Photovoltaic cleaning gain quantification method, system, and apparatus

Abstract

The application discloses a photovoltaic cleaning gain quantification method, system and device. The method comprises the following steps: obtaining power generation data of at least one complete natural year before installation of a power station cleaning device; determining an evaluation period of 12 consecutive months; calculating a reference power generation by adopting an annual total amount comparison or a monthly comparison and then accumulating; collecting actual power generation of the same period after cleaning; and calculating a difference between the two as a cleaning increment. The application only uses physical measurement data of an inverter or an electric meter, does not require additional sensors, can objectively and repeatedly quantify the cleaning effect, and solves the problems of the prior art, such as experience judgment, high cost and complex model. The application also provides dependent schemes, such as decay correction, multi-year data weighting and non-natural year period processing, and corresponding systems, storage media and electronic devices. The application is suitable for quantification and evaluation of cleaning effects of various photovoltaic power stations.

Description

Technical Field

[0001] This invention relates to the field of intelligent photovoltaic cleaning operations, and more specifically to a photovoltaic cleaning gain quantification method, system, and equipment. Background Technology

[0002] After long-term operation, photovoltaic power plants accumulate dust, bird droppings, and other contaminants on the surface of their modules, leading to a decrease in power generation efficiency. To restore power generation capacity, the modules need to be cleaned regularly. Evaluating the cleaning effect—that is, quantifying the change in power generation before and after cleaning—is a fundamental technical task in the operation and maintenance of photovoltaic power plants.

[0003] Currently, the industry primarily uses the following technical methods to evaluate cleaning effectiveness: 1. Subjective experience judgment method Maintenance personnel qualitatively judge the cleaning effect by visually observing the cleanliness of the component surfaces or roughly comparing the instantaneous output power of the power station before and after cleaning. This method relies on personal experience and lacks objectivity, repeatability, and accuracy. Different personnel may reach completely different conclusions, making it unsuitable as a technical basis for revenue accounting or contract performance.

[0004] 2. Detection method using additional sensors Some technical solutions involve installing specialized sensors such as irradiance meters, reference solar cells, and dust detectors in the power plant to monitor the level of dust accumulation on the modules or their theoretical power generation capacity in real time, thereby calculating the increase in power generation resulting from cleaning. While this method offers high accuracy, it requires the additional purchase and installation of hardware, significantly increasing the operation and maintenance costs of the power plant. This is particularly problematic for the large number of existing distributed photovoltaic power plants, making it economically unviable and difficult to promote.

[0005] 3. Estimation method based on complex physical models Some solutions attempt to establish physical models that include multiple parameters such as irradiance, temperature, and component degradation. These models predict theoretical power generation by inputting meteorological data and then compare the predictions with actual power generation to obtain the cleaning gain. However, these methods rely on accurate meteorological data (which usually needs to be obtained from third parties or measured in-house). The model parameter calibration is complex, and the models have poor universality across different power plants and seasons, making them difficult to implement and inconvenient for daily operation and maintenance.

[0006] In summary, existing technologies lack a method to objectively and repeatably quantify the power generation gain before and after cleaning using only existing power generation data from the power plant. Therefore, the technical problem this invention aims to solve is: how to objectively quantify the power generation gain brought by cleaning equipment or cleaning services using only existing historical and actual power generation data from the power plant, through a simple, transparent, and verifiable calculation method, thereby providing technical support for evaluating the operation and maintenance effectiveness of the power plant. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a photovoltaic cleaning gain quantification method, system and equipment.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for quantifying the gain of photovoltaic cleaning includes the following steps: Obtain power generation data of the photovoltaic power plant for at least one full calendar year before the installation of the cleaning equipment, wherein the power generation data is derived from physical measurement data of the inverter or grid meter. Establish a continuous 12-month evaluation cycle; The baseline power generation is calculated using one of the following two methods: Method A: Add up the historical power generation corresponding to each month within the assessment period to obtain the total historical power generation for the assessment period, which is used as the benchmark power generation. Method B: Determine the historical power generation for each month within the assessment period as the benchmark power generation for each month; The actual power generation during the evaluation period after the cleaning service is collected, and the time granularity of the actual power generation is consistent with the granularity of the benchmark power generation in the selected method. The difference between the actual power generation and the benchmark power generation is calculated, and the difference is used as the power generation increment generated by the cleaning process.

[0009] When calculating the baseline power generation, the annual degradation effect of photovoltaic modules is further considered, and the historical power generation for the same period is multiplied by the degradation discount factor. ,in The annual decay rate, This refers to the number of years between the historical year and the target assessment period.

[0010] When using historical data from the most recent full year as a benchmark For method A, take Where t is the year number (t=1,2,…) starting from the installation of the cleaning equipment, and the annual base power generation is calculated according to the following formula: = in The total annual power generation for the most recent full year; For method B, take ,in The year to which the target month belongs. For the most recent full year, the baseline power generation for each month is calculated using the following formula: ( , ) = × in For the most recent full year Monthly power generation.

[0011] When you have historical data for multiple complete years For method A, take The annual base power generation is calculated using the following formula: = or = , For method B, take The baseline power generation for each month is calculated using the following formula: ( , ) = or ( , ) = , in The number of complete historical years. For each historical year, For the first Total annual power generation For the first Year Monthly power generation These are the weighting coefficients. For the target year.

[0012] The weighting coefficient The following condition must be met: the closer a historical year is to the target year, the greater its weight. < < < .

[0013] When the evaluation period spans two calendar years For months belonging to the first calendar year within the assessment period, take... ; For months within the assessment period that fall in the second calendar year, take... , in The starting year of the evaluation period. The year of the most recent full calendar year before the cleaning equipment was installed.

[0014] The annual decay rate δ is determined by any of the following methods: We take the industry experience value δ=0.3%; Alternatively, the linear degradation rate given in the photovoltaic module product manual may be used; Alternatively, fit an exponential curve to the power generation data for the same month of at least three complete historical years or to the power generation data for a complete year. δ is solved using the least squares method.

[0015] The method is executed by the local controller of the photovoltaic power plant or a cloud server.

[0016] A photovoltaic cleaning gain quantization system, comprising: The data acquisition module is used to acquire the power generation data of the photovoltaic power station for at least one full natural year before the installation of the cleaning equipment, as well as the actual power generation data within the evaluation period after the installation of the cleaning equipment. The data comes from the physical measurement data of the inverter or the grid meter. The evaluation cycle determination module is used to determine an evaluation cycle of 12 consecutive months; The baseline calculation module is used to calculate the baseline power generation using one of the following two methods: Method A: Add up the historical power generation corresponding to each month within the assessment period to obtain the total historical power generation for the assessment period, which is used as the benchmark power generation. Method B: Determine the historical power generation for each month within the assessment period as the benchmark power generation for each month; The incremental calculation module is used to calculate the difference between the actual power generation and the benchmark power generation, and use the difference as the power generation increment generated by the cleaning process.

[0017] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0018] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described above.

[0019] Compared with the prior art, the advantages of the present invention are as follows: This invention utilizes only the physical measurement data of power generation output from the power plant's existing inverters or grid meters. By directly comparing historical data with the cleaned data, the power generation gain brought about by the cleaning is obtained. Compared to existing technologies that require additional deployment of dedicated sensors such as irradiance meters, reference solar cells, and dust detectors, or rely on third-party meteorological data services, this invention achieves objective quantification of the cleaning effect while significantly reducing hardware costs and implementation difficulty. It has a particularly significant advantage in cost-sensitive scenarios such as distributed photovoltaic power plants.

[0020] The core step of this invention is to subtract two sets of physical measurement data (historical power generation and actual power generation after cleaning) to obtain the difference as the cleaning gain. This process does not rely on the operator's experience or involve the parameter calibration of complex physical models, thus eliminating the technical defects of subjective experience methods, such as variability and non-reproducibility of results. The obtained power generation increment value can directly reflect the degree of restoration of the power plant's power generation capacity by the cleaning operation, and it is traceable—any third party can reproduce the result based on the same data source, meeting the basic requirement of verifiability of the technical solution.

[0021] This invention provides two methods for calculating baseline power generation: Method A (Annual Total Comparison): The historical power generation for each month within the assessment period is summed to obtain the historical total power generation for the assessment period as a benchmark. This method is suitable for scenarios requiring an overall annual gain and where the assessment period aligns with the calendar year, and it is the simplest to calculate.

[0022] Method B (Monthly Comparison and Accumulation): The historical power generation for each month within the assessment period is determined as the benchmark for each month, and the data is compared month by month and then accumulated. This method is suitable for scenarios where the assessment period does not start in January (e.g., from March to March of the following year), and can effectively suppress the interference of seasonal weather fluctuations on a single comparison, obtaining more stable technical results.

[0023] Both methods set a continuous 12-month evaluation period and constrained the time granularity of the actual power generation after cleaning to be consistent with the selected benchmark granularity, thus ensuring the technical standardization of data comparison.

[0024] This invention limits the power generation data to be sourced from inverters or grid meters. These devices are legally certified metrological equipment in the power industry, and their output data has a clear physical meaning (kilowatt-hours), a controllable error range (typically less than 2%), and complete time-series records. Using such data sources ensures that the cleaning gain calculation results are not tampered with, and allows for third-party verification of the original metering records in case of disputes, thus solving the technical problems of non-standard data sources and difficulty in verifying results in existing technologies.

[0025] The cleaning gain output by this invention is measured in kilowatt-hours (kWh), a physical quantity that directly characterizes the actual improvement effect of the cleaning operation on the power generation of the power plant. Maintenance personnel can perform various technical analyses based on this indicator, such as comparing the gain values ​​of different seasons or different cleaning equipment to optimize cleaning cycle decisions; and jointly analyzing the gain value with other maintenance parameters such as component degradation rate and system efficiency to assess the overall health of the power plant. These applications demonstrate the specific technical value of this solution in the field of photovoltaic power generation technology. Detailed Implementation

[0026] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0027] Data preparation: Taking a 1MW photovoltaic power station as an example, the cleaning equipment was installed on March 1, 2025. The evaluation period is 12 consecutive months: from March 1, 2025 to February 28, 2026.

[0028] (1) Historical power generation data: The most recent complete calendar year before the installation of the cleaning equipment was 2024. Actual power generation in each month of 2024 (kWh) is shown in Table 1: Total electricity generation in 2024: =90,000+82,000+105,000+98,000+102,000+95,000+108,000+110,000+96,000+89,000+78,000+88,000=1,141,000kWh.

[0029] (2) The annual decay rate of the component is δ=0.3%=0.003 (industry experience value).

[0030] (3) Actual power generation in each month during the evaluation period after the installation of the cleaning equipment (kWh): Actual total power generation during the assessment period: =108,000+101,500+105,000+98,500+111,000+113,500+99,000+92,000+81,000+91,000+93,500+85,000=1,179,000kWh.

[0031] (4) Other parameters: electricity price p = 0.62 yuan / kWh, revenue sharing ratio s = 70%.

[0032] Based on the above measured data, the gain quantization method of the present invention will be described in detail below.

[0033] Example 1: Based on Method A, Annual Total Comparison Method.

[0034] Step 1: Determine the baseline power generation (1) Basic scheme The historical power generation for each month within the assessment period is summed up to obtain the total historical power generation for the assessment period, which is used as the benchmark power generation.

[0035] The corresponding 2024 power generation for each month of the assessment period (March 2025 to February 2026) is retrieved from the data in Part 1: Summing up, we get: =105,000+98,000+102,000+95,000+108,000+110,000+96,000+89,000+78,000+88,000+90,000+82,000=1,141,000kWh.

[0036] (2) Single-year benchmark scheme considering attenuation correction The formula given in method A is: = in =1,141,000 kWh, δ=0.003, =t=1 (first year of the contract).

[0037] Substitute: =1,141,000×(1 0.003) 1 =1,141,000×0.997=1,137,577kWh.

[0038] (3) Multi-year weighted average scheme considering attenuation correction Method A provides two formulas (arithmetic mean or weighted average). Here, we take the weighted average as an example: = , ,Pick Take N=3, historical years =2022, =2023, =2024.

[0039] Total annual power generation (supplemented from historical data): =1,100,000 kWh =1,120,000 kWh =1,141,000 kWh.

[0040] The weights are determined by proximity to the target year, with higher weights for closer years. =0.2, =0.3, =0.5. Target year =2025.

[0041] Substitute: 2024 item: 0.5×1,141,000×(1 0.003)(2025 2024)=0.5×1,141,000×0.997=568,788.5kWh 2023 items: 0.3×1,120,000×(1 0.003)(2025 2023)=0.3×1,120,000×0.994009=334,987.0kWh 2022 items: 0.2×1,100,000×(1 0.003)(2025 2022)=0.2×1,100,000×0.991027=218,025.9kWh Summing yields: =568,788.5+334,987.0+218,025.9=1,121,801.4kWh.

[0042] Step 2: Collect actual power generation data From the data in the first part, the actual total power generation =1,179,000 kWh.

[0043] Step 3: Calculate the incremental power generation (1) Basic scheme: Δ= =1,179,000 1,141,000 = 38,000 kWh.

[0044] (2) Annual attenuation correction: Δ= =1,179,000 1,137,577 = 41,423 kWh.

[0045] (3) Multi-year weighted average attenuation correction: Δ= =1,179,000 1,121,801.4 = 57,198.6 kWh.

[0046] Step 4: Revenue Sharing Calculation Total excess profit = Δ × 0.62 = Δ × 0.62 yuan / kWh Cleaning company's revenue = excess revenue × 70%; Power plant company's revenue = excess revenue × 30%; For example, taking the single-year decay correction result: excess return = 41,423 × 0.62 = 25,682 yuan. The cleaning company's revenue is 25,682 × 0.7 = 17,977 yuan, and the power plant's revenue is 25,682 × 0.3 = 7,705 yuan.

[0047] Method A (Annual Total Comparison): The historical power generation for each month within the assessment period is summed to obtain the historical total power generation for the assessment period as a benchmark. This method is suitable for scenarios requiring an overall annual gain and where the assessment period aligns with the calendar year, and it is the simplest to calculate.

[0048] Example 2: Based on Method B, monthly comparison and accumulation method.

[0049] Step 1: Determine the baseline power generation on a monthly basis (1) Basic scheme The historical power generation for each month within the assessment period is determined as the benchmark power generation for each month.

[0050] From the data in the first part, the values ​​for the same month in 2024 are used as the baseline for each month: (2) Single-year benchmark scheme considering attenuation correction The formula for method B is: ( , ) = × ,Pick ,in The year to which the target month belongs. The most recent full year. =2024, δ=0.003δ=0.003.

[0051] for =Months in 2025: =1, multiplier (1 0.003) 1 =0.997 for =Months in 2026: =2, multiplier (1 0.003) 2 =0.994009 The monthly calculation basis is as follows: (3) Handling of cross-year cycles When the evaluation period spans two calendar years (with the starting month not being January), for the month belonging to the first calendar year, take... For months within the assessment period that fall in the second calendar year, take... .

[0052] The evaluation period for this example begins in March 2025, that is... =2025, =2024.

[0053] Months in 2025: =2025 2024=1 Months in 2026: =2026 2024=2 The index is consistent with the single-year benchmark scheme in the attenuation correction in (2) above, and the subsequent calculation process will not be repeated here.

[0054] Step 2: Calculate and accumulate the increments month by month. Calculate the difference for each month ( , )= Then, sum them up to get the total increment.

[0055] Taking the baseline in the table above, which considers attenuation correction, as an example, we can directly accumulate the values: =(108,000 104,685)+(101,500 97,706)+(105,000 101,694)+(98,500 94,715)+(111,000 107,676)+(113,500 109,670)+(99,000 95,712)+(92,000 88,733)+(81,000 77,766)+(91,000 87,736)+(93,500 89,461)+(85,000 81,509)=3,315+3,794+3,306+3,785+3,324+3,830+3,288+3,267+3,234+3,264+4,039+3,491=41,423kWh The result is consistent with the total subtraction result of the annual attenuation correction in Example 1, verifying the equivalence of the two methods.

[0056] Method B (Monthly Comparison and Accumulation): Historical power generation for each month within the assessment period is determined as the benchmark for each month, and the data is compared month by month and then accumulated. This method is particularly suitable for scenarios where the assessment period begins outside of January (e.g., March to the following March), effectively suppressing the interference of seasonal weather fluctuations on a single comparison and obtaining more stable technical results. Furthermore, this method is also applicable to settlements based on any number of months, providing a more flexible quantitative calculation method.

[0057] This invention utilizes only the physical measurement data of power generation output from the power plant's existing inverters or grid meters. By directly comparing historical data with the cleaned data, the power generation gain brought about by the cleaning is obtained. Compared to existing technologies that require additional deployment of dedicated sensors such as irradiance meters, reference solar cells, and dust detectors, or rely on third-party meteorological data services, this invention achieves objective quantification of the cleaning effect while significantly reducing hardware costs and implementation difficulty. It has a particularly significant advantage in cost-sensitive scenarios such as distributed photovoltaic power plants.

[0058] The core step of this invention is to subtract two sets of physical measurement data (historical power generation and actual power generation after cleaning) to obtain the difference as the cleaning gain. This process does not rely on the operator's experience or involve the parameter calibration of complex physical models, thus eliminating the technical defects of subjective experience methods, such as variability and non-reproducibility of results. The obtained power generation increment value can directly reflect the degree of restoration of the power plant's power generation capacity by the cleaning operation, and it is traceable—any third party can reproduce the result based on the same data source, meeting the basic requirement of verifiability of the technical solution.

[0059] Example 3: Fitting and Determination of Annual Decay Rate δ 1. Data Selection The total annual power generation data for the power station over the past five full calendar years (2020-2024) was used to eliminate seasonal fluctuations. The total annual power generation data is as follows: 2. Fitting Model Logarithmic linearization: .

[0060] 3. Use the base year =2022 Calculate X for each year. 2022 and : 4. Solving for the slope using the least squares method Because X is symmetric (-2, -1, 0, 1, 2). .

[0061] calculate : =( 2×14.0000)+( (1×13.9935)+(0×13.9910)+(1×13.9885)+(2×13.9860) = 28.0000 13.9935 + 0 + 13.9885 + 27.9720 = 0.0330 =4+1+0+1+4=10 = 0.00330 5. By slope Solve Therefore = = ≈0.99670, = 1 0.99670 = 0.00330 = 0.330%, which is close to the industry experience value of 0.3%.

[0062] Example 4: Example of the aforementioned method being executed by the local controller of the photovoltaic power station. A local controller (such as an industrial computer, data acquisition unit, or smart gateway) is deployed at the power plant site. This controller connects to the inverter or meter via RS485, Ethernet, or wirelessly, periodically reading and storing power generation data locally. The controller internally contains the program for the aforementioned cleaning gain quantification method, including: a historical data storage module, an evaluation cycle setting module, a baseline calculation module (supporting methods A and B), and an incremental calculation module. At the end of the evaluation cycle, the controller automatically performs calculations and displays the results (power generation increment, excess revenue, and profit sharing) on ​​a local touchscreen, or uploads them to the owner's and cleaning service provider's mobile apps via 4G / 5G.

[0063] Example 5: Example of the aforementioned method being executed by the cloud server of a photovoltaic power station. If the power plant is already connected to a cloud platform (such as a photovoltaic monitoring system), the above calculations can also be performed on a cloud server. The power plant's inverters or data acquisition devices upload power generation data to the cloud server in real time (e.g., every 15 minutes). The server database stores historical annual data. When the evaluation period ends, a scheduled task on the server (or a manually triggered API) calls the same algorithm module, reads the historical and cleaned data from the database, performs calculations, stores the results in the database, and generates a report for users to download.

[0064] The cloud-based execution advantages of this embodiment are: it can simultaneously process incremental cleaning calculations for hundreds or thousands of power plants and generate reports in batches; when the algorithm is updated, only the server-side software needs to be upgraded, without the need for on-site modifications; and it supports multiple users (owners and cleaning service providers) to remotely view the calculation results, improving transparency.

[0065] Example 6: Embodiment of the Cleaning Gain Quantization System of the Present Invention The system includes the following modules: 1. Data Acquisition Module The module connects to the power station's inverter via an RS485 bus, periodically reading primary power generation data (kWh) and simultaneously acquiring grid-connected power generation data from the grid meter. This module has collected monthly power generation data for the entire calendar year from January 1, 2024 to December 31, 2024. And continuously collect real-time power generation data after the cleaning equipment is installed (starting from March 2025). ( , ).

[0066] All data is timestamped and stored in a local SQLite database.

[0067] 2. Evaluation Cycle Determination Module The system accepts the start date (March 1, 2025) and end date (February 28, 2026) of the evaluation period input by the user, automatically verifies whether the period is 12 consecutive months, and determines the start year. =2025, starting month =3, End Year =2026, Ending Month =2.

[0068] This module determines whether the evaluation period spans a calendar year based on the starting month and passes this information to the benchmark calculation module.

[0069] 3. Benchmark Calculation Module Users select either method A or method B through the system interface. In this example, method A is selected. This module reads data from the database based on the historical corresponding months for each month within the evaluation period (March 2024 to February 2026 corresponds to March 2024 to February 2025). And sum them up to get the baseline power generation. =1,141,000 kWh. Option B can also be selected in other embodiments, which will not be elaborated here.

[0070] 4. Incremental Calculation Module This module extracts the actual values ​​for each month within the evaluation period from the actual power generation database. Calculate the difference ( , )= .

[0071] For method A, the total difference is calculated directly. The final output includes the total increment, excess revenue, and profit sharing amount, which is then sent to the operations and maintenance personnel's mobile app via Ethernet.

[0072] Each module of the above system can be implemented by an industrial controller (such as a Raspberry Pi 4B+) or a cloud server installed at the power plant site, and the modules can interact with each other through function calls or REST APIs.

[0073] Example 7: Example of a computer-readable storage medium In this embodiment, a USB flash drive is used as the computer-readable storage medium, on which a computer program is stored. When the program is executed by the processor, it implements the calculation method described in any of the aforementioned quantitative method embodiments. Specifically, the method of the present invention is written as a Python script or a C++ program and saved on the readable storage medium. The user inserts the storage medium into the power plant data acquisition device, which automatically installs and runs the program. The program reads the historical power generation CSV file and outputs incremental results according to the evaluation period and calculation method selected by the user.

[0074] Example 8: Electronic Device Example This embodiment uses a cloud server running a Linux system. Users submit the power plant ID and evaluation cycle parameters via a web frontend. The backend service reads historical data of the power plant stored in the cloud database, executes the method of this invention, and returns incremental results in JSON format. This electronic device (cloud server) can handle requests from tens of thousands of power plants simultaneously, possessing high concurrency capabilities. In other embodiments, industrial tablet PCs or other electronic devices can also be used as the carrier for the method and system of this invention.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for quantifying the gain of photovoltaic cleaning, characterized in that, Includes the following steps: Obtain power generation data of the photovoltaic power plant for at least one full calendar year before the installation of the cleaning equipment, wherein the power generation data is derived from physical measurement data of the inverter or grid meter. Establish a continuous 12-month evaluation cycle; The baseline power generation is calculated using one of the following two methods: Method A: Add up the historical power generation corresponding to each month within the assessment period to obtain the total historical power generation for the assessment period, which is used as the benchmark power generation. Method B: Determine the historical power generation for each month within the assessment period as the benchmark power generation for each month; The actual power generation during the evaluation period after the cleaning service is collected, and the time granularity of the actual power generation is consistent with the granularity of the benchmark power generation in the selected method. The difference between the actual power generation and the benchmark power generation is calculated, and the difference is used as the power generation increment generated by the cleaning process.

2. The photovoltaic cleaning gain quantification method according to claim 1, characterized in that, When calculating the baseline power generation, the annual degradation effect of photovoltaic modules is further considered, and the historical power generation for the same period is multiplied by the degradation discount factor. ,in The annual decay rate, This refers to the number of years between the historical year and the target assessment period.

3. The photovoltaic cleaning gain quantification method according to claim 2, characterized in that, When using historical data from the most recent full year as a benchmark For method A, take Where t is the year number counting from the installation of the cleaning equipment (t=1,2,…), and the annual base power generation is calculated according to the following formula: = in The total annual power generation for the most recent full year; For method B, take ,in The year to which the target month belongs. For the most recent full year, the baseline power generation for each month is calculated using the following formula: ( , ) = × in For the most recent full year Monthly power generation.

4. The photovoltaic cleaning gain quantification method according to claim 2, characterized in that, When you have historical data for multiple complete years For method A, take The annual base power generation is calculated using the following formula: = or = ， For method B, take The baseline power generation for each month is calculated using the following formula: ( , ) = or ( , ) = ， in The number of complete historical years. For each historical year, For the first Total annual power generation For the first Year Monthly power generation These are the weighting coefficients. For the target year.

5. The photovoltaic cleaning gain quantification method according to claim 4, characterized in that, The weighting coefficient The following condition must be met: the closer a historical year is to the target year, the greater its weight. < < < .

6. The photovoltaic cleaning incremental quantification method according to any one of claims 2 to 5, characterized in that, When the evaluation period spans two calendar years For months belonging to the first calendar year within the assessment period, take... ; For months within the assessment period that fall in the second calendar year, take... , in The starting year of the evaluation period. The year of the most recent full calendar year before the cleaning equipment was installed.

7. The method according to any one of claims 2 to 5, characterized in that, The annual decay rate δ is determined by any of the following methods: We take the industry experience value δ=0.3%; Alternatively, the linear degradation rate given in the photovoltaic module product manual may be used; Alternatively, fit an exponential curve to the power generation data for the same month of at least three complete historical years or to the power generation data for a complete year. δ is solved using the least squares method.

8. The method according to claim 1, characterized in that, The method is executed by the local controller of the photovoltaic power plant or a cloud server.

9. A photovoltaic cleaning gain quantification system, characterized in that, include: The data acquisition module is used to acquire the power generation data of the photovoltaic power station for at least one full natural year before the installation of the cleaning equipment, as well as the actual power generation data within the evaluation period after the installation of the cleaning equipment. The data comes from the physical measurement data of the inverter or the grid meter. The evaluation cycle determination module is used to determine an evaluation cycle of 12 consecutive months; The baseline calculation module is used to calculate the baseline power generation using one of the following two methods: Method A: Add up the historical power generation corresponding to each month within the assessment period to obtain the total historical power generation for the assessment period, which is used as the benchmark power generation. Method B: Determine the historical power generation for each month within the assessment period as the benchmark power generation for each month; The incremental calculation module is used to calculate the difference between the actual power generation and the benchmark power generation, and use the difference as the power generation increment generated by the cleaning process.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

Images