A method, system and device for detecting contamination layer on the surface of photovoltaic modules

Abstract

The present invention discloses a method, system and device for detecting the surface dirt layer of a photovoltaic module. The detection method includes: selecting a single photovoltaic module for cleaning and obtaining a sample of the cleaned photovoltaic module; obtaining hyperspectral images of all photovoltaic modules to be detected by using a hyperspectral imager mounted on an unmanned aerial vehicle; preprocessing the obtained hyperspectral images, including reflectivity correction and outlier removal; extracting the endmember spectrum of the photovoltaic module from the hyperspectral image of the clean photovoltaic module sample; performing sparse constrained mixed pixel decomposition of the hyperspectral image of the photovoltaic module area with a surface dirt layer, where some endmembers are known, to obtain the endmember spectrum of the surface dirt layer and the corresponding abundance matrix; and determining the composition and occlusion ratio of the surface dirt layer. Using the endmember spectrum information provided by the hyperspectral image, sparse constrained mixed pixel decomposition with some endmembers known is used to obtain the endmember spectrum of the surface dirt layer and the corresponding abundance matrix, so as to achieve accurate detection of the state of the surface dirt layer of the photovoltaic module.

Description

Technical Field

[0001] The present invention relates to the technical field of optical image processing, and more specifically, to a method, system and device for detecting a contamination layer on the surface of a photovoltaic module. Background Art

[0002] With the rapid development of photovoltaic power generation, the safe operation and maintenance and efficient and optimized operation of photovoltaic power stations face greater challenges. In photovoltaic power stations, the problem of dirt layer on the surface of photovoltaic modules is an important factor affecting the stable, safe, efficient and economical operation of the system. In its operating environment, various dirt such as dust, sand, bird droppings, fallen leaves, etc. are deposited on the surface of photovoltaic modules to form a shielding dirt layer. On the one hand, the surface dirt layer leads to a decrease in the transmittance and irradiance of photovoltaic modules, thereby limiting the photoelectric conversion efficiency and output power of photovoltaic cells. On the other hand, the non-uniform dirt layer forms point-shaped and strip-shaped shielding on the surface of the module, which is prone to produce hot spot effects, seriously affecting the operation safety and service life of photovoltaic power stations.

[0003] At present, in the operation and maintenance of photovoltaic power stations, there are technical solutions that use drones equipped with visible light and infrared image sensors to obtain visible light and infrared image data of photovoltaic modules, and combine computer vision technology to realize fault detection of photovoltaic modules. The above solution can detect point-shaped, strip-shaped shielding and hot spot problems, but it cannot effectively obtain quantitative status information such as surface contamination layer component type and shielding ratio. For the problem of surface contamination of photovoltaic modules, traditional monitoring and operation methods are still mainly based on experience judgment and manual observation. The status of the contamination layer is obtained through qualitative analysis, which is difficult to meet the needs of efficient, real-time, accurate and effective intelligent monitoring and operation. Existing cleaning solutions such as pipeline flushing, cleaning robot operations, and cleaning vehicle operations still lack effective information guidance for cleaning date scheduling, spatial path planning and operation management. It is difficult to achieve a good balance between power generation efficiency and operation and maintenance costs by implementing a simple regular cleaning strategy, which is easy to cause waste of operation and maintenance manpower and cleaning resources.

[0004] Therefore, it is necessary to propose a new solution to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies of the above-mentioned prior art and to provide a method, system and device for detecting the surface contamination layer of a photovoltaic module. The method utilizes the endmember spectral information provided by a hyperspectral image and adopts sparse constrained mixed pixel decomposition with some endmembers known to obtain the endmember spectrum and the corresponding abundance matrix of the surface contamination layer, so as to effectively realize the accurate detection of the surface contamination state of the photovoltaic module.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for detecting a contamination layer on a photovoltaic module surface, the method comprising:

[0008] Step 1, selecting a single photovoltaic module for cleaning and obtaining a cleaned photovoltaic module sample;

[0009] Step 2, using a drone equipped with a hyperspectral imager to obtain hyperspectral images of all photovoltaic modules to be inspected;

[0010] Step 3, preprocessing the acquired hyperspectral image, including reflectance correction and outlier removal;

[0011] Step 4, extracting the endmember spectrum of the photovoltaic module from the hyperspectral image of the clean photovoltaic module sample;

[0012] Step 5, performing sparse constrained mixed pixel decomposition with some known endmembers on the hyperspectral image of the photovoltaic module area with the surface contamination layer to obtain the endmember spectrum of the surface contamination layer and the corresponding abundance matrix;

[0013] Step 6: Determine the composition and occlusion ratio of the surface dirt layer.

[0014] Furthermore, in step 3, the reflectivity correction calculation method is:

[0015] ,

[0016] In the formula, For the wavelength The reflectivity at For the wavelength The radiant brightness at is the dark current response obtained by measuring the black calibration plate, is the radiant brightness measured for a white reference plate, is the known reflectivity value of the white reference plate.

[0017] Furthermore, in step 3, the method for removing outliers is:

[0018] ,

[0019] In the formula, For the wavelength The reflectivity at .

[0020] Furthermore, in step 4, the endmember spectrum extraction method adopts one of the pure pixel index, single body growth, and vertex component analysis, and the estimated endmember spectrum of the photovoltaic module is .

[0021] Furthermore, in step 5, the sparse constrained mixed pixel decomposition with some endmembers known includes the following steps:

[0022] Step 5.1: Extract the initial end member value of the photovoltaic module with surface contamination layer by end member extraction method ;

[0023] Step 5.2, Each end member is calculated with The similarity of each end member in The end member with the highest similarity is replaced by After rearrangement, the end member matrix of the photovoltaic module with surface contamination is constructed. ,in It remains fixed during the subsequent solution process. It represents the end-member spectrum of the surface pollution layer and needs to be solved and updated;

[0024] Step 5.3, estimate the PV panel abundance matrix by fully constrained least squares method and surface pollution layer abundance matrix The initial value of

[0025] Step 5.4, repeat the iteration and update by the following calculation formula , and ,

[0026] ,

[0027] ,

[0028] ,

[0029] In the formula, is the sparsity regularization constraint parameter.

[0030] Furthermore, in step 6, determining the composition and occlusion ratio of the surface dirt layer includes the following steps:

[0031] Step 6.1: Surface pollution endmember spectroscopy Matching the spectrum with the spectrum library to determine the component type of the surface pollution layer; the spectrum matching uses one or more of the spectral angle distance, correlation coefficient, Euclidean distance, and similarity coefficient as a similarity measure;

[0032] Step 6.2, based on the solved surface pollution layer abundance matrix Calculate the occlusion ratio of the surface dirt layer. The calculation formula is as follows:

[0033] ,

[0034] ,

[0035] In the formula, For the The occlusion ratio of the surface dirt layer, is the total occlusion ratio of the surface dirt layer.

[0036] The present invention also provides a system for detecting a contamination layer on the surface of a photovoltaic module, the system comprising:

[0037] A hyperspectral image acquisition unit, used for photographing the photovoltaic assembly to be inspected to acquire a hyperspectral image;

[0038] A preprocessing unit, used for preprocessing the hyperspectral image by correcting the reflectivity and removing abnormal values;

[0039] An end member extraction unit, used for extracting the end member spectrum of the photovoltaic module and the initial value of the end member spectrum of the photovoltaic module with a surface contamination layer;

[0040] A mixed pixel decomposition unit is used to perform sparse constrained mixed pixel decomposition of a hyperspectral image of a photovoltaic module area with a surface contamination layer, where some end members are known, to obtain an end member spectrum of the surface contamination layer and a corresponding abundance matrix;

[0041] The surface dirt layer state detection unit is used to determine the composition and occlusion ratio of the surface dirt layer.

[0042] The present invention also provides a device for detecting a contamination layer on the surface of a photovoltaic module, comprising:

[0043] Memory, used to store computer programs and data;

[0044] The processor is used to implement the steps of the above-mentioned method for detecting the surface contamination layer of photovoltaic modules when executing the computer program.

[0045] The beneficial effects of the present invention are:

[0046] 1. The present invention utilizes the endmember spectral information provided by the hyperspectral image and adopts sparse constrained mixed pixel decomposition with some endmembers known to obtain the endmember spectrum and corresponding abundance matrix of the surface contamination layer, so as to realize accurate detection of the surface contamination state of the photovoltaic module.

[0047] 2. The present invention can accurately detect the state of the surface contamination layer of photovoltaic modules, thereby helping to continuously monitor, track and warn of power generation efficiency losses and potential photovoltaic hot spot problems caused by changes in the surface contamination layer of modules during the operation and maintenance of photovoltaic power stations, providing a quantitative and precise basis for decision-making and optimization of cleaning strategies for photovoltaic modules, ensuring the safe operation of photovoltaic power stations and effectively saving cleaning resources, and having important application value for the safe operation and maintenance and energy saving and efficiency improvement of photovoltaic power stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A flow chart of a method for detecting a contamination layer on the surface of a photovoltaic module in this embodiment;

[0049] Figure 2A structural framework diagram of a detection system for a surface contamination layer of a photovoltaic module in this embodiment;

[0050] Figure 3 This is a structural framework diagram of a device for detecting contamination on the surface of a photovoltaic module in this embodiment.

[0051] Reference numerals: hyperspectral image acquisition unit 1 , preprocessing unit 2 , end member extraction unit 3 , mixed pixel decomposition unit 4 , surface pollution layer state detection unit 5 , memory 6 , processor 7 . DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Embodiment: A method for detecting contamination on the surface of a photovoltaic module, such as Figure 1 As shown, the method includes:

[0054] Step 1, selecting a single photovoltaic module for cleaning and obtaining a cleaned photovoltaic module sample;

[0055] Step 2, using a drone equipped with a hyperspectral imager to obtain hyperspectral images of all photovoltaic modules to be inspected;

[0056] Step 3, preprocessing the acquired hyperspectral image, including reflectance correction and outlier removal;

[0057] The reflectivity correction calculation method is:

[0058] ,

[0059] In the formula, For the wavelength The reflectivity at For the wavelength The radiance at (available from the sensor), is the dark current response obtained by measuring the black calibration plate, is the radiant brightness measured for a white reference plate, is the known reflectivity value of the white reference plate, usually 1 or a known high reflectivity value;

[0060] The method for removing outliers is:

[0061] ,

[0062] In the formula, For the wavelength The reflectivity at

[0063] Step 4, extracting the endmember spectrum of the photovoltaic module from the hyperspectral image of the clean photovoltaic module sample;

[0064] Among them, the endmember spectrum extraction method adopts one of the pure pixel index, single body growth, and vertex component analysis. The estimated endmember spectrum of the photovoltaic module is ;

[0065] Step 5, performing sparse constrained mixed pixel decomposition with some known endmembers on the hyperspectral image of the photovoltaic module area with the surface contamination layer to obtain the endmember spectrum of the surface contamination layer and the corresponding abundance matrix;

[0066] Specifically, the steps include:

[0067] Step 5.1: Extract the initial end member value of the photovoltaic module with surface contamination layer by end member extraction method ;

[0068] Step 5.2, Each end member is calculated with The similarity of each end member in The end member with the highest similarity is replaced by After rearrangement, the end member matrix of the photovoltaic module with surface contamination is constructed. ,in It remains fixed during the subsequent solution process. It represents the end-member spectrum of the surface pollution layer and needs to be solved and updated;

[0069] Step 5.3, estimate the PV panel abundance matrix by fully constrained least squares method and surface pollution layer abundance matrix The initial value of

[0070] Step 5.4, repeat the iteration and update by the following calculation formula , and ,

[0071] ,

[0072] ,

[0073] ,

[0074] In the formula, is the sparse regularization constraint parameter;

[0075] Step 6, determining the composition and shading ratio of the surface dirt layer;

[0076] Specifically, the steps include:

[0077] Step 6.1: Surface pollution endmember spectroscopy Matching the spectrum with the spectrum library to determine the component type of the surface pollution layer; the spectrum matching uses one or more of the spectral angle distance, correlation coefficient, Euclidean distance, and similarity coefficient as a similarity measure;

[0078] Step 6.2, based on the solved surface pollution layer abundance matrix Calculate the occlusion ratio of the surface dirt layer. The calculation formula is as follows:

[0079] ,

[0080] ,

[0081] In the formula, For the The occlusion ratio of the surface dirt layer, is the total occlusion ratio of the surface dirt layer.

[0082] In the technical solution provided in the embodiment of the present application, the endmember spectral information provided by the hyperspectral image is used, and the sparse constrained mixed pixel decomposition with some known endmembers is adopted to obtain the endmember spectrum and the corresponding abundance matrix of the surface dirt layer, so as to realize the accurate detection of the state of the surface dirt layer of the photovoltaic module. By realizing the accurate detection of the state of the surface dirt layer of the photovoltaic module, it is helpful to continuously monitor, track and warn the power generation efficiency loss and potential photovoltaic hot spot problems caused by the change of the surface dirt layer of the module during the operation and maintenance of the photovoltaic power station, and provide quantitative and precise basis for the decision-making and optimization of the cleaning strategy of the photovoltaic module, so as to ensure the safe operation of the photovoltaic power station and effectively save cleaning resources, which has important application value for the safe operation and maintenance and energy saving and efficiency improvement of the photovoltaic power station.

[0083] See also Figure 2 , is a schematic diagram of a structure of a photovoltaic module surface contamination detection system provided in an embodiment of the present application. The system comprises a hyperspectral image acquisition unit 1, a preprocessing unit 2, a mixed pixel decomposition unit 4, and a surface contamination state detection unit 5.

[0084] The hyperspectral image acquisition unit 1 is used to photograph the photovoltaic assembly to be inspected to acquire a hyperspectral image;

[0085] Preprocessing unit 2, used for preprocessing the hyperspectral image by correcting the reflectivity and removing abnormal values;

[0086] An end member extraction unit 3 is used to extract the end member spectrum of the photovoltaic module and the initial value of the end member spectrum of the photovoltaic module with a surface contamination layer;

[0087] A mixed pixel decomposition unit 4 is used to perform sparse constrained mixed pixel decomposition of the hyperspectral image of the photovoltaic module area with the surface pollution layer, in which some end members are known, to obtain the end member spectrum of the surface pollution layer and the corresponding abundance matrix;

[0088] The surface dirt layer state detection unit 5 is used to determine the composition and occlusion ratio of the surface dirt layer.

[0089] In the technical solution provided in the embodiment of the present application, first, a single photovoltaic module is selected for cleaning to obtain a cleaned photovoltaic module sample; then, a hyperspectral imager is mounted on an unmanned aerial vehicle to obtain hyperspectral images of all photovoltaic modules to be inspected; then, the obtained hyperspectral images are preprocessed, including reflectance correction and outlier removal; then, the endmember spectra of the photovoltaic modules are extracted from the hyperspectral images of the clean photovoltaic module samples; then, the hyperspectral images of the photovoltaic module areas with surface dirt layers are subjected to sparse constrained mixed pixel decomposition with some endmembers known, to obtain the endmember spectra of the surface dirt layer and the corresponding abundance matrix; finally, the composition and occlusion ratio of the surface dirt layer are determined.

[0090] See also Figure 3 , is a schematic diagram of a structure of a device for detecting a surface dirt layer of a photovoltaic module provided in an embodiment of the present application. The device includes a memory 6 and a processor 7 connected to the memory 6.

[0091] The memory 6 is used to store computer programs and data; the processor 7 is used to implement the steps of the above-mentioned method for detecting the surface contamination layer of photovoltaic modules when executing the computer program.

[0092] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for detecting a contamination layer on the surface of a photovoltaic module, characterized in that: The method comprises: Step 1, selecting a single photovoltaic module for cleaning and obtaining a cleaned photovoltaic module sample; Step 2, using a drone equipped with a hyperspectral imager to obtain hyperspectral images of all photovoltaic modules to be inspected; Step 3, preprocessing the acquired hyperspectral image, including reflectance correction and outlier removal; Step 4, extracting the endmember spectrum of the photovoltaic module from the hyperspectral image of the clean photovoltaic module sample; Step 5, performing sparse constrained mixed pixel decomposition with some known endmembers on the hyperspectral image of the photovoltaic module area with the surface contamination layer to obtain the endmember spectrum of the surface contamination layer and the corresponding abundance matrix; Step 6, determining the composition and shading ratio of the surface dirt layer; In step 5, the sparse constrained mixed pixel decomposition with some known endmembers includes the following steps: Step 5.1: Extract the initial end member value of the photovoltaic module with surface contamination layer by end member extraction method ; Step 5.2, Each end member is calculated with The similarity of each end member in The end member with the highest similarity is replaced by After rearrangement, the end member matrix of the photovoltaic module with surface contamination is constructed. ,in It remains fixed during the subsequent solution process. represents the end-member spectrum of the surface pollution layer, which needs to be solved and updated. is the estimated endmember spectrum of the PV module; Step 5.3, estimate the PV panel abundance matrix by fully constrained least squares method and surface pollution layer abundance matrix The initial value of Step 5.4, repeat the iteration and update by the following calculation formula , and , ， ， ， In the formula, is the sparse regularization constraint parameter; In step 6, the composition and occlusion ratio of the surface dirt layer are determined, including the following steps: Step 6.1: Surface pollution end-member spectroscopy Matching the spectrum with the spectrum library to determine the component type of the surface pollution layer; the spectrum matching uses one or more of the spectral angle distance, correlation coefficient, Euclidean distance, and similarity coefficient as a similarity measure; Step 6.2, based on the solved surface pollution layer abundance matrix Calculate the occlusion ratio of the surface dirt layer. The calculation formula is as follows: ， ， In the formula, For the The occlusion ratio of the surface dirt layer, is the total occlusion ratio of the surface dirt layer.

2. A method for detecting a contamination layer on the surface of a photovoltaic module according to claim 1, characterized in that: In step 3, the reflectivity correction calculation method is: ， In the formula, For the wavelength The reflectivity at For the wavelength The radiant brightness at is the dark current response obtained by measuring the black calibration plate, is the radiant brightness measured for a white reference plate, is the known reflectivity value of the white reference plate.

3. A method for detecting a contamination layer on the surface of a photovoltaic module according to claim 1, characterized in that: In step 3, the outlier removal method is: ， In the formula, For the wavelength The reflectivity at .

4. A method for detecting a contamination layer on the surface of a photovoltaic module according to claim 1, characterized in that: In step 4, the endmember spectrum extraction method adopts one of the clean pixel index, simplex growth, and vertex component analysis.

5. A system for detecting contamination on the surface of a photovoltaic module for implementing the method of claim 1, characterized in that: The system comprises: A hyperspectral image acquisition unit (1) is used to photograph the photovoltaic module to be inspected to acquire a hyperspectral image; A preprocessing unit (2) is used to perform preprocessing of the hyperspectral image by correcting the reflectivity and removing abnormal values; An end member extraction unit (3) is used to extract the end member spectrum of the photovoltaic module and the initial value of the end member spectrum of the photovoltaic module with a surface contamination layer; A mixed pixel decomposition unit (4) is used to perform sparse constrained mixed pixel decomposition with some known end members on the hyperspectral image of the photovoltaic module area with the surface contamination layer, so as to obtain the end member spectrum of the surface contamination layer and the corresponding abundance matrix; The surface dirt layer state detection unit (5) is used to determine the composition and occlusion ratio of the surface dirt layer.

6. A device for detecting contamination on the surface of a photovoltaic module, characterized in that: include: Memory (6) for storing computer programs and data; A processor (7) is used to implement the steps of the method for detecting contamination on the surface of a photovoltaic module as claimed in any one of claims 1 to 4 when executing a computer program.

Images