Photovoltaic power station power generation efficiency evaluation method, device, medium and product

Abstract

The invention discloses a photovoltaic power station power generation efficiency evaluation method and device, a medium and a product, and relates to the technical field of new energy. The method comprises the steps of obtaining typical annual meteorological data of a position where a target photovoltaic power station is located; according to the power station characteristics and the typical annual meteorological data, generating simulation power generation time sequence data; determining an approximate time period and reference irradiation data in the typical year by combining the simulated power generation time sequence data and the first actual output power data of the first inverter in the target time period; determining first theoretical generating capacity data based on the reference irradiation data; determining direct current side system energy efficiency data according to the first actual generating capacity data and the first theoretical generating capacity data; and under the condition of determining the power generation energy efficiency reference data, determining a power generation efficiency evaluation result of the target photovoltaic power station in the target time period based on the direct current side system energy efficiency data and the power generation energy efficiency reference data. According to the invention, the power generation efficiency of the photovoltaic power station can be effectively evaluated in the absence of meteorological real-time monitoring data.

Description

Technical Field

[0001] The present invention relates to the field of new energy technologies, and in particular to a method, equipment, medium and product for evaluating the power generation efficiency of a photovoltaic power station. Background Art

[0002] With the rapid development of new energy technologies, photovoltaic power plants are being built and operated in large numbers. Evaluating the power generation efficiency of photovoltaic power plants facilitates real-time monitoring of the photovoltaic system's operating status, identifying efficiency deviations and potential failures, and optimizing operation and maintenance strategies to improve power generation efficiency and economic benefits.

[0003] In related technologies, to evaluate the power generation efficiency of a photovoltaic power station over a certain period of time, it is generally necessary to rely heavily on the irradiation data from meteorological monitoring during this period. The total daytime irradiation is calculated by integration, and the theoretical conversion efficiency of the photovoltaic modules and the system loss coefficient are combined to derive the theoretical power generation. This is then compared with the actual power generation to evaluate the power generation efficiency.

[0004] However, in actual applications, some photovoltaic projects are restricted by geographical conditions (such as remote areas, areas with restricted data transmission, etc.), making it difficult to obtain sufficiently reliable real-time meteorological monitoring data over a period of time. As a result, the traditional method of relying on irradiation data cannot effectively evaluate the power generation efficiency of the power station. Summary of the Invention

[0005] In view of the above technical problems and defects, the purpose of the present invention is to provide a photovoltaic power station power generation efficiency evaluation method, equipment, medium and product, which can effectively evaluate the power generation efficiency of photovoltaic power stations in the absence of real-time meteorological monitoring data.

[0006] To achieve the above-mentioned objectives, in a first aspect, the present invention provides a method for evaluating the power generation efficiency of a photovoltaic power station, comprising: obtaining typical annual meteorological data of a target photovoltaic power station; generating simulated power generation time series data of the target photovoltaic power station based on the power station characteristics of the target photovoltaic power station and the typical annual meteorological data; determining an approximate time period and reference irradiation data in a typical year based on the simulated power generation time series data and the first actual output power data of the first inverter of the target photovoltaic power station in the target time period, the reference irradiation data being the irradiation data of the approximate time period in the typical annual meteorological data; determining first theoretical power generation data of the first inverter in the target time period based on the reference irradiation data; determining DC side system energy efficiency data of the target photovoltaic power station based on the first actual power generation data and the first theoretical power generation data of the first inverter in the target time period; and determining a power generation efficiency evaluation result of the target photovoltaic power station in the target time period based on the DC side system energy efficiency data and the power generation energy efficiency benchmark data when the power generation energy efficiency benchmark data is determined.

[0007] When there is a lack of real-time meteorological monitoring data, the present invention can generate simulated power generation time series data with the help of the typical annual meteorological data of the target photovoltaic power station in combination with the characteristics of the power station, thereby getting rid of the dependence on real-time meteorological monitoring. By matching the simulated data with the actual output power of the inverter, the approximate time period of the typical year and the reference irradiation data are determined to calculate the theoretical power generation. The actual power generation is then compared with the theoretical power generation to determine the energy efficiency data of the DC side system, and finally the evaluation is completed in combination with the power generation energy efficiency benchmark data. The theoretical model is constructed using historical typical data to circumvent the obstacle of the lack of real-time meteorological data, ensuring that the power generation efficiency of the photovoltaic power station can still be accurately evaluated in scenarios where real-time data is insufficient, providing a reliable basis for power station operation optimization, fault diagnosis, etc., effectively solving the problem of difficult power generation efficiency evaluation when meteorological monitoring data is scarce, and improving the adaptability and practicality of the evaluation method.

[0008] Optionally, in some embodiments, before determining the power generation efficiency evaluation result of the target photovoltaic power station based on the DC side system energy efficiency data and the power generation energy efficiency benchmark data, it also includes: obtaining the second inverters of multiple photovoltaic power stations within a preset area, the second actual power generation data and the second theoretical power generation data within the target time period, the preset area is determined based on the location of the target photovoltaic power station, and the second theoretical power generation data is determined based on the irradiation data in the typical annual meteorological data; generating the power generation energy efficiency benchmark data based on the second actual power generation data and the second theoretical power generation data.

[0009] By adopting the technical solution of the above embodiment, power generation energy efficiency benchmark data is generated by obtaining data from multiple photovoltaic power stations in a preset area, making the benchmark data more regional and representative, providing an objective reference for the target photovoltaic power station, improving the accuracy and rationality of power generation efficiency evaluation, and avoiding the limitations of single power station data.

[0010] Optionally, in some embodiments, power generation energy efficiency benchmark data is generated based on the second actual power generation data and the second theoretical power generation data, including: normalizing the second actual power generation data to obtain actual unit power power generation data, and normalizing the second theoretical power generation data to obtain theoretical unit power power generation data; using the actual unit power power generation data as sample data, and verifying whether the sample data conforms to the normal distribution; if so, determining the distribution test method based on the sample size of the sample data; based on the distribution test method, determining the distribution critical value corresponding to the target confidence level; determining the confidence interval based on the distribution critical value, the sample size and the mean and standard deviation of the sample data; and determining the power generation energy efficiency benchmark data based on the confidence interval and the theoretical unit power power generation data.

[0011] By adopting the technical solution of the above embodiment, the process of generating power generation energy efficiency benchmark data is standardized in detail. After normalization, power differences are eliminated. Through normal distribution verification, sample size judgment and confidence interval calculation, the scientific rigor of the benchmark data is ensured, its reliability and persuasiveness are enhanced, and a solid foundation is laid for subsequent evaluation.

[0012] Optionally, in some embodiments, the power station characteristics include the inclination angle, azimuth angle and inverter efficiency of the photovoltaic modules, and the simulated power generation timing data includes a simulated power generation curve; based on the power station characteristics of the target photovoltaic power station and the typical annual meteorological data, the simulated power generation timing data of the target photovoltaic power station is generated, including: generating an inclined surface irradiation curve based on the inclination angle, azimuth angle and horizontal plane irradiance in the typical annual meteorological data; generating a theoretical inverter power curve based on the inverter efficiency and the inclined surface irradiation curve; through a dynamic time warping algorithm, aligning the time axis of the theoretical inverter power curve with the actual power curve of the target photovoltaic power station to obtain the simulated power generation curve of the target photovoltaic power station.

[0013] By adopting the technical solution of the above embodiment, the power station characteristics (inclination, azimuth, etc.) are combined with typical annual meteorological data to generate a simulated power generation curve. The actual power curve is aligned through dynamic time regularization to make the simulated data fit the actual operating status, thereby improving the accuracy of the simulated power generation timing data and facilitating more accurate performance analysis.

[0014] Optionally, in some embodiments, based on the simulated power generation timing data and the first actual output power data of the first inverter of the target photovoltaic power station in the target time period, an approximate time period in a typical year is determined, including: dividing the typical year into multiple time segments based on the target time period; generating irradiation curves for each time segment based on the irradiation data of the typical year; determining an approximate power curve from multiple irradiation curves based on the first actual output power data; and determining the time period corresponding to the approximate power curve as the approximate time period.

[0015] By adopting the technical solution of the above embodiment, the typical year is divided into time segments to determine the approximate time period, and the approximate power curve is screened based on the actual output power, so that the reference irradiation data is more closely matched with the target time period, ensuring the accuracy of the calculation of the first theoretical power generation data, and providing a reliable basis for energy efficiency evaluation.

[0016] Optionally, in some embodiments, the DC side system energy efficiency data of the target photovoltaic power station is determined based on the first actual power generation data and the first theoretical power generation data of the first inverter in the target time period, including: determining the first energy efficiency data based on the first actual power generation data and the first theoretical power generation data; determining the second energy efficiency data based on the first actual power generation data and the third theoretical power generation data of the first inverter, the third theoretical power generation data being generated based on real-time meteorological data of the location of the target photovoltaic power station within the target time period; and determining the DC side system energy efficiency data based on the first energy efficiency data and the second energy efficiency data.

[0017] By adopting the technical solution of the above embodiment, the energy efficiency of the DC side system is determined by combining the first energy efficiency data based on a typical year and the second energy efficiency data based on real-time meteorological conditions. This takes into account both historical benchmarks and real-time status, comprehensively reflects the power generation efficiency of the photovoltaic power station, avoids single data deviation, and improves the comprehensiveness and credibility of the evaluation results.

[0018] Optionally, in some embodiments, the DC side system energy efficiency data is determined based on the first energy efficiency data and the second energy efficiency data, including: determining the weight coefficient of the first energy efficiency data and the second energy efficiency data based on the completeness of the real-time meteorological data; and determining the DC side system energy efficiency data based on the weight coefficient, the first energy efficiency data and the second energy efficiency data.

[0019] By adopting the technical solution of the above embodiment, the DC side energy efficiency is calculated by allocating weights according to the completeness of real-time meteorological data, flexibly adapting to changes in data quality, enhancing adaptability to different scenarios, making the results more reasonable, ensuring objective evaluation when the data completeness is different, and improving the reliability of power generation efficiency evaluation.

[0020] In a second aspect, an embodiment of the present invention provides an electronic device, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the electronic device to execute the method described in the first aspect or the second aspect, and any possible implementation of the first aspect or the second aspect.

[0021] In a third aspect, the present invention provides a computer-readable storage medium comprising instructions, which, when executed on the electronic device, enables the electronic device to execute the method described in the first aspect or the second aspect, and any possible implementation of the first aspect or the second aspect.

[0022] In a fourth aspect, the present invention provides a computer program product comprising instructions, which, when the computer program product is run on the electronic device, enables the electronic device to execute the method described in the first aspect or the second aspect, and any possible implementation of the first aspect or the second aspect.

[0023] It is understood that the electronic device provided in the second aspect, the storage medium provided in the third aspect, and the computer program product provided in the fourth aspect are all used to execute the method provided by the present invention. Therefore, the beneficial effects that can be achieved can be referred to the beneficial effects of the corresponding methods and will not be repeated here.

[0024] One or more technical solutions provided by the present invention have at least the following technical effects or advantages: 1. Multi-source data integration improves scientific evaluation: By deeply integrating typical annual meteorological data with actual power plant power generation data, simulated power generation time series data is generated and matched to approximate time periods. This ensures that theoretical power generation calculations closely align with the actual power plant operating environment, avoiding the biases of relying solely on real-time monitoring data or historical models. Energy efficiency benchmark data is constructed by combining data from multiple power plants in the region. Through normalization, normal distribution testing, and confidence interval calculation, the objectivity and statistical representativeness of the benchmark data are ensured, providing a scientific reference for horizontal comparison with target power plants and enhancing the rigor of performance evaluation.

[0025] 2. Eliminate dependence on real-time meteorological data: When real-time meteorological data is missing or incomplete, a simulated power generation model is constructed using typical annual data. The theoretical and actual power curves are aligned through a dynamic time warping algorithm, and the historical approximate time period that matches the target time period is accurately located. This ensures that the theoretical power generation calculation does not rely on real-time environmental monitoring, effectively solving the evaluation problem in data-scarce scenarios, enhancing the universality and engineering practicality of the method, and is particularly suitable for areas with limited meteorological monitoring conditions.

[0026] 3. Dynamic weight allocation to adapt to complex scenarios: By incorporating multi-dimensional factors such as the completeness, temporal resolution, and outlier ratio of real-time meteorological data to determine weight coefficients, the contribution of typical year data and real-time data to energy efficiency calculations is nonlinearly adjusted, ensuring that evaluation results reflect both long-term benchmark performance and capture real-time operational anomalies. For example, when data is missing or abnormal, the weight of typical year data is automatically increased to ensure result stability; when real-time data quality is high, the emphasis is placed on immediate performance, enabling intelligent adaptation to different scenarios and significantly improving the reliability and flexibility of evaluation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present invention, and together with the specification, are used to explain the principles of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and those skilled in the art can derive other drawings based on these drawings without inventive effort. In the drawings: Figure 1 This is a flow chart of a method for evaluating power generation efficiency of a photovoltaic power station according to an embodiment of the present invention; Figure 2 is a flow chart of another photovoltaic power station power generation efficiency evaluation method according to an embodiment of the present invention; Figure 3 It is a schematic diagram of the architecture of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. As used in the specification of the present invention, the singular expressions "a," "an," "above," "the," and "this" are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used in the present invention refers to any and all possible combinations of one or more of the listed items.

[0029] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying relative importance or implicitly indicating the quantity of the technical features indicated. Thus, a feature designated "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, unless otherwise specified, "plurality" means two or more.

[0030] It should also be noted that, unless otherwise clearly specified and limited, in the embodiments of the present invention, terms such as "setting" and "connection" should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal connection of two components; it can be a wired communication connection or a wireless communication connection. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. The embodiments of the present invention are described in detail below.

[0031] In related technologies, when evaluating whether the power generation efficiency of a photovoltaic power station meets expectations, a common method is to use a weather station to monitor the irradiance data on the surface of the photovoltaic modules in real time, integrate the irradiance data to obtain the total irradiation, and then calculate the theoretical power generation by combining the conversion efficiency of the photovoltaic modules, inverter efficiency and other parameters. Specifically, first, the irradiance data is collected in real time by the weather station, and the irradiance is integrated in time series to obtain the total irradiation energy received by the photovoltaic modules on that day; then, based on the rated power, efficiency and system configuration parameters of the modules, the theoretical power generation under the irradiation conditions is calculated; finally, the theoretical power generation is compared with the actual power generation output of the photovoltaic power station, and the difference between the two is used to determine whether the power generation efficiency meets expectations.

[0032] However, this method is highly dependent on meteorological data. When it is difficult to obtain sufficient and reliable meteorological data, it is difficult to effectively evaluate the power generation efficiency of the power station.

[0033] Therefore, the embodiment of the present invention provides a photovoltaic power station power generation efficiency evaluation scheme, which can effectively evaluate the photovoltaic power station power generation efficiency in the absence of real-time meteorological monitoring data. First, by obtaining the typical annual meteorological data of the target photovoltaic power station and generating simulated power generation time series data, there is no need to rely on real-time meteorological monitoring equipment or satellite data, and the influence of geographical restrictions on meteorological data acquisition is eliminated. It is particularly suitable for remote areas or scenarios where meteorological data is scarce, and the universality and operability of the method are improved. Based on the simulated power generation time series data generated based on the power station characteristics and typical annual data, the power generation characteristics of the target photovoltaic power station under different meteorological conditions can be accurately characterized. The approximate time period and reference irradiation data in the typical year are determined through a matching algorithm, so that the calculation of theoretical power generation is closely aligned with the actual power generation scenario, the matching degree between theoretical data and actual operation data is improved, and the evaluation deviation caused by the lack of real-time data is avoided.

[0034] During the energy efficiency assessment phase, the theoretical power generation of the first inverter is calculated by referencing irradiation data. This is then compared with actual power generation to determine the DC-side system energy efficiency data, enabling a quantitative analysis of the PV system's energy conversion efficiency. Combined with power generation efficiency benchmark data, this objectively determines whether the target power station's power generation performance meets expectations within the target time period, providing a scientific, quantitative basis for operation and maintenance decisions.

[0035] The embodiments of the present invention make full use of available historical meteorological data and inverter operation data, reduce dependence on real-time meteorological monitoring conditions, save hardware deployment costs, and ensure the accuracy and reliability of evaluation results through statistical means and simulation technology. It effectively solves the problems of single data input and poor regional adaptability in related technologies, and provides an innovative technical path for efficient operation and maintenance and performance optimization of photovoltaic power stations. It has significant practical value and promotion significance, especially in scenarios where meteorological data is scarce.

[0036] The following combination Figure 1 , to specifically illustrate a photovoltaic power station power generation efficiency evaluation method provided by an embodiment of the present invention, comprising the following steps: Step 201: Acquire typical annual meteorological data at the location of the target photovoltaic power station.

[0037] Specifically, it's first necessary to ensure that the target power plant doesn't have any backflow prevention devices or power generation restrictions. If the target PV plant doesn't have meteorological monitoring equipment installed and complete irradiation data is unavailable, it's possible to filter out multiple years (typically more than 10 years) of meteorological data for the target PV plant's location from meteorological agencies, satellite remote sensing databases, or long-term local monitoring data. This data covers key parameters such as hourly or minute-by-minute solar irradiance (including direct and diffuse irradiance), ambient temperature, and humidity.

[0038] The data for each year are then preprocessed to remove outliers and fill in missing values. The average irradiance distribution for each time period is calculated through statistical analysis (such as a monthly irradiance curve similar to a Gaussian function), and finally a set of annual meteorological data representing the typical climate characteristics of the region is formed, namely the typical annual meteorological data, among which solar irradiance data (referred to as irradiation data) is the core parameter.

[0039] Step 202 : Generate simulated power generation time series data of the target photovoltaic power station based on the power station characteristics of the target photovoltaic power station and typical annual meteorological data.

[0040] Among them, power station characteristics include but are not limited to photovoltaic module type (such as monocrystalline silicon, polycrystalline silicon) and rated efficiency, module installation inclination and azimuth, inverter model and conversion efficiency, system DC and AC side configuration parameters, etc.

[0041] The solar irradiance data in the typical annual meteorological data is converted into the inclined surface irradiance actually received by the photovoltaic module through the inclined surface irradiance conversion model (such as the Perez model) combined with the module installation inclination angle and azimuth angle. At the same time, the influence of ambient temperature on module efficiency (such as temperature coefficient correction) is taken into account to convert the inclined surface irradiance into the theoretical input power of the module DC side.

[0042] Further combined with the inverter's conversion efficiency curve (taking into account efficiency changes under different load rates), the DC power is converted into the theoretical output power on the AC side, and simulated power generation data is generated in a time series (such as hourly).

[0043] Finally, through hour-by-hour calculation of typical annual meteorological data throughout the year, a simulated power generation time series curve covering each time period is generated. This curve reflects the theoretical power generation capacity of the target photovoltaic power station under typical meteorological conditions, providing benchmark data for subsequent approximate time period matching and energy efficiency evaluation.

[0044] In this embodiment, the simulated power generation time series data is obtained by integrating the hardware configuration parameters of the target photovoltaic power station (such as the type and rated efficiency of photovoltaic modules, installation inclination and azimuth, inverter model and conversion efficiency, etc.) with the typical annual meteorological data of the area (including hourly solar irradiance, ambient temperature, etc.), and generating a theoretical power generation data sequence at each time point through physical model calculation. Specifically, it can be formed into a continuous simulated power generation curve in chronological order (such as hourly or daily).

[0045] The simulated power generation time series data represents the theoretical power generation capacity of the target photovoltaic power station under typical meteorological conditions, and can accurately map the impact of environmental factors such as light and temperature in different time periods on power generation performance.

[0046] Step 203, based on the simulated power generation timing data and the first actual output power data of the first inverter of the target photovoltaic power station in the target time period, determine the approximate time period and reference irradiation data in a typical year, where the reference irradiation data is the irradiation data of the approximate time period in the typical year meteorological data.

[0047] The target time period may be a certain day, a certain hour, or a sunshine period of a certain day (such as sunrise to sunset).

[0048] Taking a specific day as an example for the target time period, the actual output power data of the first inverter of the target PV power station during that day's time period (such as a power series recorded by minute or hour) needs to be matched day by day with the simulated power generation power curve for each day of a typical year in the simulated power generation time series data. In specific implementation, the sunshine duration of the day is used as the time window, and the simulated power curve of the same duration for each day in the simulated power generation time series data (such as the hourly power data corresponding to the local sunrise time to sunset time) is extracted. By calculating the similarity index between the actual power curve and each simulated curve (such as the Pearson correlation coefficient, root mean square error (RMSE), or dynamic time warping distance (DTW), the date with the highest similarity is selected as the "approximate day" (i.e., the approximate time period).

[0049] The hourly irradiance data (including direct irradiance, diffuse irradiance and total irradiance) during the sunshine period of the approximate day corresponding to the typical annual meteorological data, that is, the reference irradiance data used in subsequent calculations, is essentially the irradiance distribution under typical meteorological conditions that is closest to the actual power generation fluctuation characteristics of the target day.

[0050] Step 204 : determining first theoretical power generation data of the first inverter within the target time period based on the reference irradiation data.

[0051] The first theoretical power generation data is the ideal power generation expected by the first inverter within the target time period.

[0052] Specifically, based on the example in the previous step, first integrate the hourly irradiance during the sunshine period of the day in the reference irradiance data to calculate the total irradiance received by the PV modules. The specific method is: integrate the irradiance (unit W / m 2 ) multiplied by the time interval (e.g. 1 hour is converted to 3600 seconds) to obtain the irradiation energy (Wh / m 2 ), and the total radiation exposure for the whole day is obtained by accumulating it in each time period.

[0053] Then, combined with the component parameters of the target photovoltaic power station (such as the total area of ​​photovoltaic components S, unit m 2 ; The rated conversion efficiency of the component η_mod, which needs to be corrected for factors such as temperature and attenuation, is used to calculate the theoretical input energy on the DC side of the component E_dc = total radiation × S × η_mod.

[0054] Further, according to the actual operating efficiency η_inv of the inverter (the typical efficiency or real-time efficiency curve in the equipment parameters), the DC side energy is converted into the AC side theoretical output energy, that is, the first theoretical power generation data E_ac=E_dc×η_inv.

[0055] This process needs to ensure that the time resolution of the reference irradiation data is consistent with the actual output power data (for example, both are 15-minute intervals), and achieve quantitative calculation from irradiation to theoretical power generation through physical model conversion, providing benchmark comparison data for evaluating the actual power generation efficiency of the inverter.

[0056] Step 205 : determining DC side system energy efficiency data of the target photovoltaic power station according to the first actual power generation data and the first theoretical power generation data of the first inverter in the target time period.

[0057] Specifically, it is necessary to quantitatively compare the first actual power generation data of the first inverter in the target time period (that is, the actual electric energy output on the DC side of the inverter. If it is AC side data, the DC side energy needs to be inferred through the inverter efficiency, and the formula is "actual DC power generation = actual AC power generation / inverter efficiency") with the above-mentioned first theoretical power generation data.

[0058] The DC side system energy efficiency data can be specifically referred to as the DC side system energy efficiency ratio. When specifically implemented, it is calculated using the formula “DC side system energy efficiency ratio = (actual DC power generation / theoretical DC power generation) × 100%”. The DC side system energy efficiency ratio reflects the proportional relationship between the actual efficiency of photovoltaic modules in converting solar energy into DC power and the theoretical efficiency.

[0059] If the actual power generation data is the AC side measurement value, it must first be corrected to the DC side energy using the inverter rated efficiency or real-time efficiency curve to ensure consistency with the physical dimensions of the theoretical power generation data. Ultimately, energy efficiency data is obtained that characterizes the energy conversion efficiency of the system's DC side.

[0060] Step 206 : When the power generation energy efficiency benchmark data is determined, the power generation efficiency evaluation result of the target photovoltaic power station within the target time period is determined based on the DC side system energy efficiency data and the power generation energy efficiency benchmark data.

[0061] Specifically, first, the source of the power generation energy efficiency benchmark data is clarified. The power generation energy efficiency benchmark data can be the statistical average of the DC side energy efficiency ratio of similar photovoltaic power stations in the same region, the average energy efficiency ratio of the target photovoltaic power station in normal operation during the same period in history, or the confidence interval of the target confidence level (such as 95% or 98%) calculated by statistical methods (such as t-test or z-test). In specific implementation, the calculated DC side system energy efficiency data is compared with the benchmark data: if the benchmark data is a confidence interval, it is determined whether the energy efficiency data falls within the interval. If it is within the interval, the power generation efficiency is normal. If it exceeds the interval, it is considered abnormal and the deviation direction (higher than the upper limit or lower than the lower limit) is combined to analyze whether there is a measurement error or equipment failure; if the benchmark data is a fixed threshold (such as the industry standard energy efficiency ratio), the percentage difference between the energy efficiency data and the threshold is calculated, and whether it meets expectations is determined based on the preset allowable deviation range (such as ±5% or ±2%).

[0062] Finally, through quantitative comparison, the power generation efficiency evaluation results of the target photovoltaic power station within the target time period are output, providing clear conclusions such as "in line with expectations", "low efficiency" or "abnormal efficiency" for operation and maintenance decisions, and can further locate the causes of efficiency loss (such as component degradation, abnormal inverter loss, etc.).

[0063] This embodiment, through the above-mentioned method steps, effectively solves the problem of traditional methods' over-reliance on real-time irradiation data by introducing typical annual meteorological data and simulated power generation time series data driven by power station characteristics, significantly improving the feasibility and effectiveness of evaluation in scenarios where real-time meteorological monitoring conditions are insufficient or data transmission is limited. By constructing a historical irradiation benchmark using typical annual meteorological data, a time-series simulated power generation curve is generated in combination with the actual characteristic parameters of the power station, and dynamically matching the actual output power of the inverter within the target time period with the meteorological conditions of the approximate time period, a refined dynamic calculation of theoretical power generation is achieved. Therefore, even without the support of real-time irradiation data, potential problems such as photovoltaic module efficiency degradation and line loss anomalies can still be accurately identified through comparative analysis of DC-side system energy efficiency with benchmark data.

[0064] Compared to traditional methods, this implementation not only overcomes the bottleneck of missing regional meteorological monitoring data, but also, through the collaborative analysis of simulated time-series data and real-time operational data, enhances the evaluation model's adaptability to dynamic factors such as seasonal fluctuations and equipment aging, effectively supporting optimized operations and maintenance decisions. Furthermore, based on the long-term stability of typical annual data, this method eliminates the interference of short-term meteorological fluctuations on evaluation results, enabling performance evaluation to more objectively reflect the actual operating status of the PV system, providing a reliable basis for plant health assessment, power generation performance benchmarking, and prioritization of technological transformation.

[0065] In some embodiments, before step 206, the following steps are further included: Step 301: Obtain second inverters of multiple photovoltaic power stations within a preset area, and second actual power generation data and second theoretical power generation data within a target time period. The preset area is determined based on the location of the target photovoltaic power station, and the second theoretical power generation data is determined based on irradiation data in typical annual meteorological data.

[0066] Specifically, a preset area is first defined based on the geographic coordinates (latitude and longitude) of the target PV power station. Usually, a square area with a side length (such as 10km, 20km) or a circular area with a radius of 10km or 20km is set with the target power station as the center. Multiple PV power stations in the area that are connected to the same operation and maintenance platform or data sharing network are screened out through the geographic information system (GIS) or power station location database.

[0067] For each screened PV power station (which may include the target PV power station), obtain the second actual power generation data of its second inverter (which may include the first inverter) within the target time period (such as the sunshine period of the day). This data must be collected in real time from the inverter monitoring system or SCADA (Supervisory Control And Data Acquisition) platform. If it is the AC side power generation, the actual DC side power generation must be inferred from the inverter efficiency parameters (the formula is "DC actual power generation = AC actual power generation / inverter efficiency"). At the same time, the inverter rated power is recorded for subsequent normalization processing.

[0068] The determination of the second theoretical power generation data needs to be combined with the actual hardware parameters of each power station, including the type of photovoltaic modules, installation inclination, total module area and rated conversion efficiency, etc., and use the reference irradiation data corresponding to the target time period in the typical annual meteorological data (such as the irradiation data of the approximate time period determined by the matching algorithm). The total irradiation received by the module is calculated by the slope irradiation conversion model, and then calculated and accumulated hourly by the formula "theoretical power generation = total irradiation × total module area × module efficiency × inverter efficiency" to form the second theoretical power generation data within the target time period.

[0069] During implementation, the time resolution of data from each power station needs to be unified (e.g., 15-minute intervals), and missing data or outliers need to be interpolated or removed to ensure the accuracy of subsequent analysis.

[0070] Step 302: Generate power generation energy efficiency benchmark data based on the second actual power generation data and the second theoretical power generation data.

[0071] Specifically, the energy efficiency ratio of the actual and theoretical data of each second inverter is first calculated using the formula "energy efficiency ratio = (second actual power generation / second theoretical power generation) × 100%), and the energy efficiency ratio data is normalized according to factors such as inverter model and rated power (such as converted into the energy efficiency ratio corresponding to the power of a single watt inverter) to eliminate the impact of equipment differences on the statistical results.

[0072] Subsequently, all normalized energy efficiency ratio data are statistically analyzed. The Shapiro-Wilk test method can be used to verify whether the data conforms to the normal distribution. If so, the sample mean, standard deviation and sample size are calculated. The critical value under the target confidence level (such as 95%) is determined based on the t distribution for small samples (n<30) and the z distribution for large samples (n≥30), and then the confidence interval of the power generation energy efficiency is constructed as the benchmark data; if the data does not conform to the normal distribution, the non-parametric test method (such as the Wilcoxon signed-rank test) is used to calculate the median and percentile range to form the benchmark data.

[0073] In addition, the power generation energy efficiency benchmark data can also be set by calculating the average, median or industry standard threshold of the energy efficiency ratio of all inverters in the region, and filtering similar data subsets based on the type of target photovoltaic power station (such as centralized, distributed) and equipment configuration to ensure the targeted nature of the benchmark data.

[0074] The final generated power generation energy efficiency benchmark data must include statistics (such as mean, standard deviation), confidence interval range or performance threshold, to provide a comparable benchmark reference for the power generation efficiency evaluation of the target PV power station and achieve an objective judgment of its operating efficiency within the target time period.

[0075] This embodiment uses the daily power generation data of multiple inverters within a certain geographical area as evaluation data for three reasons: 1. The AC side data after maximum power point tracking, optimization, and inversion within the inverter is more stable than the data of each string. This can, to a certain extent, reduce the power generation deviation between strings caused by the influence of installation conditions.

[0076] 2. The amount of inverter data is larger than the entire station data. The larger the sample size, the more data there is for statistical analysis, which facilitates improving the accuracy of subsequent statistical test methods.

[0077] 3. Data accessibility is relatively good. Compared with irradiation data, the power generation monitoring data of the inverter background is easier to obtain. Although there are occasional missing and inaccurate data, it is still usable overall. If there is a possibility of improving the monitoring accuracy in the future, further improvements will be made.

[0078] In some embodiments, step 302 may specifically include: S3021 , normalizing the second actual power generation data to obtain actual unit power power generation data, and normalizing the second theoretical power generation data to obtain theoretical unit power power generation data.

[0079] Among them, the power generation per unit power can be recorded as single watt power generation, that is, the actual or theoretical amount of electricity generated per watt of rated power in a specific time period, which is used to measure the power generation capacity and efficiency per unit power.

[0080] In this embodiment, data conversion needs to be performed on the rated power (in W) of each second inverter. The specific method is to divide the actual power generation of the inverter during the target time period (in Wh, if it is AC side data, the actual power generation on the DC side needs to be inferred through the inverter efficiency, the formula is "DC actual power generation = AC actual power generation / inverter efficiency") by its rated power to obtain "actual unit power generation data" (in Wh / W). This data represents the actual power generation capacity of the inverter with a rated power of each watt during the target time period.

[0081] When normalizing the second theoretical power generation data, the same logic is used to divide the theoretical power generation calculated based on the reference irradiation data (in Wh, which also needs to be consistent with the energy dimension of the actual power generation. If it is the theoretical value on the AC side, the theoretical value on the DC side needs to be forward calculated through the inverter efficiency) by the rated power of the inverter to obtain the "theoretical unit power power generation data" (in Wh / W). This eliminates the dimensional impact caused by differences in parameters such as the rated power of different inverters, allowing the actual and theoretical power generation data to be directly compared under the same unit standard, thereby accurately reflecting the power generation efficiency.

[0082] During implementation, it is necessary to ensure that the rated power parameters of all inverters are accurate, identify and eliminate outliers in the data (such as power generation that is obviously beyond the reasonable range), and interpolate missing data through the mean of adjacent time periods or mark it as invalid samples to ensure that the normalized data can truly reflect the unit power generation efficiency of the inverter.

[0083] S3022: Use the actual unit power generation data as sample data and verify whether the sample data conforms to the normal distribution.

[0084] Specifically, the sample must first be verified for normality. The Shapiro-Wilk test can be used. This method is suitable for testing normality with small samples (n ≤ 5000). It is performed by calculating the goodness-of-fit statistic, W, and the corresponding p-value for the sample data's fit to the normal distribution. To implement this, all sample data are entered into statistical analysis software (such as the Python scipy library or the R language's shapiro.test function), with a significance level of α = 0.05. If the p-value obtained is greater than α (e.g., p > 0.05), the sample data are considered to have no significant evidence to reject the normality hypothesis, meaning they conform to a normal distribution. If the p-value is less than or equal to α, the data are considered to be non-normal. Before testing, a preliminary observation of the data distribution can be made by plotting a histogram, QQ plot, or calculating descriptive statistics such as skewness and kurtosis to assist in verifying the reliability of the results. Extreme outliers in the data (such as data points beyond three times the standard deviation) must first be identified and processed (such as correction or elimination) to avoid significant interference with the normality test results and ensure that the premise assumptions of subsequent statistical analysis are valid.

[0085] If it conforms to the normal distribution, proceed to step S3023. If the data does not conform to the normal distribution, a non-parametric test is performed: based on the median and distribution shape of the sample data, the Wilcoxon signed-rank test is used to compare the median difference between the actual power generation of the target power station and the reference data, or a non-parametric confidence interval is constructed through the quantile method (such as calculating the interquartile range (IQR) based on the 25% and 75% quantiles of the sample data, generating the interval [Q1-1.5IQR, Q3+1.5IQR] as the judgment threshold), thereby avoiding dependence on the data distribution shape; if the actual value of the target power station exceeds this interval, it is judged to be abnormal. This method is suitable for skewed, multi-peaked data or data containing outliers. Although the statistical power is lower than that of the parametric test, its robustness is guaranteed.

[0086] S3023, determine the distribution test method based on the sample size of the sample data.

[0087] Specifically, the appropriate distribution test method is selected based on the size of the sample size n. If the sample size n is less than 30, it is considered a small sample situation, and the t distribution is used for statistical inference. This distribution is suitable for scenarios where the population standard deviation is unknown and the sample size is small. Its shape is determined by the degrees of freedom df = n-1. If the sample size n is greater than or equal to 30, it is considered a large sample situation. According to the central limit theorem, the sample mean approximately follows a normal distribution, and the z distribution (standard normal distribution) can be used for analysis. In this case, there is no need to consider the influence of degrees of freedom, and the symmetry and known probability density characteristics of the normal distribution are directly used for calculation.

[0088] This step divides the sample size threshold to ensure that the subsequent statistical methods are consistent with the data characteristics and avoid benchmark data deviations caused by incorrect distribution assumptions.

[0089] S3024: Determine the distribution critical value corresponding to the target confidence level based on the distribution test method.

[0090] Based on the selected distribution test method, determine the distribution critical value corresponding to the target confidence level (e.g., 95%). If using the t-distribution, consult a t-distribution table based on the degrees of freedom (df = n-1) and the confidence level. For example, at a 95% confidence level (two-sided test) with 20 degrees of freedom (df = 20), the critical value is approximately ±2.086. This value represents the quantile under the t-distribution curve when the probability of each tail is 2.5%.

[0091] If using the z-distribution, directly select the critical value of the standard normal distribution corresponding to the 95% confidence level of ±1.96. This value is calculated by calculating the cumulative probability of the normal distribution, ensuring that 95% of the area under the curve is within the critical value range. Accurately obtaining the critical value is key to constructing a confidence interval. Its essence is to quantify the fluctuation range of the sample statistic at the target confidence level.

[0092] In some specific embodiments, the t-value can be obtained based on the query rules of the t-distribution table. Assuming the degrees of freedom (df) is 60, that is, the sample size minus 1 (assuming the sample size n = 61), the test method is a one-sided test, and the significance level α = 0.05 (i.e., the one-sided tail probability is 5%). In the t-distribution table, the left column represents the degrees of freedom, and the top column represents the one-sided or two-sided test critical values ​​corresponding to different significance levels. Here, searching horizontally along the row with df = 60, the corresponding t-value can be directly read under the column with a one-sided tail of 0.05 (or labeled t(0.05, 60)), which is 1.671. This t-value represents the quantile corresponding to the 5% right tail area in the t-distribution with 60 degrees of freedom. That is, when the statistic is greater than 1.671, the probability of rejecting the null hypothesis in a one-sided test is less than 0.05. This t-value is used to construct a one-sided confidence interval or perform a one-sided t-test to ensure the reliability of statistical inference of power generation energy efficiency benchmark data at the set significance level.

[0093] S3025, determine the confidence interval based on the distribution critical value, sample size, and the mean and standard deviation of the sample data.

[0094] Specifically, the confidence interval is calculated as follows: Where CI represents the confidence interval, represents the mean value of the sample data and is the center of the confidence interval; t is a statistic determined based on the t distribution, and its value is determined by the degrees of freedom and the target confidence level, reflecting the degree of deviation allowed under a specific confidence level; s represents the standard deviation of the sample data, which measures the degree of data dispersion; n is the sample size, is the square root of the sample size. What is calculated is the standard error, which reflects the size of the sampling error of the sample mean.

[0095] The meaning of the whole formula is to take the sample mean Centered by t statistic and standard error The product of determines the error margin and constructs an interval that contains the population mean.

[0096] For example, if If t=2, s=10, and n=25, the standard error is 10 / 5=2, the error range is 2×2=4, and the confidence interval is 50±4, that is, [46,54]. This means that at the corresponding confidence level, there is reason to believe that the population mean falls within this interval.

[0097] S3026: Determine power generation energy efficiency benchmark data based on the confidence interval and theoretical unit power generation data.

[0098] Specifically, we first need to clarify the physical meaning of the confidence interval, which is the normal fluctuation range of the unit power generation efficiency of similar inverters in a preset area at the target confidence level (e.g., a 95% confidence interval indicates a 95% probability of containing the true energy efficiency level).

[0099] Subsequently, the theoretical unit power generation data (ideal values ​​calculated based on typical annual irradiation data and equipment parameters) is associated with the confidence interval. If the theoretical value is within the confidence interval, it indicates that the theoretical power generation capacity of the target PV power station conforms to the statistical laws of similar equipment in the region; if it exceeds the interval, it is necessary to check whether there are deviations in the theoretical calculation parameters (such as component efficiency correction errors).

[0100] Ultimately, the power generation energy efficiency benchmark data is presented in the form of a confidence interval, including lower and upper limit values ​​(such as [0.82, 0.88]Wh / W), providing a benchmark reference for subsequent evaluation of the actual energy efficiency of the target power station. When the actual unit power generation falls within the interval, it is judged to be normal performance; otherwise, it is considered abnormal and triggers further diagnosis.

[0101] In some embodiments, the plant characteristics include the tilt angle, azimuth angle, and inverter efficiency of the photovoltaic modules.

[0102] The inclination angle of a photovoltaic module refers to the angle formed by the plane of the module and the horizontal plane, which directly affects the amount of solar radiation received and the power generation efficiency; the azimuth angle is the horizontal angle of the photovoltaic module (usually based on true north), which determines the light conditions it receives at different times and thus affects the power generation; the inverter efficiency refers to the energy conversion ratio when the inverter converts direct current into alternating current. The higher the efficiency, the smaller the conversion loss and the more effective power output.

[0103] The simulated power generation time series data includes the simulated power generation curve, which is a curve used to show the changes in the theoretical power generation power or electricity of the power station at different times. It can be used to evaluate power generation performance and optimize system design.

[0104] Step 202 may specifically include: S2021: Generate an inclined surface irradiance curve based on the inclination angle, azimuth angle, and horizontal surface irradiance in typical annual meteorological data.

[0105] Specifically, first clarify the inclination angle of the photovoltaic module (that is, the angle between the module plane and the horizontal plane, usually ranging from 0° to 90°, which affects the incident angle of the sun's rays) and azimuth angle (the angle rotated clockwise with due north as 0°, 0° is due south, which determines the direction of the module). Combined with the minute-by-minute or hourly horizontal irradiance (including direct and diffuse irradiance) in the typical annual meteorological data, the solar declination angle (latitude of the sun's direct point) and hour angle (the offset angle of the sun relative to local noon) at different times are determined through the solar position calculation model, and then the solar altitude angle (the angle between the sun's rays and the horizontal plane) and solar azimuth angle (the angle between the projection of the sun's rays on the horizontal plane and the due north direction) are calculated.

[0106] On this basis, the horizontal surface irradiance is decomposed into three parts: direct irradiance, diffuse irradiance and ground reflected irradiance. The direct irradiance needs to calculate the incident angle according to the solar altitude angle and the component inclination angle, and is converted into the direct irradiance of the inclined surface through the cosine correction formula. The diffuse irradiance is converted using an isotropic model (such as uniformly distributing the horizontal surface scattering to the inclined surface) or a more accurate anisotropic model (such as the Perez model, taking into account the sky scattering distribution). The ground reflected irradiance is calculated based on the geometric relationship between the inclined surface and the ground using the reflectivity parameter (usually 0.2 to 0.3).

[0107] The above conversion calculation is performed moment by moment (e.g., every 15 minutes), ultimately forming an inclined surface irradiance curve that includes the total irradiance (direct + scattered + reflected) of the inclined surface at each moment. This curve accurately reflects the temporal variation characteristics of the solar radiation energy actually received by the photovoltaic module, providing key input data for subsequent power conversion.

[0108] S2022: Generate a theoretical inverter power curve based on the inverter efficiency and the tilted surface irradiation curve.

[0109] Specifically, first calculate the irradiance value (unit W / m 2 ) is converted into DC input power of photovoltaic modules. When making specific calculations, the actual total installation area of ​​the modules (unit: m2) must be taken into account. 2 ) and photoelectric conversion efficiency (taking temperature correction into account, such as the module efficiency decreases with increasing temperature, the correction formula is η = η_ref × [1-γ(T-T_ref)], where γ is the temperature coefficient, T is the ambient temperature, and T_ref is the standard test temperature of 25°C), then the module DC power P_dc = inclined surface irradiance × total module area × conversion efficiency.

[0110] Subsequently, the DC power at each moment is converted based on the inverter's efficiency characteristics (usually provided by the equipment manufacturer, which shows the relationship between input DC power and output AC power, such as approximately 95% efficiency at low load and approximately 98% efficiency at rated load). If the inverter efficiency curve is a piecewise linear or nonlinear function, an interpolation method (such as linear interpolation or cubic spline interpolation) is required to determine the efficiency value at the corresponding input power, and calculate the theoretical AC output power P_ac = P_dc × inverter efficiency. For inverters with multiple MPPT (maximum power point tracking) channels, calculations must be performed for each sub-array separately and then accumulated.

[0111] After completing the above power conversion calculations moment by moment, a theoretical inverter power curve is formed with time as the horizontal axis and theoretical AC power as the vertical axis. This curve takes into account the photovoltaic conversion characteristics of the components and the energy loss of the inverter, and truly simulates the power output capacity of the photovoltaic system under ideal operating conditions, providing a theoretical benchmark for subsequent comparison with actual data.

[0112] S2023: Using a dynamic time warping algorithm, align the time axis of the theoretical inverter power curve with the actual power curve of the target photovoltaic power station to obtain a simulated power generation curve of the target photovoltaic power station.

[0113] To address the possible time axis misalignment problem between the theoretical inverter power curve and the actual power curve of the target photovoltaic power station (such as time point misalignment caused by different sampling frequencies, clock deviation, and data missing), the Dynamic Time Warping (DTW) algorithm is used to perform time axis alignment.

[0114] Specifically, the dynamic time warping algorithm is used to correct the time axis deviation between the theoretical inverter power curve and the actual power curve, ensuring that the morphological characteristics of the two in the time dimension are accurately matched. The process starts with data input: the theoretical power curve is generated based on meteorological data and power station parameters, and is usually a smooth curve under ideal lighting conditions; the actual power curve comes from the SCADA system or data collector of the target power station, which may produce time phase offsets due to cloud cover, inverter response delays, or data acquisition clock asynchrony. First, the two curves are normalized to eliminate dimensional differences (such as converting power values ​​into per-unit values ​​or percentages), and then a distance matrix is ​​constructed between the points in the two sequences. The matrix elements are the Euclidean distance or Manhattan distance between the theoretical value and the actual value at the corresponding time point.

[0115] The core of DTW lies in finding an optimal path from the matrix starting point (1,1) to the end point (n,m), minimizing the cumulative distance of all points along the path. The path extension direction is only allowed to move rightward, upward, or diagonally, thereby achieving elastic scaling of the time axis. For example, if the actual curve experiences a sudden power drop and delayed recovery due to rapid cloud movement during a certain period of time, the DTW algorithm will stretch the time axis of the corresponding section of the theoretical curve so that the starting point of the power drop and the end point of the power recovery are aligned with the actual curve. After the path search is completed, a time warping function is generated based on the optimal path. The timestamps of the theoretical curve are mapped to the actual time axis according to the warping function, and the final output is the time-aligned simulated power generation curve.

[0116] To ensure the quality of alignment, the root mean square error (RMSE) of the two curves after warping and the curvature of the dynamic warping path (Warping Path Cost) can also be calculated. If the RMSE exceeds a preset threshold (such as 5%) or the curvature is too high (indicating an abnormal degree of time distortion), the alignment is judged to have failed and an alarm is triggered, prompting manual review of the data quality or adjustment of the theoretical model parameters. This method is particularly suitable for scenarios where actual power fluctuates frequently in cloudy weather. It can effectively eliminate timing misalignments caused by asynchronous data acquisition clocks or dynamic changes in local shadows, and improve the accuracy of subsequent energy efficiency ratio calculations. In terms of computational complexity, the original time complexity of DTW is O(nm), which can be optimized to near-linear complexity through sliding window constraints (such as Sakoe-Chiba Band) or multi-level hierarchical warping strategies to meet the real-time processing requirements of high-resolution (such as 1-minute) data.

[0117] The simulated power generation curve accurately matches the fluctuation characteristics of theoretical power and actual power in the time dimension, eliminating the interference of time misalignment on subsequent energy efficiency evaluation, and laying the foundation for accurate comparison of theoretical and actual power generation performance.

[0118] In some embodiments, step 203 may further include: S2031: Divide a typical year into multiple time segments based on the target time period.

[0119] Specifically, first determine the length of the target time period (e.g., the target time period is the daylight period, assuming a duration of T hours). Based on this duration, divide the typical year's meteorological data into multiple continuous and non-overlapping time segments in chronological order. The division is done on an hourly or minutely basis, ensuring that the time span of each time segment is consistent with the target time period (e.g., T hours). For example, if the target time period is 6 hours (08:00-14:00), the same period of each day in the typical year (08:00-14:00) is used as a time segment. Alternatively, a non-full-day target time period (e.g., 3 hours) can be divided using a sliding window approach (e.g., sliding from 00:00-03:00, 01:00-04:00, in that order). During the segmentation process, the time resolution of typical annual meteorological data (such as 15-minute intervals) needs to be processed to ensure that each time segment contains complete time point data. For target time periods spanning days (such as 23:00-02:00 the next day), the data of two adjacent days need to be merged into one segment, ultimately forming multiple time segments of equal or custom length covering the entire year, providing structured data units for subsequent irradiance curve generation.

[0120] S2032: Generate irradiation curves for each time segment based on irradiation data of a typical year.

[0121] Specifically, for each time segment, the irradiance data (including direct irradiance, diffuse irradiance and total irradiance) in the corresponding time period is extracted from the typical annual meteorological data. The time resolution must be consistent with the actual power data of the target time period, such as 15 minutes, and the irradiance curve of the segment is formed in chronological order. If the component inclination and azimuth of the target photovoltaic power station are known, the horizontal plane irradiance must first be converted into the component inclined surface irradiance through the inclined plane irradiance conversion model (such as the Perez model or the isotropic model). The specific steps include calculating the solar altitude angle, azimuth and incident angle, decomposing the direct and diffuse irradiance and performing geometric correction, and considering the ground reflected irradiance (the reflectivity is usually taken as 0.2). For scenes where the inclined plane conversion is not performed (such as directly using the horizontal plane irradiance as a reference), the horizontal plane total irradiance data of the corresponding time period is directly extracted. Each irradiance curve uses time as the horizontal axis (such as the specific time period from 00:00 to 24:00), irradiance (unit W / m 2 ) is the vertical axis, which clearly presents the real-time change characteristics of solar radiation energy in this time segment, providing an environmental data basis for subsequent power curve matching.

[0122] S2033: Determine an approximate power curve from a plurality of irradiation curves based on the first actual output power data.

[0123] Specifically, the first actual output power data of the first inverter of the target PV power station during the target time period (e.g., AC power values ​​in kW every 15 minutes) is obtained and converted into a power sequence consistent with the temporal resolution of the irradiance curve. (If the actual data contains missing or outliers, it must first be corrected using linear interpolation or neighborhood mean.) For each time segment of the irradiance curve, the theoretical model of the PV system is used to convert it into the corresponding theoretical inverter power curve: first, the irradiance is multiplied by the total module area and the photovoltaic conversion efficiency (accounting for temperature correction) to obtain the DC input power, which is then multiplied by the inverter efficiency (using the rated efficiency of the device or a typical efficiency curve) to obtain the theoretical AC power. Subsequently, a similarity matching algorithm (such as dynamic time warping (DTW), Pearson correlation coefficient (PCC), or root mean square error (RMSE)) is used to calculate the similarity between the actual power curve and all theoretical power curves. For example, DTW constructs a distance matrix and finds the optimal time mapping path to quantify the morphological match between the two curves in the time series. For PCC, the linear correlation coefficient between the actual power and the theoretical power is calculated. The closer the coefficient is to 1, the higher the similarity. Finally, the theoretical power curve with the highest similarity is selected as the "approximate power curve", and its corresponding irradiation curve is the environmental irradiation condition that best matches the actual power generation state.

[0124] S2034: Determine the time period corresponding to the approximate power curve as the approximate time period.

[0125] Specifically, after determining the approximate power curve, the specific time period of the time segment corresponding to the curve in the typical year (such as 09:00-15:00 on May 10, 2023) is extracted, and this time period is the "approximate time period". During the specific operation, each time segment has recorded its corresponding calendar time (such as year, month, day, hour, and minute) when it is divided. By matching the index or label of the approximate power curve, the corresponding start and end timestamps in the typical year meteorological data are directly located. For example, if the theoretical power curve with the highest similarity comes from the 08:00-14:00 segment on the 120th day of the typical year, the approximate time period is 08:00-14:00 on that date, and its corresponding irradiance data (including direct and scattered irradiance at each moment) is used as the reference irradiance data for the subsequent calculation of the first theoretical power generation.

[0126] This process ensures that the time attributes of the approximate time period strictly correspond to the typical annual meteorological data through data label association and index matching, providing accurate environmental parameter input for theoretical power generation calculation and achieving quantitative matching between actual power generation status and typical meteorological conditions.

[0127] In some embodiments, step 205 specifically includes: S2051: Determine first energy efficiency data based on first actual power generation data and first theoretical power generation data.

[0128] Specifically, the first actual power generation data of the first inverter in the target time period (in Wh, if it is the AC side power, the actual DC side power generation needs to be inferred through the inverter efficiency, the formula is "DC actual power generation = AC actual power generation / inverter efficiency") and the first theoretical power generation data (theoretical DC side input energy calculated based on the irradiation data of the typical year approximate time period, in Wh) are quantitatively compared. During the specific operation, the energy efficiency ratio is directly calculated using the formula "first energy efficiency data = (first actual power generation / first theoretical power generation) × 100%". This ratio reflects the proportional relationship between the actual energy conversion efficiency and the theoretical efficiency of the photovoltaic system under typical meteorological conditions. If the actual power generation data is an AC side measurement value, it must first be corrected according to the average efficiency of the inverter in the target time period (which can be obtained through real-time monitoring or equipment parameters) to ensure that it is consistent with the energy dimension of the first theoretical power generation data (both DC side or both AC side).

[0129] Before calculation, the data must be eliminated for outliers (such as the extreme case where the actual power generation exceeds the theoretical value by 110%) to avoid interference with the energy efficiency calculation caused by equipment failure or measurement errors, and ultimately obtain the first energy efficiency data that characterizes the operating efficiency of the system under typical conditions.

[0130] S2052: Determine second energy efficiency data based on the first actual power generation data and third theoretical power generation data of the first inverter, where the third theoretical power generation data is generated based on real-time meteorological data of the target photovoltaic power station location within a target time period.

[0131] Specifically, first obtain real-time meteorological data of the target photovoltaic power station location within the target time period, including minute-by-minute or hourly solar irradiance (inclined surface irradiance, if it is horizontal surface data, it needs to be combined with the component inclination and azimuth angle through the Perez model conversion), ambient temperature, humidity and other parameters.

[0132] Based on real-time irradiance data and combined with PV module parameters (total area, rated conversion efficiency, and taking into account the temperature correction factor, such as η = η_ref × [1-γ(T-T_ref)], where γ is the temperature coefficient, T is the real-time temperature, and T_ref is the standard test temperature of 25°C), the theoretical DC input energy of the module is calculated using the formula "DC theoretical energy = real-time total irradiance × total module area × corrected conversion efficiency."

[0133] The third theoretical power generation data (theoretical output energy on the AC side = theoretical DC energy × inverter efficiency) is further calculated based on the real-time efficiency of the first inverter (if the real-time efficiency is not available, the efficiency curve of the device at the corresponding load rate is interpolated).

[0134] The third theoretical power generation data is compared with the first actual power generation data (if it is AC side data, it is directly compared; if it is DC side data, it needs to be converted to AC side through inverter efficiency), and the energy efficiency ratio is calculated using "second energy efficiency data = (first actual power generation / third theoretical power generation) × 100%". This data reflects the degree of match between the actual power generation efficiency of the photovoltaic system under real-time meteorological conditions and the immediate theoretical efficiency, providing a real-time benchmark for dynamic evaluation of system performance.

[0135] S2053: Determine DC side system energy efficiency data according to the first energy efficiency data and the second energy efficiency data.

[0136] Specifically, the DC-side system energy efficiency data is determined by combining the first and second energy efficiency data through statistical analysis. First, ensure that the energy dimensions of the two are consistent (both are DC-side energy efficiency ratios). If the second energy efficiency data is calculated based on the third theoretical power generation on the AC side, it must first be converted to a DC-side ratio using the formula "DC-side second energy efficiency = AC-side second energy efficiency × inverter efficiency."

[0137] Subsequently, the two energy efficiency data sets were checked for plausibility, with outliers exceeding 20% ​​(e.g., extreme values ​​caused by sudden changes in real-time meteorological data or temporary equipment failures) removed. The final DC-side system energy efficiency data was calculated using either a weighted average method (weights assigned based on data reliability, such as 0.6 for typical year data and 0.4 for real-time data) or a median method. If the difference between the two was within 5%, the arithmetic mean was taken as the final result. If the difference exceeded 10%, the data source was investigated (e.g., whether the typical year data matched the current season, whether the real-time data contained sensor errors), and the data was corrected and recalculated.

[0138] The final DC-side system energy efficiency data integrates the long-term performance under typical meteorological conditions and the immediate performance in a real-time environment. It can more comprehensively reflect the energy conversion efficiency of the photovoltaic system within the target time period, and provide a quantitative basis for the accurate evaluation of power generation efficiency in multiple dimensions.

[0139] In some embodiments, step S2053 may specifically include: (1) Determine the weight coefficients of the first energy efficiency data and the second energy efficiency data according to the completeness of the real-time meteorological data.

[0140] Specifically, a completeness assessment indicator system must be established first, including the data missing rate (the proportion of time when real-time meteorological data is missing during the target time period), the proportion of outliers (the proportion of data points where the irradiance exceeds 120% of the theoretical maximum value or the temperature exceeds the equipment operating range) and the consistency of time resolution (whether the sampling intervals of real-time data and actual power data match, such as whether both are 15 minutes).

[0141] During implementation, the missing rate of real-time meteorological data is first calculated. If the missing time ratio is ≤5%, it is judged to be highly complete, and a higher weight (such as 0.6 to 0.8) is assigned to the second energy efficiency data. If the missing rate is between 5% and 20%, it is judged to be moderately complete, and the weight is adjusted to 0.4 to 0.6. If the missing rate is >20%, it is considered to be low completeness, and the weight of the second energy efficiency data is reduced to below 0.3, and the weight of the first energy efficiency data is correspondingly increased (based on typical annual data, with higher stability).

[0142] At the same time, outliers are identified. If the proportion of outliers exceeds 10%, the real-time data weight is adjusted down by 0.1 to 0.2. The final weight coefficient is determined through linear interpolation or piecewise function to ensure that the sum of the weights is 1. For example, when the completeness is high, the weight distribution is (0.3, 0.7), and when the completeness is low, it is (0.7, 0.3), reflecting quantitative feedback on data reliability.

[0143] In some embodiments, in addition to the completeness of meteorological data (MR, missing rate), it is also necessary to consider the impact of temporal resolution consistency (TRC), outlier ratio (OR), seasonal correlation (SC) and equipment status parameters (such as inverter temperature deviation ΔT, component cleanliness C, etc.) on the weights of the first energy efficiency data and the second energy efficiency data.

[0144] When allocating weight coefficients to the first energy efficiency data and the second energy efficiency data based on the completeness of meteorological data, the completeness of real-time meteorological data is first used as the core indicator, and the exponential decay function (such as e -MR / 0.2 , MR is the missing rate) to quantify its impact on the credibility of real-time data. The higher the missing rate, the lower the weight of real-time data.

[0145] At the same time, the consistency of time resolution is taken into consideration. If the sampling interval of real-time data is consistent with that of actual power data, a weight correction factor of 1 is assigned. If they are inconsistent, the weight correction factor is reduced to 0.5 or 0, which directly weakens the impact of time-misaligned data.

[0146] For the proportion of outliers, a piecewise function is used for processing (the correction factor is 1 when the proportion of outliers is ≤5%, and it is reduced to 0.2 when it is above 30%) to suppress the interference of noise data.

[0147] Seasonal correlation is achieved by calculating the correlation coefficient (such as Pearson coefficient) between the target period and the corresponding seasonal irradiation curve of a typical year, and using it as a weight scaling factor (the higher the correlation coefficient, the higher the weight of the real-time data).

[0148] In the device status parameters, the inverter temperature deviation is calculated by a logic function (such as 1 / (1+e ΔT / 5)) simulates the effect of high temperature on efficiency degradation, and component cleanliness is used directly as a linear correction factor (cleanliness score 0-1). The final weight coefficient is obtained by normalizing the product of the above factors. The calculation formula for the real-time data weight (the weight coefficient corresponding to the second energy efficiency data) W2 is as follows: W1=1-W2; W1 is the weight coefficient of typical year data, that is, the weight coefficient corresponding to the first energy efficiency data.

[0149] This embodiment implements the nonlinear influence of multi-dimensional factors on weights, ensuring that the energy efficiency evaluation results reflect both the real-time operating status and rely on a stable historical benchmark.

[0150] (2) Determine the DC side system energy efficiency data according to the weight coefficient, the first energy efficiency data, and the second energy efficiency data.

[0151] Specifically, first unify the energy dimensions of the two. If the second energy efficiency data is calculated based on the third theoretical power generation on the AC side, it needs to be converted into a DC side ratio through "DC side second energy efficiency = second energy efficiency data × average efficiency of the inverter in the target time period" to ensure consistency with the unit of the first energy efficiency data (already DC side).

[0152] The weighted average method is then used to calculate the comprehensive energy efficiency data. The formula is "DC side system energy efficiency data = first energy efficiency data × first weight + second energy efficiency data × second weight", where the first weight and the second weight are determined by the completeness of the real-time meteorological data (such as the first weight = 1-second weight).

[0153] During implementation, the rationality of the weight coefficients must be verified. If the completeness of the real-time data is 0 (completely missing), the second weight = 0, and the first energy efficiency data is directly used as the result. If the completeness is 100%, the second weight takes the maximum value (such as 0.8) to emphasize real-time performance feedback. During the calculation process, the energy efficiency data is boundary-checked to ensure that the results are within the range of 0% to 100%. Abnormal values ​​outside the reasonable range (such as the second energy efficiency data >110% due to equipment failure) are truncated (taken as 100%). The final result is the DC side system energy efficiency data that integrates typical long-term performance and real-time environmental performance, providing a quantitative basis for power generation efficiency evaluation that balances stability and real-time performance.

[0154] like Figure 2 As shown, the method of this embodiment can be divided into the following main branches and steps: Part 1: Calculation and comparison of theoretical power generation based on meteorological data (left branch): Obtain actual meteorological data from historical weather stations within a 10-kilometer radius. Determine whether complete actual meteorological monitoring data is available. If so, calculate the total irradiation of the PV array for that day. Next, calculate the theoretical inverter power generation per watt for that day. Combined with the actual power generation, calculate the DC-side energy efficiency ratio of the PV system. If not, meaning that complete meteorological data is unavailable, obtain data from a typical year and generate a daily power generation curve for the entire year based on plant characteristics and curve fitting. Using a matching algorithm, match the inverter power generation curve for that day to an approximate day in a typical year. Calculate the theoretical inverter power generation per watt for that day. Combined with the actual power generation, calculate the DC-side energy efficiency ratio of the PV system.

[0155] The second part is the statistical analysis based on the actual power generation data of the inverter (right branch): Conduct a search within a 10-kilometer radius around the power station being evaluated. Obtain the per-watt power generation of all inverters for the evaluated power station and all power stations within the search area on that day. Verify that the data is normally distributed. If so, use a t-test or z-test, depending on the sample size. Construct a 95% confidence interval based on the actual per-watt power generation of all inverters compared to the theoretical per-watt power generation. If not, indicating a non-normal distribution, perform a nonparametric test. This allows for valid statistical inference and comparison even when the data do not meet the assumptions of parametric tests (such as normal distribution).

[0156] Part 3: Calculation and final determination of energy efficiency ratio (combination of the middle and lower parts): Following the output from the left branch of the first part and the right branch of the second part: The DC system energy efficiency ratio calculated based on different data is weighted and calculated using a weighting factor. The left branch calculates the energy efficiency ratio based on meteorological data and theoretical models, while the right branch constructs a confidence interval for per-watt power generation based on actual power generation data. This confidence interval can be considered a statistical range of actual performance and can be combined with the theoretical value to obtain the range of the energy efficiency ratio. More specifically, the right branch of the second part produces a confidence interval for per-watt power generation, which, combined with the theoretical per-watt power generation, can be converted into a confidence interval for the energy efficiency ratio. The "DC system energy efficiency ratio calculated based on different data" may refer to a single-point energy efficiency ratio calculated using actual daily irradiation (if available) or a single-point energy efficiency ratio calculated using typical daily data. These single-point actual energy efficiency ratios, or a comprehensive weighted actual energy efficiency ratio, are then compared with the "energy efficiency ratio confidence interval" derived from the statistical analysis on the right.

[0157] The actual energy efficiency ratio is compared with the confidence interval for evaluation. The "confidence interval" here refers to a reasonable range of fluctuations for normal energy efficiency ratio performance, established through statistical methods (such as t-tests, z-tests, or non-parametric tests). The "actual energy efficiency ratio" refers to the actual performance of the power plant (or inverter) during the current evaluation period. If the actual energy efficiency ratio is within the confidence interval, it is considered normal; if it is outside the interval, it is considered abnormal.

[0158] The above method begins with obtaining basic data (meteorological data and inverter power generation data), calculates theoretical power generation performance through theoretical modeling, and establishes a normal fluctuation range (confidence interval) through statistical analysis of actual power generation data. Finally, the actual power generation efficiency (which may be a weighted actual energy efficiency ratio) is compared with this statistically derived normal range to evaluate and determine the power station's daytime power generation performance.

[0159] The photovoltaic power station power generation efficiency evaluation method provided by the embodiment of the present invention effectively solves the limitation of traditional technology that relies on real-time meteorological data, and significantly improves the evaluation accuracy and applicability in non-irradiation monitoring scenarios.

[0160] By integrating typical annual meteorological data with the actual operating data of the target power station, a dynamic theoretical model based on time series matching is constructed, bypassing the traditional method's strong dependence on real-time irradiance input.

[0161] During the data input stage, typical annual meteorological data is introduced as the basis, and the characteristic parameters of the power station are combined to generate a simulated power generation timing curve. The theoretical curve is aligned with the actual inverter power curve through the dynamic time warping algorithm, and the approximate time period in the typical year is accurately located and reference irradiation data is extracted to ensure that the theoretical power generation calculation is highly adapted to the meteorological conditions of the target time period.

[0162] During the model building phase, normality testing and confidence interval modeling were performed using inverter single-watt power generation data from neighboring power stations to establish statistically driven energy efficiency benchmark data, thus addressing the risk of failure of a single meteorological model when data is missing.

[0163] Through a dynamic weighting mechanism, the weight coefficients of the physical model and the statistical model are adaptively adjusted according to the completeness of the meteorological data, which not only retains the high precision advantage of meteorological data but also enhances the robustness of the statistical model, so that the energy efficiency ratio calculation results remain stable and reliable in complex scenarios.

[0164] This embodiment supports fine-grained evaluation of DC-side system performance, accurately identifying efficiency deviations caused by abnormal factors such as component degradation, dust obstruction, and line loss. It also quantitatively categorizes abnormal conditions through confidence interval comparisons, providing an intuitive basis for operational and maintenance decisions. Furthermore, by normalizing per-watt power generation, the impact of differences in power plant capacity and equipment models is eliminated, enabling cross-regional and multi-plant performance comparisons, significantly enhancing the method's universality.

[0165] The final output of power generation efficiency evaluation results takes into account both short-term abnormal alarms and long-term performance trend analysis, providing reliable technical support for the intelligent operation and maintenance and energy efficiency optimization of photovoltaic power stations.

[0166] The method provided in the above embodiment can be executed by an electronic device. The following describes the electronic device in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , which is a schematic diagram of a physical device structure of an electronic device in an embodiment of the present invention.

[0167] It should be noted that Figure 3 The structure of the electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0168] like Figure 3 As shown, the electronic device includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 into the random access memory (RAM) 403, such as the method described in the above embodiment. In the random access memory (RAM) 403, various programs and data required for system operation are also stored. The central processing unit (CPU) 401, the read-only memory (ROM) 402 and the random access memory (RAM) 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0169] The following components are connected to the input / output (I / O) interface 405: an input section 406 including an audio input device, a push button switch, etc.; an output section 407 including a display, an audio output device, an indicator light, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output (I / O) interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed in the drive 410 as needed so that a computer program read therefrom can be installed into the storage section 408 as needed.

[0170] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 409 and / or installed from a removable medium 411. When the computer program is executed by the central processing unit (CPU) 401, the various functions defined in the present invention are performed.

[0171] It should be noted that specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0172] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0173] Specifically, the electronic device of this embodiment includes a processor and a memory, the memory is coupled to one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and one or more processors call the computer instructions to enable the electronic device to execute the method provided by the above embodiment.

[0174] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments, or may exist independently and not incorporated into the electronic device. The storage medium carries one or more computer programs, and when executed by a processor of the electronic device, the electronic device implements the methods provided in the above embodiments.

[0175] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

[0176] As used in the above embodiments, the term “when…” may be interpreted to mean “if…” or “after…” or “in response to determining…” or “in response to detecting…”, depending on the context. Similarly, the phrases “upon determining…” or “if (stated condition or event) is detected” may be interpreted to mean “if determining…” or “in response to determining…” or “upon detecting (stated condition or event)” or “in response to detecting (stated condition or event)”, depending on the context.

[0177] Those skilled in the art will appreciate that all or part of the process steps in the above-described method embodiments can be implemented by a computer program instructing the relevant hardware. The program can be stored in a computer-readable storage medium, and when executed, the program can include the process steps in the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for evaluating the power generation efficiency of a photovoltaic power station, characterized in that: include: Obtain typical annual meteorological data for the location of the target PV power station; generating simulated power generation time series data of the target photovoltaic power station according to the power station characteristics of the target photovoltaic power station and the typical annual meteorological data; Determining an approximate time period and reference irradiation data in a typical year based on the simulated power generation time series data and first actual output power data of the first inverter of the target photovoltaic power station in a target time period, where the reference irradiation data is irradiation data for the approximate time period in the meteorological data of the typical year; determining first theoretical power generation data of the first inverter within the target time period based on the reference irradiation data; determining DC side system energy efficiency data of the target photovoltaic power station according to first actual power generation data of the first inverter in the target time period and the first theoretical power generation data; When the power generation energy efficiency benchmark data is determined, a power generation efficiency evaluation result of the target photovoltaic power station within the target time period is determined based on the DC side system energy efficiency data and the power generation energy efficiency benchmark data.

2. The method according to claim 1, characterized in that Before determining the power generation efficiency evaluation result of the target photovoltaic power station based on the DC side system energy efficiency data and the power generation energy efficiency benchmark data, the method further includes: Obtaining second actual power generation data and second theoretical power generation data of second inverters of multiple photovoltaic power stations within a preset area within the target time period, wherein the preset area is determined based on the location of the target photovoltaic power station, and the second theoretical power generation data is determined based on irradiation data in the typical year meteorological data; The power generation energy efficiency benchmark data is generated according to the second actual power generation data and the second theoretical power generation data.

3. The method according to claim 2, characterized in that The generating the power generation energy efficiency benchmark data according to the second actual power generation data and the second theoretical power generation data includes: Normalizing the second actual power generation data to obtain actual unit power power generation data, and normalizing the second theoretical power generation data to obtain theoretical unit power power generation data; Using the actual unit power generation data as sample data, and verifying whether the sample data conforms to a normal distribution; If yes, determining the distribution test method according to the sample size of the sample data; Determining a distribution critical value corresponding to a target confidence level based on the distribution test method; Determining a confidence interval based on the distribution critical value, the sample size, and the mean and standard deviation of the sample data; The power generation energy efficiency benchmark data is determined based on the confidence interval and the theoretical unit power generation data.

4. The method according to claim 1, wherein The power station characteristics include the inclination angle, azimuth angle and inverter efficiency of the photovoltaic modules, and the simulated power generation time series data includes a simulated power generation curve; generating the simulated power generation time series data of the target photovoltaic power station based on the power station characteristics of the target photovoltaic power station and the typical annual meteorological data includes: generating an inclined surface irradiance curve according to the inclination angle, the azimuth angle, and the horizontal surface irradiance in the typical year meteorological data; generating a theoretical inverter power curve based on the inverter efficiency and the inclined surface irradiation curve; The theoretical inverter power curve is aligned with the time axis of the actual power curve of the target photovoltaic power station through a dynamic time warping algorithm to obtain a simulated power generation curve of the target photovoltaic power station.

5. The method according to claim 1, wherein Determining an approximate time period in a typical year based on the simulated power generation time series data and first actual output power data of the first inverter of the target photovoltaic power station in a target time period includes: dividing a typical year into a plurality of time segments based on the target time period; generating an irradiation curve for each of the time segments according to the irradiation data of the typical year; determining an approximate power curve from a plurality of the irradiance curves based on the first actual output power data; The time period corresponding to the approximate power curve is determined as the approximate time period.

6. The method according to any one of claims 1 to 5, characterized in that The determining, based on the first actual power generation data of the first inverter in the target time period and the first theoretical power generation data, the DC side system energy efficiency data of the target photovoltaic power station includes: determining first energy efficiency data based on the first actual power generation data and the first theoretical power generation data; determining second energy efficiency data based on the first actual power generation data and third theoretical power generation data of the first inverter, wherein the third theoretical power generation data is generated based on real-time meteorological data of the location of the target photovoltaic power station within the target time period; The DC side system energy efficiency data is determined according to the first energy efficiency data and the second energy efficiency data.

7. The method according to claim 6, characterized in that The determining the DC side system energy efficiency data according to the first energy efficiency data and the second energy efficiency data includes: Determining weight coefficients of the first energy efficiency data and the second energy efficiency data according to the completeness of the real-time meteorological data; The DC side system energy efficiency data is determined according to the weight coefficient, the first energy efficiency data, and the second energy efficiency data.

8. An electronic device, characterized in that: including one or more processors and memory; The memory is coupled to the one or more processors, and is configured to store computer program codes, where the computer program codes include computer instructions. The one or more processors call the computer instructions to enable the electronic device to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium storing computer instructions, characterized in that: When the computer instructions are executed on an electronic device, the electronic device is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on an electronic device, the electronic device is enabled to perform the method according to any one of claims 1 to 7.

Images