CN111444615A - Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve - Google Patents
Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve Download PDFInfo
- Publication number
- CN111444615A CN111444615A CN202010228062.3A CN202010228062A CN111444615A CN 111444615 A CN111444615 A CN 111444615A CN 202010228062 A CN202010228062 A CN 202010228062A CN 111444615 A CN111444615 A CN 111444615A
- Authority
- CN
- China
- Prior art keywords
- characteristic
- fault
- curve
- photovoltaic array
- different
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000003745 diagnosis Methods 0.000 title claims abstract description 25
- 239000013598 vector Substances 0.000 claims abstract description 37
- 238000004088 simulation Methods 0.000 claims abstract description 29
- 230000009467 reduction Effects 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000005259 measurement Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000007635 classification algorithm Methods 0.000 claims abstract description 7
- 238000009826 distribution Methods 0.000 claims description 20
- 230000007613 environmental effect Effects 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000009499 grossing Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 claims description 3
- 230000001174 ascending effect Effects 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 7
- 238000012423 maintenance Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000032683 aging Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000007847 structural defect Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Mathematical Optimization (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Algebra (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Testing Of Individual Semiconductor Devices (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention discloses a photovoltaic array fault diagnosis method based on K nearest neighbor and IV curves, which comprises the steps of extracting characteristic parameters in simulated IV characteristic curves of different fault types, different irradiances and different temperatures and carrying out standardization treatment to obtain characteristic vectors of the simulated IV characteristic curves of different fault types, different irradiances and different temperatures; carrying out dimensionality reduction on the feature vector of the simulation IV characteristic curve; and taking the feature vectors subjected to the dimension reduction processing of the simulation IV characteristic curves under all fault types as a training data set, taking the feature vectors subjected to the dimension reduction processing of the actual measurement IV characteristic curves of the current photovoltaic array as an actual measurement data set, and performing fault diagnosis on the current photovoltaic array by adopting a K neighbor classification algorithm. The method can accurately judge the faults of the photovoltaic array, can realize fault state detection and timely maintenance, reduces the fault risk and ensures the stable operation of the photovoltaic power station.
Description
Technical Field
The invention belongs to the technical field of photovoltaic array fault diagnosis, and particularly relates to a photovoltaic array fault diagnosis method based on K nearest neighbor and IV curves.
Background
The photovoltaic array is an important component of a photovoltaic system, and is susceptible to pollution and turbidity such as sunlight, snow, dust and the like due to long-term work outdoors with a severe environment, and may cause various concurrent faults of the direct current side of the photovoltaic array, such as a bypass diode short-circuit fault, a bypass diode open-circuit fault, a cable aging fault, local shadow shielding, and a bypass diode short-circuit and open-circuit accompanied by shadow shielding. This will reduce the power generation efficiency of the photovoltaic power plant, reduce its service life, even cause a fire to cause safety problems and increase the maintenance cost of the photovoltaic system, and therefore the state detection and fault diagnosis technology of the photovoltaic array is of particular interest to many scholars.
The related fault diagnosis methods proposed by the present scholars can be roughly divided into the following methods: the first detection based on the infrared thermal imager is usually used for detecting hot spots or some structural defects in the photovoltaic module by shooting the photovoltaic module through the infrared thermal imager because the photovoltaic module has obvious temperature difference characteristics in normal and fault states, but the infrared thermal imager needs to be additionally added and is expensive, so that the detection is not suitable for a small photovoltaic power station.
The second is based on measuring the voltage, current, power at the operating point of the photovoltaic array. The fault discrimination is mainly performed by calculating the deviation between the analog output and the measured output. For example, a student builds a Simulink simulation model and calculates the deviation between the power and the measured power to diagnose the fault. The method has low hardware technical requirements, but cannot judge the fault type more accurately.
A third current scholars also propose a fault detection method based on current-voltage (IV) curve measurement, and the commonly used methods include artificial neural networks, support vector machines, decision trees and the like, which have limited sample characteristic parameters, but the selection of the sample characteristic parameters directly affects the precision of the whole diagnostic model and cannot traitor multiple concurrent faults.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve, which can accurately judge various fault types such as array shadow occlusion, bypass diode short circuit, cable aging, bypass diode short circuit accompanying shadow occlusion, bypass diode open circuit and the like, can realize fault state detection and timely maintenance, reduce fault risk and ensure stable operation of a photovoltaic power station.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a photovoltaic array fault diagnosis method based on K nearest neighbor and an IV curve comprises the following steps:
acquiring simulation IV characteristic curves of different fault types, different irradiance and different temperatures;
extracting characteristic parameters in simulation IV characteristic curves of different fault types, different irradiances and different temperatures;
standardizing the extracted characteristic parameters to obtain standardized simulated IV characteristic curve characteristic vectors of different fault types, different irradiances and different temperatures;
carrying out dimensionality reduction on the feature vector of the simulation IV characteristic curve;
and taking the feature vectors of the simulation IV characteristic curve under all fault types after the dimension reduction processing as a training data set, taking the feature parameters of the current photovoltaic array actual measurement IV characteristic curve and the feature vectors after the dimension reduction processing as an actual measurement data set, and performing fault diagnosis on the current photovoltaic array by adopting a K neighbor classification algorithm based on the training data set and the actual measurement data set.
Further, the obtaining of the simulated IV characteristic curves of different fault types and different irradiance and temperatures includes:
acquiring environmental parameters of a previous month measured in a photovoltaic array monitoring system, wherein the environmental parameters comprise photovoltaic assembly coplanar irradiance G and assembly backboard temperature T;
uniformly sampling the actually measured environmental parameters of the previous month, wherein the number of the sampled samples is m;
the photovoltaic system fault modeling simulation method or the Simulink simulation model is utilized to simulate m IV characteristic curves under different irradiance and temperature corresponding to each fault type under n fault types, and m x n simulated IV characteristic curves under different fault types are obtained.
Further, the characteristic parameters in the simulated IV characteristic curve under different fault types, different irradiance and different temperatures are extracted, and the characteristic parameters comprise S1~S13:
S1=G,
S2=T,
S4=V|I=0=Voc,
S5=I|V=0=Isc,
S6=max(V×I)=Pm,
S12=Pks_num,
S13=Vall_num,
Wherein G represents the coplanar irradiance of the photovoltaic assembly, T represents the backboard temperature of the photovoltaic assembly, and IjRepresents the current value of the jth point on the simulated IV characteristic curve, f (I)j) The current value on the simulation IV characteristic curve is represented as IjVoltage value corresponding to time, V (I) represents voltage value corresponding to current I, VocRepresents the open circuit voltage, IscRepresenting short-circuit current, PmRepresents the maximum power point, I represents the current value on the simulated IV characteristic curve, V represents the voltage value on the simulated IV characteristic curve, VmRepresents the voltage at the maximum power point, ImShowing the current at the maximum power point, FF the fill factor, Isc_d1Represents a short-circuit current IscAt a first derivative value, Voc_d1Represents the open circuit voltage VocAnd a first derivative value is obtained, Pks _ num represents the number of peak points in the second derivative curve of the simulation IV characteristic curve after smoothing, and Vall _ num represents the number of valley points in the second derivative curve of the simulation IV characteristic curve after smoothing.
Further, the normalizing the extracted feature parameters includes:
wherein, s'ijRepresents the ith simulated IV characteristic curve, the jth normalized value of the characteristic parameter, sijRepresents the values of the ith simulated IV characteristic, the jth characteristic parameter,denotes the mean value, σ, of the jth characteristic parameterjAnd d represents the number of characteristic parameters in the extracted simulated IV characteristic curve, and m represents the number of simulated IV characteristic curves corresponding to different irradiance and temperature for each fault type.
Further, the dimension reduction processing is carried out on the characteristic vector of the simulation IV characteristic curve by adopting a t-distribution neighborhood embedding algorithm, and the method comprises the following steps:
D1) and (3) reducing the dimension by taking the feature vector normalized by the feature parameter as raw data, wherein the raw data is expressed as:
D2) calculating the joint probability distribution of the low-dimensional mapping points:
wherein, s'iRepresents the ith row, s ' in the s ' matrix 'jRepresents the j-th row, σ, in the s' matrixiIs s'iGaussian variance of (1), Pj|iIs s'jTo s'iIs s'iAnd s'jSimilarity between high-dimensional data points;
D3) defining a joint probability distribution of high-dimensional data points:
D4) calculating low-dimensional similarity qij:
D5) Calculating K L divergence:
D6) calculate gradient minimization K L divergence:
D7) updating the initial solution to obtain the solution Y of the t iteration(t):
Where α (t) represents momentum, and ρ represents a learning rate;
D8) and (5) circularly iterating the steps D4) -D7) until the iteration W times are finished, and finally obtaining low-dimensional data:
Y(W)={y1,y2,y3…ym},
wherein, yiIs s'iThe vector after dimensionality reduction, i, is 1,2, … m.
Further, the fault diagnosis of the current photovoltaic array by using a K-nearest neighbor classification algorithm based on the training data set and the measured data set includes:
E1) calculating the distance between the measured data set and the training data concentration point:
wherein d is12Representing the distance, x, between the feature vectors in the measured data set and the feature vectors in the training data set1iThe i-th feature component, x, representing a feature vector in the measured data set2iRepresenting the ith characteristic component of the characteristic vector in the training data set, and dim representing the dimensionality of the characteristic vector after dimensionality reduction processing;
E2) sorting m × n distance values according to ascending order, wherein n represents the number of fault types;
E3) selecting the first K points with the minimum value in the sorting;
E4) counting the frequency of the first K points in each fault category in a circulating mode;
E5) and returning the fault category with the highest occurrence frequency of the previous K points, and taking the fault category as the fault type of the current photovoltaic array.
Further, K is an odd number.
The invention achieves the following beneficial effects:
the method can accurately judge the concurrent faults of shadow shielding of the photovoltaic array, short circuit of the bypass diode, cable aging, short circuit of the bypass diode accompanied by shadow shielding, open circuit of the bypass diode and the like, can realize fault state detection and timely maintenance, reduces fault risks and ensures stable operation of the photovoltaic power station.
Drawings
FIG. 1 is a flow chart of a fault diagnosis method of the present invention;
FIG. 2 is a schematic diagram of uniform sampling of environmental parameters according to the present invention;
FIG. 3 is a schematic diagram of the extraction of IV curve characteristic parameters according to the present invention;
FIG. 4 is a schematic diagram of data point distribution after dimensionality reduction by the t-distribution neighborhood embedding algorithm of the present invention.
Detailed Description
The invention is further described below. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
Referring to fig. 1, the invention provides a photovoltaic array fault diagnosis method based on K-nearest neighbor and IV curve, which specifically includes the following steps:
obtaining data of IV characteristic curves of different fault types, different irradiances and different temperatures through a photovoltaic system fault modeling simulation method or a Simulink simulation model, and specifically comprising the following steps:
A1) and acquiring the actually measured environmental parameters of the photovoltaic array monitoring system in the previous month, wherein the environmental parameters comprise the coplanar irradiance G of the photovoltaic assembly and the temperature T of the backboard of the assembly.
A2) The measured environmental parameters (G, T) of the previous month are uniformly sampled, referring to fig. 2, in order to improve the accuracy of the diagnostic model and make the simulated IV characteristic curve closer to the actual IV characteristic curve, so the measured environmental parameters (G, T) of the previous month are uniformly sampled, and the sampling result can better represent the distribution of the environmental parameters of the next month. The number of sampling samples in the invention is m.
A3) The photovoltaic system fault modeling simulation method or the Simulink simulation model is utilized to simulate m IV characteristic curves under different irradiance and temperature corresponding to each fault type under n fault types, and m x n IV characteristic curve data under different fault types are obtained.
B, extracting characteristic parameters in the IV characteristic curve under different fault types, different irradiance and different temperatures and actually measuring the characteristic parameters in the obtained IV characteristic curve, wherein the method specifically comprises the following steps:
B1) environmental parameters: coplanar irradiance: s1G, module back plate temperature S2=T;
B2) Area of IV characteristic:due to the IV characteristic voltage, the sequence of currents is discrete values with non-uniform spacing, and according to the trapezoidal rule, the approximate IV characteristic area integral is achieved using the i-th region values:
wherein, IjCurrent value f (I) at j-th point on IV characteristic curvej) Current value of I on the characteristic curve of IVjA corresponding voltage value;
B3) open circuit voltage: s4=V|I=0=Voc;
B4) Short-circuit current: s5=I|V=0=Isc;
B5) Maximum power point power: s6=max(V×I)=Pm;
B9) performing smoothing processing such as interpolation, filtering and the like on the m-n IV characteristic curves under different fault types;
and after the smoothing treatment, solving a first derivative of the IV characteristic curve:
taking short-circuit current IscTaking the value of the first derivative as a characteristic parameter:
taking open circuit voltage VocTaking the value of the first derivative as a characteristic parameter:
and (3) carrying out second-order derivation on the smoothed IV characteristic curve:
B10) searching the number of peak points in the second derivative curve, setting a threshold, considering the point as a peak point when the peak value is greater than the preset threshold, obtaining a valley point in the same way, and counting the number S of the peak points12Pks _ num, the number of valley points S13=Vall_num。
Fig. 3 is a first derivative curve and a second derivative curve obtained by interpolating and filtering the IV characteristic curve and deriving the IV characteristic curve, and more information that can reflect the state of the photovoltaic array is obtained from the IV characteristic curve and the first derivative and the second derivative curve thereof.
Step C, extracting the obtained characteristic parameters (S)1—S13) The normalization process was performed as follows:
C1) under the same fault type, characteristic parameters of the IV characteristic curves under different irradiance and temperature form a characteristic vector matrix:
S∈Rm×dwherein d represents the number of characteristic parameters in the extracted IV characteristic curve, i.e. d is 13, and m represents the number of the IV characteristic curves; sijThe ith IV characteristic curve is shown, the j-th characteristic parameter value, i is 1,2, …, m, j is 1,2, …, d.
C2) And (3) carrying out standardization treatment on the extracted characteristic parameters:
wherein, s'ijRepresents the ith IV characteristic curve, the normalized value of the jth characteristic parameter,represents the average value of the jth characteristic parameter; sigmajThe standard deviation of the jth characteristic parameter is shown.
D, adopting a t-distribution neighborhood embedding algorithm (t-SNE) to realize the dimensionality reduction of the feature vector and the extraction of sensitive features;
the t-SNE algorithm is used for expressing the similarity between data points in an original high-dimensional Euclidean space by using a conditional outline and effectively relieving the problem of crowding between mapping data points in a low-dimensional space;
the t-SNE algorithm measures the similarity between two distributions by using K-L divergence, so that the distance between points with lower similarity in a low-dimensional space of data is larger, the distance between points with higher similarity is smaller, and good conditions are provided for fault diagnosis by using a K Nearest Neighbor (KNN) classification algorithm subsequently.
The method adopts a t-distribution neighborhood embedding algorithm (t-SNE) to realize the dimensionality reduction of the feature vector and the extraction of sensitive features, and specifically comprises the following steps:
D1) and (3) reducing the dimension by taking the feature vector matrix standardized by the feature parameters as raw data, wherein the raw data is expressed as:
D2) the number of iterations W, the learning rate ρ, the momentum α (t), the complexity factor Perp, and the dimension dim are set.
D3) Calculating the joint probability distribution of the low-dimensional mapping points:
wherein, s'iRepresents the ith row, s ' in the s ' matrix 'jRepresents the j-th row, σ, in the s' matrixiIs data point s'iGaussian variance of (1), Pj|iIs s'iAnd s'jSimilarity between high-dimensional data points, Pj|iThe larger the value is, the more s'iAnd s'jThe higher the similarity between data points, Pj|iThe smaller the value is, the smaller is s'iAnd s'jThe lower the similarity between data points.
D4) Because of the conditional probability Pj|iIs symmetric, and P is defined for optimizing K L divergence by replacing conditional probabilities with joint probability distributionsijJoint probability distribution for high-dimensional data points:
Pj|iis s'jTo s'iProbability distribution of (P)i|jIs s'iTo s'jA probability distribution of (a);
The elements in the initial solution represent the number of samples, i.e. the dimension dim.
D5) Calculating low-dimensional similarity qij:
D6) Calculating K L divergence:
D7) calculate gradient minimization K L divergence:
D8) updating the initial solution to obtain the solution Y of the t iteration(t):
D9) Loop iteration steps D5) -D8) until the end of iteration W.
D10) Finally, low-dimensional data are obtained:
Y(W)={y1,y2,y3…ym},
wherein, yiIs s'iReduced vector, i ═ 1,2, … m, yiDimension of (d) is dim.
And further, acquiring an actually measured IV characteristic curve of the current photovoltaic array, extracting characteristic parameters and performing dimension reduction treatment.
And E, adopting a K nearest neighbor classification algorithm (KNN) to realize the fault diagnosis of the photovoltaic array, which comprises the following steps:
E1) extracting characteristic parameters from simulation IV characteristic curves under all fault types and using the characteristic vectors subjected to dimensionality reduction as a training data set, extracting the characteristic parameters from actual measurement IV characteristic curves and using the characteristic vectors subjected to dimensionality reduction as an actual measurement data set, calculating the distance between the actual measurement data set and a training data concentration point, wherein the distance calculation formula adopts an Euclidean distance:
wherein x is1iThe i-th feature component, x, representing a feature vector in the measured data set2iAn ith feature component representing a feature vector in the training dataset;
E2) sequencing m × n distance values according to an increasing order, wherein m represents that m eigenvectors exist under each fault type, and n represents that n fault types exist;
E3) determining the selection value of K, wherein K selects odd numbers as much as possible; selecting the first K points with the minimum value in the sorting;
E4) counting the frequency of the first K points in each fault category in a circulating mode;
E5) and returning the fault category with the highest occurrence frequency of the previous K points, and taking the fault category as the fault type of the current photovoltaic array.
Fig. 4 is a data point distribution diagram after the dimension reduction by the t-distribution neighborhood embedding algorithm, and it can be seen that the distance of the point with lower similarity of the data in the low-dimensional space is larger, and the distance of the point with higher similarity is smaller, and from the diagram, three faults of fault 1, fault 2 and fault 3 can be seen, the distribution among the same fault types is denser, and the distance among the different fault types is larger.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed system and method may be implemented in other ways. For example, the above-described system embodiments are merely illustrative, and for example, the division of the units is only one logical functional division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed.
Those of skill would further appreciate that the various illustrative modules and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that such implementation decisions may be made by those skilled in the art using various means for implementing the functions described herein without departing from the scope of the invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. The industry has described the principles of the invention, and variations and modifications are possible without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (7)
1. A photovoltaic array fault diagnosis method based on K nearest neighbor and IV curves is characterized by comprising the following steps:
acquiring simulation IV characteristic curves of different fault types, different irradiance and different temperatures;
extracting characteristic parameters in simulation IV characteristic curves of different fault types, different irradiances and different temperatures;
standardizing the extracted characteristic parameters to obtain standardized simulated IV characteristic curve characteristic vectors of different fault types, different irradiances and different temperatures;
carrying out dimensionality reduction on the feature vector of the simulation IV characteristic curve;
and taking the feature vectors of the simulation IV characteristic curve under all fault types after the dimension reduction processing as a training data set, taking the feature parameters of the current photovoltaic array actual measurement IV characteristic curve and the feature vectors after the dimension reduction processing as an actual measurement data set, and performing fault diagnosis on the current photovoltaic array by adopting a K neighbor classification algorithm based on the training data set and the actual measurement data set.
2. The method for photovoltaic array fault diagnosis based on K nearest neighbor and IV curve of claim 1, wherein the obtaining of simulated IV characteristic curves under different fault types, irradiance and temperatures comprises:
acquiring environmental parameters of a previous month measured in a photovoltaic array monitoring system, wherein the environmental parameters comprise photovoltaic assembly coplanar irradiance G and assembly backboard temperature T;
uniformly sampling the actually measured environmental parameters of the previous month, wherein the number of the sampled samples is m;
the photovoltaic system fault modeling simulation method or the Simulink simulation model is utilized to simulate m IV characteristic curves under different irradiance and temperature corresponding to each fault type under n fault types, and m x n simulated IV characteristic curves under different fault types are obtained.
3. The method for photovoltaic array fault diagnosis based on K nearest neighbor and IV curve of claim 1, wherein the extracting of characteristic parameters in simulation IV characteristic curves of different fault types and different irradiance and temperature comprises S1~S13:
S1=G,
S2=T,
S4=V|I=0=Voc,
S5=I|V=0=Isc,
S6=max(V×I)=Pm,
S12=Pks_num,
S13=Vall_num,
Wherein G represents the coplanar irradiance of the photovoltaic assembly, T represents the backboard temperature of the photovoltaic assembly, and IjRepresents the current value of the jth point on the simulated IV characteristic curve, f (I)j) The current value on the simulation IV characteristic curve is represented as IjVoltage value corresponding to time, V (I) represents voltage value corresponding to current I, VocRepresents the open circuit voltage, IscRepresenting short-circuit current, PmRepresents the maximum power point, I represents the current value on the simulated IV characteristic curve, V represents the voltage value on the simulated IV characteristic curve, VmRepresents the voltage at the maximum power point, ImShowing the current at the maximum power point, FF the fill factor, Isc_d1Represents a short-circuit current IscAt a first derivative value, Voc_d1Represents the open circuit voltage VocAnd a first derivative value is obtained, Pks _ num represents the number of peak points in the second derivative curve of the simulation IV characteristic curve after smoothing, and Vall _ num represents the number of valley points in the second derivative curve of the simulation IV characteristic curve after smoothing.
4. The method for photovoltaic array fault diagnosis based on K-nearest neighbor and IV curve as claimed in claim 3, wherein the normalizing the extracted characteristic parameters comprises:
wherein, s'ijRepresents the ith simulated IV characteristic curve, the jth normalized value of the characteristic parameter, sijRepresents the values of the ith simulated IV characteristic, the jth characteristic parameter,denotes the mean value, σ, of the jth characteristic parameterjAnd d represents the number of characteristic parameters in the extracted simulated IV characteristic curve, and m represents the number of simulated IV characteristic curves corresponding to different irradiance and temperature for each fault type.
5. The photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve as claimed in claim 4, wherein the dimension reduction processing is performed on the simulation IV characteristic curve feature vector by using t-distribution neighborhood embedding algorithm, comprising:
D1) and (3) reducing the dimension by taking the feature vector normalized by the feature parameter as raw data, wherein the raw data is expressed as:
D2) calculating the joint probability distribution of the low-dimensional mapping points:
wherein, s'iRepresents the ith row, s ' in the s ' matrix 'jRepresents the j-th row, σ, in the s' matrixiIs s'iGaussian variance of (1), Pj|iIs s'jTo s'iIs s'iAnd s'jSimilarity between high-dimensional data points;
D3) defining a joint probability distribution of high-dimensional data points:
D4) calculating low-dimensional similarity qij:
D5) Calculating K L divergence:
D6) calculate gradient minimization K L divergence:
D7) updating the initial solution to obtain the solution Y of the t iteration(t):
Where α (t) represents momentum, and ρ represents a learning rate;
D8) and (5) circularly iterating the steps D4) -D7) until the iteration W times are finished, and finally obtaining low-dimensional data:
Y(W)={y1,y2,y3…ym},
wherein, yiIs s'iThe vector after dimensionality reduction, i, is 1,2, … m.
6. The method for photovoltaic array fault diagnosis based on K-nearest neighbor and IV curve of claim 5, wherein the fault diagnosis of the current photovoltaic array is performed by adopting a K-nearest neighbor classification algorithm based on the training data set and the measured data set, and the method comprises the following steps:
E1) calculating the distance between the measured data set and the training data concentration point:
wherein d is12Representing the distance, x, between the feature vectors in the measured data set and the feature vectors in the training data set1iThe i-th feature component, x, representing a feature vector in the measured data set2iRepresenting the ith characteristic component of the characteristic vector in the training data set, and dim representing the dimensionality of the characteristic vector after dimensionality reduction processing;
E2) sorting m × n distance values according to ascending order, wherein n represents the number of fault types;
E3) selecting the first K points with the minimum value in the sorting;
E4) counting the frequency of the first K points in each fault category in a circulating mode;
E5) and returning the fault category with the highest occurrence frequency of the previous K points, and taking the fault category as the fault type of the current photovoltaic array.
7. The photovoltaic array fault diagnosis method based on K neighbors and IV curves according to claim 6, wherein K is an odd number.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010228062.3A CN111444615B (en) | 2020-03-27 | 2020-03-27 | Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010228062.3A CN111444615B (en) | 2020-03-27 | 2020-03-27 | Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111444615A true CN111444615A (en) | 2020-07-24 |
CN111444615B CN111444615B (en) | 2022-09-23 |
Family
ID=71648101
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010228062.3A Active CN111444615B (en) | 2020-03-27 | 2020-03-27 | Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111444615B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112164081A (en) * | 2020-09-30 | 2021-01-01 | 西南交通大学 | Method for extracting cross section contour of vehicle-mounted LiDAR point cloud railway |
CN112327189A (en) * | 2020-10-14 | 2021-02-05 | 北方工业大学 | KNN algorithm-based energy storage battery health state comprehensive judgment method |
CN112990059A (en) * | 2021-03-30 | 2021-06-18 | 安徽天能清洁能源科技有限公司 | IV curve diagnosis method and device based on curve shape description |
CN113037214A (en) * | 2021-02-02 | 2021-06-25 | 合肥工业大学 | Photovoltaic module shadow shielding fault diagnosis method based on S-V curve |
CN113094906A (en) * | 2021-04-13 | 2021-07-09 | 深圳市铂科新材料股份有限公司 | Method, device, equipment and medium for calculating direct current bias characteristic |
CN113326667A (en) * | 2021-08-03 | 2021-08-31 | 深圳格瑞特新能源有限公司 | Solar cell panel fault diagnosis method and system |
CN115455730A (en) * | 2022-09-30 | 2022-12-09 | 南京工业大学 | Photovoltaic module hot spot fault diagnosis method based on complete neighborhood preserving embedding |
CN116111948A (en) * | 2022-12-07 | 2023-05-12 | 中广核贵州安顺关岭新能源有限公司 | Photovoltaic branch equipment fault early warning system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108923748A (en) * | 2018-07-16 | 2018-11-30 | 河海大学常州校区 | A kind of diagnosing failure of photovoltaic array method based on IV curved scanning |
CN110008628A (en) * | 2019-04-17 | 2019-07-12 | 河海大学常州校区 | A kind of photovoltaic array fault parameter discrimination method |
TW202010243A (en) * | 2018-08-10 | 2020-03-01 | 魏榮宗 | Fault detection system and method for solar photovoltaic |
-
2020
- 2020-03-27 CN CN202010228062.3A patent/CN111444615B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108923748A (en) * | 2018-07-16 | 2018-11-30 | 河海大学常州校区 | A kind of diagnosing failure of photovoltaic array method based on IV curved scanning |
TW202010243A (en) * | 2018-08-10 | 2020-03-01 | 魏榮宗 | Fault detection system and method for solar photovoltaic |
CN110008628A (en) * | 2019-04-17 | 2019-07-12 | 河海大学常州校区 | A kind of photovoltaic array fault parameter discrimination method |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112164081A (en) * | 2020-09-30 | 2021-01-01 | 西南交通大学 | Method for extracting cross section contour of vehicle-mounted LiDAR point cloud railway |
CN112164081B (en) * | 2020-09-30 | 2023-04-21 | 西南交通大学 | Vehicle-mounted LiDAR point cloud railway cross section contour extraction method |
CN112327189A (en) * | 2020-10-14 | 2021-02-05 | 北方工业大学 | KNN algorithm-based energy storage battery health state comprehensive judgment method |
CN112327189B (en) * | 2020-10-14 | 2023-06-09 | 北方工业大学 | Comprehensive judging method for health state of energy storage battery based on KNN algorithm |
CN113037214B (en) * | 2021-02-02 | 2022-08-12 | 合肥工业大学 | Photovoltaic module shadow shielding fault diagnosis method based on S-V curve |
CN113037214A (en) * | 2021-02-02 | 2021-06-25 | 合肥工业大学 | Photovoltaic module shadow shielding fault diagnosis method based on S-V curve |
CN112990059A (en) * | 2021-03-30 | 2021-06-18 | 安徽天能清洁能源科技有限公司 | IV curve diagnosis method and device based on curve shape description |
CN113094906A (en) * | 2021-04-13 | 2021-07-09 | 深圳市铂科新材料股份有限公司 | Method, device, equipment and medium for calculating direct current bias characteristic |
CN113094906B (en) * | 2021-04-13 | 2024-06-11 | 深圳市铂科新材料股份有限公司 | Method, device, equipment and medium for calculating DC bias characteristics |
CN113326667A (en) * | 2021-08-03 | 2021-08-31 | 深圳格瑞特新能源有限公司 | Solar cell panel fault diagnosis method and system |
CN115455730A (en) * | 2022-09-30 | 2022-12-09 | 南京工业大学 | Photovoltaic module hot spot fault diagnosis method based on complete neighborhood preserving embedding |
CN115455730B (en) * | 2022-09-30 | 2023-06-20 | 南京工业大学 | Photovoltaic module hot spot fault diagnosis method based on complete neighborhood preserving embedding |
CN116111948A (en) * | 2022-12-07 | 2023-05-12 | 中广核贵州安顺关岭新能源有限公司 | Photovoltaic branch equipment fault early warning system |
CN116111948B (en) * | 2022-12-07 | 2024-04-23 | 中广核贵州安顺关岭新能源有限公司 | Photovoltaic branch equipment fault early warning system |
Also Published As
Publication number | Publication date |
---|---|
CN111444615B (en) | 2022-09-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111444615B (en) | Photovoltaic array fault diagnosis method based on K nearest neighbor and IV curve | |
CN109873610B (en) | Photovoltaic array fault diagnosis method based on IV characteristic and depth residual error network | |
CN109842373B (en) | Photovoltaic array fault diagnosis method and device based on space-time distribution characteristics | |
CN109766952B (en) | Photovoltaic array fault detection method based on partial least square method and extreme learning machine | |
Liu et al. | Fault diagnosis approach for photovoltaic array based on the stacked auto-encoder and clustering with IV curves | |
CN110008628B (en) | Photovoltaic array fault parameter identification method | |
CN111754026B (en) | Photovoltaic power station group power prediction method and device, computer equipment and storage medium | |
CN111327271B (en) | Photovoltaic array fault diagnosis method based on semi-supervised extreme learning machine | |
CN109670553B (en) | Photovoltaic array fault diagnosis method based on adaptive neural fuzzy inference system | |
CN110503153B (en) | Photovoltaic system fault diagnosis method based on differential evolution algorithm and support vector machine | |
CN110795690A (en) | Wind power plant operation abnormal data detection method | |
CN109145948A (en) | A kind of injection molding machine putty method for detecting abnormality based on integrated study | |
CN109992911B (en) | Photovoltaic module rapid modeling method based on extreme learning machine and IV characteristics | |
CN111625399A (en) | Method and system for recovering metering data | |
CN116629416A (en) | Photovoltaic power station power prediction method and device | |
CN116821832A (en) | Abnormal data identification and correction method for high-voltage industrial and commercial user power load | |
CN116435998A (en) | Prediction method of photovoltaic power generation power | |
CN114117921A (en) | Intelligent diagnosis method for faults of photovoltaic array | |
CN112598148A (en) | Fan variable pitch motor temperature fault early warning method based on collaborative expression and LightGBM algorithm | |
CN114118401A (en) | Neural network-based power distribution network flow prediction method, system, device and storage medium | |
CN111614489A (en) | Transient recording fault data cleaning method and system based on sparse self-coding | |
CN111277221A (en) | Photovoltaic fault diagnosis method and device | |
CN114841266A (en) | Voltage sag identification method based on triple prototype network under small sample | |
CN114818827A (en) | Non-invasive load decomposition method based on seq2point network | |
CN114066239A (en) | User power consumption abnormity detection method and device |
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 |