesian network model according to the distribution characteristics of each multi-dimensional characteristic variable per se from an SCADA (supervisory control and data acquisition) system with large data volume, and the vine-Copula Bayesian network model is established on the basis of a vine-Copula structure;
the vine-Copula Bayesian network model comprehensively considers the state parameters and the relativity of all dimensions of the characteristic variables of the wind turbine generator, carries out prediction evaluation on the occurrence probability of the high-temperature capacity reduction state of the wind turbine generator, and has good prediction evaluation effect; and evaluating the output result of the model by a cross entropy algorithm.
The intelligent evaluation method and system for the high-temperature capacity reduction state of the wind turbine generator set are close to the actual working condition of the wind turbine generator set, can provide reference for field operation and maintenance personnel, better deal with and handle the high-temperature capacity reduction state of the gear box, and reduce and avoid further larger loss.
Finally, it should be pointed out that: the above examples are only for illustrating the technical solutions of the present invention, and are not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. The intelligent evaluation method for the high-temperature capacity reduction state of the wind turbine generator is characterized by comprising the following steps of:
collecting multi-dimensional characteristic variables of the wind turbine generator, and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
establishing a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
obtaining correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function, constructing a relation tree based on the correlation coefficients, constructing a vine-Copula structure by using the relation tree, taking the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network, and generating a vine-Copula Bayesian network model;
calculating a conditional probability table of each multi-dimensional characteristic variable, and inputting the conditional probability table into a vine-Copula Bayesian network model to obtain the conditional probability table of the vine-Copula Bayesian network model;
detecting the test samples in the state point test sample set by using the conditional probability table of the vine-Copula Bayesian network model to obtain a state detection result; evaluating the state detection result;
the state point test sample set is formed after marking the high-temperature capacity reduction state points of the multi-dimensional characteristic variables.
2. The method according to claim 1, wherein the forming of the training dataset of the vine-Copula bayes net model based on the multi-dimensional feature variables comprises the sub-steps of:
preprocessing the collected multidimensional characteristic variables of the wind turbine generator to form preprocessed multidimensional characteristic variables;
marking high-temperature capacity reduction state points in the preprocessed multidimensional characteristic variables, taking the marked preprocessed multidimensional characteristic variables as state point sample datasets, and dividing the state point sample datasets into state point training sample sets and state point testing sample sets;
and determining an edge distribution function of the multidimensional characteristic variables in the state point training sample set, and taking the edge distribution function as training data of a vine-Copula Bayesian network model to form a training data set of the vine-Copula Bayesian network model.
3. The method according to claim 1, wherein the determining of the optimal Copula function and the parameters of the optimal Copula function between two preprocessed multidimensional characteristic variables comprises the following sub-steps:
performing pairwise relevance modeling on the preprocessed multidimensional characteristic variables by a pair-Copula method to obtain a pair-Copula function and parameters of the pairwise multidimensional characteristic variables;
and determining the optimal Copula function type and the optimal Copula function parameter between every two multidimensional characteristic variables through an AIC (automatic information center) criterion based on the pair-Copula functions and parameters of every two multidimensional characteristic variables.
4. The method of claim 1, wherein the generating of the vine-Copula bayes network model comprises the sub-steps of:
a correlation coefficient obtaining step: establishing a plurality of groups of pair-Copula functions according to the optimal Copula function, and obtaining Kendall coefficients tau of the groups of pair-Copula functionsi,jAnd Spearman coefficient
Selecting: selecting taui,jAnd
taking a pair-Copula function corresponding to the selected multidimensional characteristic variable as an initial unit of a vine-Copula Bayes model;
a relation tree construction step: for Kendall coefficient taui,jSumming to obtain Kendall coefficient sum
The Kendall coefficients are summed
The characteristic variable corresponding to the maximum value in the multi-dimensional characteristic variables is recorded as a first root node, and the multi-dimensional characteristic variables are selected according to the selected multi-dimensional characteristic variablesConstructing a relation tree by the corresponding pair-Copula function;
and repeating the relational tree steps until all the multidimensional characteristic variables are calculated to serve as the relational tree of the root node, building a vine-Copula structure by utilizing the relational tree, and generating a vine-Copula Bayesian network model by taking the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network.
5. The method of claim 1, wherein the obtaining the conditional probability table of the vine-Copula bayes network model comprises the sub-steps of:
obtaining a conditional probability table of each node based on a DAG network structure of the vine-Copula Bayesian network and through a maximum likelihood estimation algorithm;
and inputting the conditional probability table of each node into a vine-Copula Bayesian network model to obtain the conditional probability table of the Bayesian network model.
6. The method according to claim 1, wherein said evaluating said state detection result comprises the sub-steps of:
obtaining the prior probability of the test samples in the test sample set of the state points;
inputting the prior probability into a vine-Copula Bayesian network model, carrying out Bayesian inference through a conditional probability table of the Bayesian network model, and inspecting the test sample to obtain a state detection result;
and evaluating the state detection result by utilizing a cross entropy algorithm.
7. The method of claim 2, wherein the pre-processed multi-dimensional feature variables are obtained by screening the collected multi-dimensional feature variables by an optimal intra-group variance method.
8. The method of claim 2, wherein the edge distribution function of the multidimensional feature variables in the state point training sample set is determined using a kernel density estimation method.
9. Method according to any of claims 1 to 8, characterized in that the multidimensional feature variable of the wind turbines is derived from a SCADA system of a wind farm.
10. The utility model provides a wind turbine generator system high temperature capacity reduction state intelligence evaluation system which characterized in that, the system includes: the system comprises an acquisition module, a determination module, a vine-Copula Bayesian network model acquisition module and a probability table acquisition module;
the vine-Copula Bayesian network model acquisition module comprises a correlation coefficient acquisition unit, a relation tree acquisition unit, a vine-Copula structure construction unit, a generation unit and an evaluation unit;
the acquisition module is used for acquiring multi-dimensional characteristic variables of the wind turbine generator and forming a training data set of a vine-Copula Bayesian network model based on the multi-dimensional characteristic variables;
the determining module establishes a pair-Copula function according to the training data set of the vine-Copula Bayesian network model to determine an optimal Copula function between every two multidimensional characteristic variables;
the correlation coefficient obtaining unit obtains correlation coefficients of each group of pair-Copula functions in the optimal Copula function according to the optimal Copula function;
the relation tree obtaining unit constructs a relation tree based on the correlation coefficient;
the vine-Copula structure building unit builds a vine-Copula structure by using the relation tree and takes the vine-Copula structure as a DAG network structure of a vine-Copula Bayesian network;
the generation unit generates a vine-Copula Bayesian network model according to the DAG network structure of the vine-Copula Bayesian network;
the probability table obtaining module calculates a conditional probability table of each multi-dimensional characteristic variable, and inputs the conditional probability table into a vine-Copula Bayesian network model to obtain a conditional probability table of the vine-Copula Bayesian network model;
the vine-Copula Bayesian network model detects the test samples in the state point test sample set to obtain a state detection result; the evaluation unit evaluates the state detection result;
the state point test sample set is formed after marking the high-temperature capacity reduction state points of the multi-dimensional characteristic variables.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910604545.6A CN110334951B (en) | 2019-07-05 | 2019-07-05 | Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910604545.6A CN110334951B (en) | 2019-07-05 | 2019-07-05 | Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110334951A CN110334951A (en) | 2019-10-15 |
| CN110334951B true CN110334951B (en) | 2022-02-08 |
Family
ID=68144317
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910604545.6A Expired - Fee Related CN110334951B (en) | 2019-07-05 | 2019-07-05 | Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110334951B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114389312B (en) * | 2021-11-02 | 2024-05-14 | 国网江苏省电力有限公司苏州供电分公司 | Distributed state estimation method for power distribution network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106651163A (en) * | 2016-12-12 | 2017-05-10 | 南京理工大学 | Capacity confidence level evaluation method for multiple wind power plants on the basis of Copula function |
| CN106650204A (en) * | 2016-09-27 | 2017-05-10 | 北京航空航天大学 | Product failure behavior coupling modeling and reliability evaluation method |
| CN107038292A (en) * | 2017-04-01 | 2017-08-11 | 三峡大学 | A kind of many output of wind electric field correlation modeling methods based on adaptive multivariable nonparametric probability |
| CN108345961A (en) * | 2018-01-30 | 2018-07-31 | 上海电力学院 | The prediction of wind farm group output and analysis method |
| CN109685371A (en) * | 2018-12-25 | 2019-04-26 | 华能陕西定边电力有限公司 | Dynamic based on Bayesian network generally weighs running of wind generating set state comprehensive estimation method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10468136B2 (en) * | 2016-08-29 | 2019-11-05 | Conduent Business Services, Llc | Method and system for data processing to predict health condition of a human subject |
-
2019
- 2019-07-05 CN CN201910604545.6A patent/CN110334951B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106650204A (en) * | 2016-09-27 | 2017-05-10 | 北京航空航天大学 | Product failure behavior coupling modeling and reliability evaluation method |
| CN106651163A (en) * | 2016-12-12 | 2017-05-10 | 南京理工大学 | Capacity confidence level evaluation method for multiple wind power plants on the basis of Copula function |
| CN107038292A (en) * | 2017-04-01 | 2017-08-11 | 三峡大学 | A kind of many output of wind electric field correlation modeling methods based on adaptive multivariable nonparametric probability |
| CN108345961A (en) * | 2018-01-30 | 2018-07-31 | 上海电力学院 | The prediction of wind farm group output and analysis method |
| CN109685371A (en) * | 2018-12-25 | 2019-04-26 | 华能陕西定边电力有限公司 | Dynamic based on Bayesian network generally weighs running of wind generating set state comprehensive estimation method |
Non-Patent Citations (1)
| Title |
|---|
| 基于交叉熵理论的配电变压器寿命组合预测方法;栗然等;《电力系统保护与控制》;20140116;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110334951A (en) | 2019-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Reliability assessment of wind turbine bearing based on the degradation-Hidden-Markov model | |
| Li et al. | Wind turbine fault diagnosis based on Gaussian process classifiers applied to operational data | |
| Dinmohammadi et al. | A fuzzy-FMEA risk assessment approach for offshore wind turbines | |
| CN111537219B (en) | Fan gearbox performance detection and health assessment method based on temperature parameters | |
| CN110259646B (en) | Wind generating set component state early warning method based on historical data | |
| CN111426950B (en) | Wind driven generator fault diagnosis method of multi-scale space-time convolution depth belief network | |
| CN112115999B (en) | Wind turbine generator fault diagnosis method of space-time multi-scale neural network | |
| CN110362045B (en) | Marine doubly-fed wind turbine generator fault discrimination method considering marine meteorological factors | |
| Butler et al. | A feasibility study into prognostics for the main bearing of a wind turbine | |
| CN107728059B (en) | Pitch system state evaluation method | |
| CN113657662B (en) | Downscaling wind power prediction method based on data fusion | |
| Peng et al. | Wind turbine failure prediction and health assessment based on adaptive maximum mean discrepancy | |
| CN113591359A (en) | Cut-in/cut-out wind speed adjusting and optimizing method, system and equipment medium of wind turbine generator | |
| CN109800931A (en) | Wind power plant generated energy loss measurement method and system based on blower SCADA data | |
| Shi et al. | Study of wind turbine fault diagnosis and early warning based on SCADA data | |
| Yang et al. | Fault early warning of wind turbine gearbox based on multi‐input support vector regression and improved ant lion optimization | |
| Du et al. | A SCADA data based anomaly detection method for wind turbines | |
| CN110991701A (en) | Wind power plant fan wind speed prediction method and system based on data fusion | |
| Zhu et al. | Operational state assessment of wind turbine gearbox based on long short-term memory networks and fuzzy synthesis | |
| CN116609055A (en) | Method for diagnosing wind power gear box fault by using graph convolution neural network | |
| CN116771610A (en) | Method for adjusting fault evaluation value of variable pitch system of wind turbine | |
| CN112417612A (en) | Method for tracking degradation state and evaluating failure aggregation risk of wind power gear box | |
| CN110334951B (en) | Intelligent evaluation method and system for high-temperature capacity reduction state of wind turbine generator | |
| CN114169718A (en) | Method for improving reliability of wind turbine generator based on state evaluation of wind turbine generator | |
| Yang et al. | Assessment of equipment operation state with improved random forest |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220208 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |