CN114614766B - Photovoltaic and photo-thermal integrated component abnormity detection method and test system - Google Patents
Photovoltaic and photo-thermal integrated component abnormity detection method and test system Download PDFInfo
- Publication number
- CN114614766B CN114614766B CN202210114898.XA CN202210114898A CN114614766B CN 114614766 B CN114614766 B CN 114614766B CN 202210114898 A CN202210114898 A CN 202210114898A CN 114614766 B CN114614766 B CN 114614766B
- Authority
- CN
- China
- Prior art keywords
- photovoltaic
- characteristic
- class
- characteristic curve
- fusion
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 25
- 238000012360 testing method Methods 0.000 title claims abstract description 14
- 239000013598 vector Substances 0.000 claims abstract description 52
- 230000002159 abnormal effect Effects 0.000 claims abstract description 46
- 238000005070 sampling Methods 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000005856 abnormality Effects 0.000 claims abstract description 21
- 238000005286 illumination Methods 0.000 claims abstract description 13
- 238000013528 artificial neural network Methods 0.000 claims abstract description 12
- 230000004927 fusion Effects 0.000 claims description 80
- 230000007613 environmental effect Effects 0.000 claims description 26
- 230000008569 process Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000003064 k means clustering Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 5
- 230000007547 defect Effects 0.000 description 38
- 238000007796 conventional method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004146 energy storage Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02S—GENERATION OF ELECTRIC POWER BY CONVERSION OF INFRARED RADIATION, VISIBLE LIGHT OR ULTRAVIOLET LIGHT, e.g. USING PHOTOVOLTAIC [PV] MODULES
- H02S50/00—Monitoring or testing of PV systems, e.g. load balancing or fault identification
- H02S50/10—Testing of PV devices, e.g. of PV modules or single PV cells
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
Landscapes
- Testing Of Individual Semiconductor Devices (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention relates to the field of photovoltaic system abnormity detection, in particular to a photovoltaic and photothermal integrated component abnormity detection method and a test system, which comprises the following steps: the method comprises the steps that the probability of all abnormal types and the uncertainty of characteristic curves are obtained by utilizing a neural network according to the characteristic curves, when the uncertainty is smaller than a preset threshold, the abnormal type with the maximum probability is used as an abnormal detection result, when the uncertainty of the characteristic curves is not smaller than the preset threshold, the first type formed by all the characteristic curves obtained in a preset time period is fused with the middle characteristic curve of each first type, the final type of each first type and the fused characteristic vector of each first type are obtained according to the effective degree and the characteristic vector of the fused result, and the abnormality of the photovoltaic photo-thermal component is obtained according to the fused characteristic vector of all the first types. The invention can avoid the condition of larger error of the characteristic curve caused by the change of illumination intensity and temperature and less sampling times, and ensures the accuracy of abnormal detection.
Description
Technical Field
The invention relates to the field of photovoltaic system abnormity detection, in particular to a photovoltaic and photothermal integrated assembly abnormity detection method and a test system.
Background
The photovoltaic and photothermal integrated component mainly comprises a photovoltaic part and a photothermal part. The photovoltaic part adopts a solar photovoltaic panel with mature technology, provides required electric energy for a building through a control system, and mainly comprises a photovoltaic cell, a storage battery, an inverter, a controller and other components. The photo-thermal part is mainly a heat collector, converts solar energy into heat energy, and simultaneously uses a thermal cycle mechanism to cool the solar cell, so that the photoelectric conversion efficiency is improved, and the solar heat energy is more efficiently utilized.
Photovoltaic light and heat subassembly can make the subassembly appear multiple unusual because natural environment's erosion and the ageing of subassembly in the use, for example subassembly surface is dirty, the latent disconnected grid of subassembly etc. these unusual make the subassembly very easily produce the hot spot phenomenon, increase conflagration hidden danger, also can cause the vat effect, lead to whole photovoltaic light and heat subassembly array power generation to descend by a wide margin, reduce the reliability and the life of subassembly.
In order to find out abnormal photovoltaic photo-thermal components in time, the components need to be monitored in a natural environment, the existing method is to obtain a characteristic curve of the photovoltaic photo-thermal components, also called an IV curve of the photovoltaic photo-thermal components, obtain the reasons or types of the abnormality generated by the components through the characteristic curve, and achieve the purposes of maintaining, replacing, cleaning and the like of the components. However, the characteristic curve of the photovoltaic photo-thermal module is generally obtained by sampling the voltage and current of the module under different loads under the fixed illumination intensity and temperature in the laboratory environment; due to interference in the natural environment and the fact that a certain time is spent in the process of acquiring the characteristic curve, normal operation of the whole assembly array cannot be influenced when voltage and current of the assembly under different loads are sampled, so that the illumination intensity and temperature cannot be guaranteed to be unchanged when the characteristic curve is acquired, the sampling times cannot be too many, the acquired characteristic curve has large errors, the acquired abnormal type has errors, and the method is not beneficial to timely discovering the abnormity of the assembly and formulating an abnormity processing method, so that loss is caused.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a photovoltaic and photothermal integrated component abnormity detection method and a test system, and the adopted technical scheme is as follows:
the invention provides a photovoltaic and photothermal integrated component abnormity detection method, which comprises the following steps:
in a time period when the sun directly irradiates the photovoltaic and photo-thermal component, obtaining a characteristic curve of the photovoltaic and photo-thermal component, taking a two-dimensional vector formed by combining the average illumination intensity and the average temperature in the process of obtaining the characteristic curve as an environmental characteristic of the characteristic curve, inputting the characteristic curve obtained last time at the current moment and the environmental characteristic thereof into a neural network to obtain probabilities of all abnormal types, taking the entropy of the probabilities of all abnormal types as the uncertainty of the characteristic curve, and taking the abnormal type with the maximum probability as an abnormal detection result of the photovoltaic and photo-thermal component when the uncertainty is smaller than a preset threshold;
when the uncertainty of the characteristic curve is greater than or equal to a preset threshold value, acquiring all characteristic curves and environmental characteristics of all the characteristic curves obtained by the photovoltaic photo-thermal component in a preset time period, performing mean shift clustering on all the characteristic curves according to the environmental characteristics of the characteristic curves and acquiring all first classes, firstly, fusing every two characteristic curves in each first class to acquire all the fused characteristic curves, and a characteristic vector and an effective degree of each fused characteristic curve, then, taking all the fused characteristic curves with the effective degree greater than zero as an effective set of each first class, clustering all the fused characteristic curves in the effective set of each first class to acquire all second classes of each first class, and fusing the fused characteristic curves in all the second classes to acquire a final class of each first class;
and for each final class of the first classes, taking the effective degree of each fusion characteristic curve in the final class as a weight, carrying out weighted summation on the characteristic vectors of all the fusion characteristic curves in the final class, taking the obtained result as the fusion characteristic vector of each first class, and obtaining the abnormity of the photovoltaic photo-thermal component according to the fusion characteristic vectors of all the first classes.
Further, the step of obtaining the characteristic curve of the photovoltaic photothermal element comprises:
the photovoltaic photo-thermal assembly is enabled to work under loads of preset number distributed at random, current and voltage output by the photovoltaic photo-thermal assembly under each load are obtained, the current and the voltage are regarded as a sampling point, all sampling points obtained under the loads of the preset number form a sampling point sequence, and the sampling point sequence is used as a characteristic curve of the photovoltaic photo-thermal assembly.
Further, the step of fusing every two characteristic curves in each first category to obtain all fused characteristic curves includes:
for each first class, combining all characteristic curves in the first class pairwise to obtain all combination results, combining two corresponding sampling point sequences into one sampling point sequence for two characteristic curves in each combination result, and taking the sampling point sequence as a fusion characteristic curve of each combination result, wherein the fusion characteristic curves of all combination results are all obtained fusion characteristic curves.
Further, the step of obtaining the feature vector and the validity degree of each fusion characteristic curve includes:
calculating the mean value of the uncertainty of the two characteristic curves in the combined result for the combined result corresponding to each fusion characteristic curve, and calling the mean value of the environmental characteristics of the two characteristic curves in the combined result as the environmental characteristics of each fusion characteristic curve, inputting the fusion characteristic curves and the environmental characteristics thereof into a neural network to obtain the probabilities of all abnormal types, and calling the vector formed by the probabilities of all abnormal types as the characteristic vector of each fusion characteristic curve;
and taking the entropy of the probabilities of all the abnormal types as the uncertainty of each fusion characteristic curve, and taking the difference between the mean value and the uncertainty of each fusion characteristic curve as the effective degree of each fusion characteristic curve.
Further, the step of obtaining all the second classes of each first class includes:
and for all the fusion characteristic curves in the effective set of each first class, the distance between any two fusion characteristic curves is the Euclidean distance of the corresponding characteristic vectors, all the fusion characteristic curves are clustered by using a K-Means clustering algorithm according to the distance between the fusion characteristic curves, the number of the classes is half of that of all the fusion characteristic curves, and all the obtained clustering results are all the second classes of each first class.
Further, the step of obtaining the final category of the first category includes:
and for all the second classes of each first class, fusing all the fusion characteristic curves in each second class into a characteristic curve, namely the characteristic curve of each second class, acquiring a set formed by the characteristic curves of all the second classes, regarding the set as a new first class, then reusing the new first class to obtain all the second classes of the first class, repeating the steps until all the second classes of the first class only contain one class, and regarding the class as the final class of the first class.
Further, the step of obtaining the abnormality of the photovoltaic photothermal element according to the fused eigenvectors of all the first categories comprises:
for all the first-class fusion feature vectors, calculating entropy of probabilities of all the abnormal types represented by each fusion feature vector, namely uncertainty of each fusion feature vector, and acquiring the minimum value of the uncertainty of all the fusion feature vectors and the probabilities of all the abnormal types represented by the fusion feature vectors when the uncertainty is the minimum value;
and when the minimum value is smaller than a preset threshold value, the abnormal type with the maximum probability is used as an abnormal detection result of the photovoltaic photo-thermal assembly, when the minimum value is larger than or equal to the preset threshold value, the characteristic curve of the photovoltaic photo-thermal assembly is obtained again, and after the characteristic curve of the photovoltaic photo-thermal assembly is obtained again, the abnormality of the photovoltaic photo-thermal assembly is obtained again.
Further, the preset time period is a time period from when the uncertainty of the characteristic curve of the photovoltaic photo-thermal component is greater than or equal to a preset threshold value to when the current time is finished.
The invention also provides a photovoltaic and photo-thermal integrated component testing system which comprises, but is not limited to, a photovoltaic and photo-thermal component load switching module, a current and voltage detection module, an illumination intensity and environment temperature detection module and a storage and calculation module.
The invention has the following beneficial effects: according to the method, all characteristic curves of the photovoltaic photo-thermal component in a period of time are obtained, clustering and fusion are carried out on all characteristic curves according to the uncertainty of the characteristic curves, the abnormal detection result of the photovoltaic photo-thermal component is further obtained according to the effective degree and the characteristic vector of the fusion result, the situation that the error of the characteristic curves is large due to the fact that the illumination intensity and the temperature cannot be guaranteed to be unchanged and the sampling times are not large when the characteristic curves are obtained can be avoided, and the accuracy of abnormal detection is guaranteed.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a method for detecting an abnormality of a pv-photothermal integrated module according to an embodiment of the present invention;
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, the following detailed description of the method and system for detecting abnormality of integrated pv-photothermal device according to the present invention with reference to the accompanying drawings and preferred embodiments will be given below. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the abnormality detection method for the photovoltaic and photothermal integrated component provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a method for detecting an abnormality of a pv-photothermal integrated module according to the present invention is shown, which includes the following steps:
and S001, constructing a test system of the photovoltaic photo-thermal assembly, and acquiring a characteristic curve of the photovoltaic photo-thermal assembly.
The photovoltaic photo-thermal component does not change the direction of the component along with the movement of the sun, but the fixed angle of the component always faces to one direction, a time period in which the sun almost vertically irradiates a component panel in one day is artificially obtained, the time period is called a direct irradiation time period, the working efficiency of the photovoltaic component is the highest empirically in the direct irradiation time period, and the obtained characteristic curve is the most accurate and reliable, for example, the direct irradiation time period is 11-14 hours.
The testing system is arranged on each assembly, the testing system can change the load of the assembly at will and can acquire the voltage and the current output by the assembly under any load, the change curve of the voltage and the current obtained under all loads is a characteristic curve, the current when the load is 0 is short-circuit current, and the voltage when the load is infinite is open-circuit voltage. Usually, the load size cannot be taken from 0 to infinity when acquiring the characteristic curve, but a limited number of loads are sampled, for example, 50-150 load sizes are sampled, then the current and voltage output by the component under the loads are acquired, namely 50-150 (current and voltage) are acquired, each group of current and voltage is called a sampling point, namely 50-150 sampling points, all the formed sampling points of the current and voltage describe a curve under a two-dimensional coordinate system, the curve reflects the characteristic curve, and the sequence formed by the sampling points is called the characteristic curve.
However, considering that when the load of the component is changed, the whole component array is under-voltage or the output power is unstable, the normal operation of the component is affected, and the fluctuation of the whole photovoltaic power grid is caused, the existing solution is to provide power compensation for the whole photovoltaic component and the photovoltaic power grid by adding an energy storage device, because the energy storage device needs to be charged in time, the load cannot be changed too frequently, the number of load samples cannot be too many, and because a certain time is needed for obtaining a characteristic curve by changing the load, the fluctuation of the power grid is easily caused if the time is too short; based on the above consideration, in order to maintain the stability of the power grid when the detection assembly is abnormal, the process of obtaining the characteristic curve needs a certain time and too many sampling times cannot be provided, so that the illumination and temperature cannot be guaranteed to be unchanged and the distribution of sampling points is reasonable in the process of obtaining the characteristic curve, and the characteristic curve has certain errors and noise.
In addition, the test system also comprises an illumination sensor used for acquiring illumination intensity in real time, and a temperature sensor used for acquiring the environmental temperature; in addition, the test system also has data storage and data calculation functions.
In a direct irradiation time period of each day, a plurality of characteristic curves of the photovoltaic photo-thermal component are obtained by using the test system, for example, 5-8 characteristic curves are obtained by using the test system in the direct irradiation time period, and a two-dimensional vector formed by combining the average illumination intensity and the average temperature of each characteristic curve in the obtaining process is called as the environmental characteristic of each characteristic curve.
And S002, obtaining the probability of each defect type of the component according to the environmental characteristics of the characteristic curve of the component.
The characteristic curve of the photovoltaic photo-thermal component can reflect the abnormal type of the photovoltaic photo-thermal component, for example, when a battery piece is damaged or a bypass diode is short-circuited, the characteristic curve has a plurality of segments, and the voltage on the segments can change rapidly along with the current; the short-circuit current of the characteristic curve becomes smaller when the component is aged and the power is attenuated; further, for example, when the component is shielded by a shadow or a dead object, the open circuit voltage of the characteristic curve becomes low.
Because the characteristic curve of the invention contains a small number of sampling points, the invention utilizes a nonlinear difference method to expand the sampling points on the characteristic curve to 300, for example.
In the invention, the characteristic curve of the photovoltaic photothermal component and the environmental characteristics thereof are input into the neural network, and the probability of each abnormal type existing on the photovoltaic photothermal component output by the neural network is also regarded as an abnormal type under the condition of no defect, namely, the abnormal type without abnormality, and other abnormal types comprise: foreign object shielding, bypass diode short circuits, power attenuation, and the like. Since the characteristic curve is a sequence of sampling points, the neural network of the present invention is an LSTM neural network. The method for acquiring the data set of the network comprises the following steps: obtaining components with different anomalies in a laboratory environment, then obtaining a characteristic curve of each component under different illumination intensities and environment temperatures, taking each characteristic curve, the illumination intensity and the environment temperature as data, taking one-hot codes of the anomaly types of the corresponding components as labels of the data, further constructing a data set, and finally training the neural network by using the data set and utilizing a random gradient descent method.
And S003, obtaining the abnormal type of the photovoltaic photo-thermal component according to the characteristic curve of the photovoltaic photo-thermal component and the probability of each abnormal type of the component.
The newly obtained characteristic curve of the component at the current moment is firstly input into the neural network to obtain the probability of each defect type on the component.
The conventional method is to take the defect type with the highest probability as the defect and the abnormity of the component, but the conventional method is inaccurate because the characteristic curve has certain errors and noises. In the invention, for all defect types, if the probability corresponding to only a few defect types is very high and the probability corresponding to most defect types approaches to 0, the characteristic curve is shown to have a definite defect type, namely, the entropy of the probability of all defect types is smaller, the characteristic curve is shown to have a definite defect type, and the defects and the abnormity of the component can be obtained by using a conventional method; on the contrary, if the entropy of the probability of all defect types is larger, it means that the defect type of the characteristic curve cannot be determined, which may be caused by the fact that the influence degree of the defect is smaller, the defect feature is disturbed and submerged by the error and noise of the characteristic curve, and the defect type cannot be output accurately.
Then, based on the above, for each probability of defect type obtained on a component, the entropy of all the probabilities of defect types is calculated, which is referred to as uncertainty of the characteristic curve of the component in the present invention; when the uncertainty is smaller than a preset threshold value, acquiring a defect type with the maximum probability, and taking the defect type as the component abnormality, wherein the larger the maximum probability is, the more serious the influence of the abnormality on the component is, and the more the defect type needs to be repaired and replaced in time; in the invention, a preset threshold value is set to be 0.2; when the uncertainty is greater than or equal to a preset threshold, the existing abnormal type cannot be determined, and then the method is obtained by the following method:
for a component with uncertainty greater than or equal to a preset threshold, acquiring all characteristic curves of the component within a preset time period; the preset time period refers to all the time periods from the moment when the uncertainty of the detected component characteristic curve is larger than or equal to the preset threshold value to the current moment. The method comprises the following steps of obtaining environmental features of all characteristic curves, clustering all the environmental features by using a mean shift clustering algorithm to obtain all categories, wherein the categories are collectively called as first categories, the environmental features in the same first category are approximately the same, the environmental features in different first categories are greatly different, and in order to avoid the influence of environmental difference on detection of component defects and abnormity, each first category is analyzed by the method, and the method comprises the following specific steps: 1) For any one of the first category, the mean of the uncertainties of any two characteristic curves therein, called two special characteristics, is obtained
The average uncertainty of the characteristic curves, and then fusing the two characteristic curves, wherein the fusion result is called a fusion characteristic curve; the fusion method comprises the following steps: the two characteristic curves are respectively corresponding to two sampling point sequences, and the two sampling point sequences are combined into one sampling point sequence which is a fusion characteristic curve.
Compared with any two characteristic curves, the fused characteristic curve has more sampling points, and can accurately and finely describe and characterize the defects and abnormal characteristics of the photovoltaic photo-thermal component.
2) And the environmental features corresponding to the fusion characteristic curves are the average value of the environmental features of the two characteristic curves, and the fusion characteristic curves and the corresponding environmental features are input into the neural network to obtain the probability of each defect type. The invention combines the probabilities of all defect types into a vector, which is called as a characteristic vector of a fusion characteristic curve;
3) And then calculating the uncertainty of the fusion characteristic curve, and then acquiring the difference between the average uncertainty of the two characteristic curves and the uncertainty of the fusion characteristic curve, wherein the difference is called as the effective degree of the fusion characteristic curve, and the larger the effective degree is, the fusion characteristic curve has more definite defects and abnormal characteristics compared with the two characteristic curves.
So far, every two characteristic curves correspondingly obtain a fusion characteristic curve and an effective degree.
4) For the first class, all characteristic curves in the class are combined pairwise to obtain all fusion characteristic curves and the effective degree of each fusion characteristic curve, all fusion characteristic curves with the effective degree larger than 0 are obtained, a set formed by the fusion characteristic curves is marked as S, the distance between any two fusion characteristic curves in the S is the Euclidean distance of the corresponding characteristic vector, then all fusion characteristic curves in the S are clustered by using a K-Means clustering algorithm according to the distance between all fusion characteristic curves, the clustering quantity is half of the quantity of all fusion characteristic curves, all clustering results are collectively called as a second class, and the fusion characteristic curves in the same second class have approximate characteristic vectors.
5) And then all the fusion characteristic curves in the same second category are fused into one characteristic curve again, still called as a fusion characteristic curve, each second category in all the second categories corresponds to one fusion characteristic curve, all the fusion curves corresponding to all the second categories are used as one first category, and the steps 1), 2), 3), 4) and 5) are continuously repeated until only one second category is obtained in the step 4), and the second category is called as a final category.
The obtained fusion characteristic curve in the final category has more definite defects and abnormal characteristics, the defects and the abnormal characteristics of the photovoltaic photo-thermal assembly can be more accurately reflected, and the interference of noise and errors existing in the characteristic curve can be eliminated to a certain degree.
6) Further, assume that there are N fused feature curves in the final class, where the feature vector of the nth fused feature curve is v n Corresponding to an effective degree of w n Then order the fused feature vectors
It is known that v is the result of weighted fusion of the feature vectors of all the fused characteristics in the final class, wherein the greater the significance, the greater the weight of the feature vectors, i.e. feature vectors with significantly increased fused characteristics with a greater focus on defect and anomaly certainty. Since the feature vector describes the probability of occurrence of each defect type, the fused feature vector v also describes the probability of occurrence of each defect type, but the fused feature vector v is more capable of reflecting the defects and abnormal features of the photovoltaic photo-thermal component.
Therefore, after the analysis of the above steps is performed on each first category, a fused feature vector is obtained for each first category, then for all the first categories, a fused feature vector is obtained by the above steps, and the probability that each fused feature vector represents the occurrence of each defect type is obtained, so that each fused feature vector corresponds to the uncertainty in step S003, a fused feature vector with the minimum uncertainty is obtained, and if the uncertainty of the fused feature vector is smaller than a preset threshold, the defect type with the maximum probability is obtained according to the fused feature vector and is used as the defect and the abnormality of the photovoltaic and photothermal component; if the uncertainty is greater than or equal to the preset threshold, the characteristic curve of the photovoltaic photo-thermal assembly is obtained again, the step S003 of the invention is implemented again until the defect type of the photovoltaic photo-thermal assembly is obtained, and then the defects and the abnormity of the assembly are determined.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (9)
1. The abnormal detection method of the photovoltaic and photothermal integrated assembly is characterized by comprising the following steps of:
in a time period when the sun directly irradiates the photovoltaic and photo-thermal component, acquiring a characteristic curve of the photovoltaic and photo-thermal component, taking a two-dimensional vector formed by combining the average illumination intensity and the average temperature in the process of acquiring the characteristic curve as the environmental characteristic of the characteristic curve, inputting the characteristic curve acquired last time at the current moment and the environmental characteristic thereof into a neural network to acquire the probabilities of all abnormal types, taking the entropy of the probabilities of all abnormal types as the uncertainty of the characteristic curve, and taking the abnormal type with the maximum probability as the abnormal detection result of the photovoltaic and photo-thermal component when the uncertainty is smaller than a preset threshold;
when the uncertainty of the characteristic curve is greater than or equal to a preset threshold value, acquiring all characteristic curves and environmental characteristics of all the characteristic curves obtained by the photovoltaic photo-thermal component in a preset time period, performing mean shift clustering on all the characteristic curves according to the environmental characteristics of the characteristic curves and acquiring all first classes, firstly, fusing every two characteristic curves in each first class to acquire all the fused characteristic curves, and a characteristic vector and an effective degree of each fused characteristic curve, then, taking all the fused characteristic curves with the effective degree greater than zero as an effective set of each first class, clustering all the fused characteristic curves in the effective set of each first class to acquire all second classes of each first class, and fusing the fused characteristic curves in all the second classes to acquire a final class of each first class;
and for each final class of the first classes, taking the effective degree of each fusion characteristic curve in the final class as a weight, carrying out weighted summation on the characteristic vectors of all the fusion characteristic curves in the final class, taking the obtained result as the fusion characteristic vector of each first class, and obtaining the abnormity of the photovoltaic photo-thermal component according to the fusion characteristic vectors of all the first classes.
2. The method for detecting the abnormality of the integrated photovoltaic and photothermal module according to claim 1, wherein the step of obtaining the characteristic curve of the integrated photovoltaic and photothermal module comprises:
the photovoltaic photo-thermal assembly is enabled to work under loads of a preset number distributed randomly, current and voltage output by the photovoltaic photo-thermal assembly under each load are obtained, the current and the voltage are regarded as a sampling point, all the sampling points obtained under the loads of the preset number form a sampling point sequence, and the sampling point sequence is used as a characteristic curve of the photovoltaic photo-thermal assembly.
3. The method for detecting the abnormality of the integrated photovoltaic-thermal module according to claim 1, wherein the step of fusing every two characteristic curves in each first category to obtain all fused characteristic curves comprises the following steps:
for each first class, combining all characteristic curves in the first class pairwise to obtain all combination results, combining two corresponding sampling point sequences into one sampling point sequence for two characteristic curves in each combination result, and taking the sampling point sequence as a fusion characteristic curve of each combination result, wherein the fusion characteristic curves of all combination results are all obtained fusion characteristic curves.
4. The method for detecting the abnormality of the integrated photovoltaic-thermal module according to claim 1, wherein the step of obtaining the feature vector and the validity degree of each fusion characteristic curve comprises the steps of:
calculating the mean value of the uncertainty of the two characteristic curves in the combined result for the combined result corresponding to each fusion characteristic curve, and calling the mean value of the environmental characteristics of the two characteristic curves in the combined result as the environmental characteristics of each fusion characteristic curve, inputting the fusion characteristic curves and the environmental characteristics thereof into a neural network to obtain the probabilities of all abnormal types, and calling the vector formed by the probabilities of all abnormal types as the characteristic vector of each fusion characteristic curve;
and taking the entropy of the probabilities of all the abnormal types as the uncertainty of each fusion characteristic curve, and taking the difference between the mean value and the uncertainty of each fusion characteristic curve as the effective degree of each fusion characteristic curve.
5. The method for detecting the abnormality of the integrated pv-thermal module according to claim 1, wherein the step of obtaining all the second categories of each of the first categories comprises:
and for all the fusion characteristic curves in the effective set of each first category, the distance between any two fusion characteristic curves is the Euclidean distance of the corresponding characteristic vector, all the fusion characteristic curves are clustered by using a K-Means clustering algorithm according to the distance between the fusion characteristic curves, the number of the categories is half of that of all the fusion characteristic curves, and all the obtained clustering results are all the second categories of each first category.
6. The method for detecting the abnormality of the integrated photovoltaic-thermal module according to claim 1, wherein the step of obtaining the final category of the first category comprises:
and for all the second classes of each first class, fusing all the fusion characteristic curves in each second class into a characteristic curve, namely the characteristic curve of each second class, acquiring a set formed by the characteristic curves of all the second classes, regarding the set as a new first class, then reusing the new first class to obtain all the second classes of the first class, repeating the steps until all the second classes of the first class only contain one class, and regarding the class as the final class of the first class.
7. The method for detecting the abnormality of the integrated pv-photothermal module according to claim 1, wherein the step of obtaining the abnormality of the pv-photothermal module according to the fused eigenvectors of all the first categories comprises:
for all the first-class fusion feature vectors, calculating entropy of probabilities of all the abnormal types represented by each fusion feature vector, namely uncertainty of each fusion feature vector, and acquiring the minimum value of the uncertainty of all the fusion feature vectors and the probabilities of all the abnormal types represented by the fusion feature vectors when the uncertainty is the minimum value;
and when the minimum value is smaller than a preset threshold value, the abnormal type when the probability is maximum is used as an abnormal detection result of the photovoltaic photo-thermal assembly, when the minimum value is larger than or equal to the preset threshold value, the characteristic curve of the photovoltaic photo-thermal assembly is obtained again, and after the characteristic curve of the photovoltaic photo-thermal assembly is obtained again, the abnormality of the photovoltaic photo-thermal assembly is obtained again.
8. The method for detecting the abnormality of the integrated photovoltaic-thermal module according to claim 1, wherein the preset time period is a period from when the uncertainty of the characteristic curve of the photovoltaic-thermal module is greater than or equal to a preset threshold value to when the current time is over.
9. A photovoltaic and photo-thermal integrated component testing system is characterized by comprising a photovoltaic and photo-thermal component load switching module, a current and voltage detection module, a light intensity and environment temperature detection module and a storage and calculation module, so as to realize the abnormal detection method of the photovoltaic and photo-thermal integrated component as claimed in any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210114898.XA CN114614766B (en) | 2022-01-31 | 2022-01-31 | Photovoltaic and photo-thermal integrated component abnormity detection method and test system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210114898.XA CN114614766B (en) | 2022-01-31 | 2022-01-31 | Photovoltaic and photo-thermal integrated component abnormity detection method and test system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114614766A CN114614766A (en) | 2022-06-10 |
| CN114614766B true CN114614766B (en) | 2022-11-29 |
Family
ID=81859468
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210114898.XA Active CN114614766B (en) | 2022-01-31 | 2022-01-31 | Photovoltaic and photo-thermal integrated component abnormity detection method and test system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114614766B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115081552B (en) * | 2022-07-28 | 2022-11-11 | 一道新能源科技(衢州)有限公司 | Solar cell data exception handling method and system based on cloud platform |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109004903A (en) * | 2018-08-14 | 2018-12-14 | 中国计量大学 | Fault detection system and method suitable for photovoltaic array |
| JP2020124090A (en) * | 2019-01-31 | 2020-08-13 | 株式会社カネカ | Abnormality diagnosis method, abnormality diagnosis system and computer program |
| CN113489457A (en) * | 2021-06-01 | 2021-10-08 | 厦门科灿信息技术有限公司 | Photovoltaic module fault detection method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109150104A (en) * | 2018-08-10 | 2019-01-04 | 江南大学 | A kind of diagnosing failure of photovoltaic array method based on random forests algorithm |
-
2022
- 2022-01-31 CN CN202210114898.XA patent/CN114614766B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109004903A (en) * | 2018-08-14 | 2018-12-14 | 中国计量大学 | Fault detection system and method suitable for photovoltaic array |
| JP2020124090A (en) * | 2019-01-31 | 2020-08-13 | 株式会社カネカ | Abnormality diagnosis method, abnormality diagnosis system and computer program |
| CN113489457A (en) * | 2021-06-01 | 2021-10-08 | 厦门科灿信息技术有限公司 | Photovoltaic module fault detection method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114614766A (en) | 2022-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Random forest based intelligent fault diagnosis for PV arrays using array voltage and string currents | |
| Dhimish | Assessing MPPT techniques on hot-spotted and partially shaded photovoltaic modules: Comprehensive review based on experimental data | |
| KR101882106B1 (en) | Method and apparatus for predicting the generated energy of the solar cell module | |
| Zaki et al. | Deep‐learning–based method for faults classification of PV system | |
| Ma et al. | Photovoltaic module current mismatch fault diagnosis based on IV data | |
| Dhimish et al. | Novel open-circuit photovoltaic bypass diode fault detection algorithm | |
| CN113037214B (en) | Photovoltaic module shadow shielding fault diagnosis method based on S-V curve | |
| CN114614766B (en) | Photovoltaic and photo-thermal integrated component abnormity detection method and test system | |
| Liu et al. | Intelligent fault diagnosis of photovoltaic array based on variable predictive models and I–V curves | |
| Ding et al. | Fault detection of photovoltaic array based on Grubbs criterion and local outlier factor | |
| CN111460644A (en) | Photovoltaic array monitoring system with fault diagnosis and power prediction functions | |
| Djalab et al. | Robust method for diagnosis and detection of faults in photovoltaic systems using artificial neural networks | |
| Shou et al. | Defect detection with generative adversarial networks for electroluminescence images of solar cells | |
| Serrano-Luján et al. | Case of study: Photovoltaic faults recognition method based on data mining techniques | |
| Kumaradurai et al. | Fault detection in photovoltaic systems using machine learning algorithms: A review | |
| Zahraoui et al. | System‐Level Condition Monitoring Approach for Fault Detection in Photovoltaic Systems | |
| CN116436405A (en) | Hot spot fault diagnosis method for photovoltaic string | |
| Trinh et al. | Application of Artificial Intelligence in Detecting and Classifying Faults of Solar Panels | |
| CN111711417B (en) | Method for diagnosing abnormal power generation amount in photovoltaic system | |
| Kumar et al. | Supervised machine leaning based anomaly detection and diagnosis in grid connected photovoltaic systems | |
| Cao et al. | A fault detection scheme for PV panels in large scale PV stations with complex installation conditions | |
| Cao et al. | A Fault Detection Scheme for PV Modules in Large Scale PV Stations With Complex Installation Conditions | |
| Voutsinas et al. | A survey of fault detection and identification methods for Photovoltaic systems based on IV curves | |
| Muñoz et al. | Fault detection and isolation in a photovoltaic system | |
| Attouri et al. | Faults classification in grid-connected photovoltaic systems |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A Anomaly Detection Method and Testing System for Photovoltaic Photothermal Integrated Modules Effective date of registration: 20231016 Granted publication date: 20221129 Pledgee: Yangzhou Branch of Zheshang Bank Co.,Ltd. Pledgor: Yangzhou Jinghua New Energy Technology Co.,Ltd. Registration number: Y2023980061180 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |