Photovoltaic module output power prediction method for photovoltaic power plant economic benefit evaluation

Abstract

The present application relates to the technical field of photovoltaic module power prediction, in particular to a photovoltaic module output power prediction method for economic benefit evaluation of photovoltaic power station, comprising: obtaining energy conversion efficiency characteristics according to the positive influence of daily irradiation conditions and environmental factors on photovoltaic module output power in a target historical light period, obtaining power generation efficiency characteristics of the target historical light period by combining the difference between daily air dew point temperature and photovoltaic module panel temperature and daily air temperature range, obtaining electric energy demand degree characteristics of the target historical light period according to the electricity price and photovoltaic power generation capacity of each electricity price period in a day, and then obtaining the light-electricity conversion benefit sensitivity degree of the target historical light period. The present application respectively analyzes from the aspects of power supply and power generation demand, determines training sample data according to the target historical light period obtained through screening, improves the accuracy of training sample data selection, and thus improves the accuracy of photovoltaic module output power prediction.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic module power prediction technology, and specifically to a method for predicting the output power of photovoltaic modules for the economic benefit assessment of photovoltaic power plants. Background Technology

[0002] Photovoltaic (PV) power generation boasts the advantages of being clean and renewable, but its output power is affected by multiple factors such as irradiance and temperature, exhibiting randomness, volatility, and intermittency. Accurate prediction of PV module output power is crucial for grid dispatch, energy storage configuration, electricity trading, and power plant economic evaluation. To ensure stable grid operation and maximize the economic benefits of power plants, accurate prediction of PV module output power has become an important research topic. Current methods typically employ network models such as machine learning models to predict PV module output power. To ensure the accuracy of these machine learning models, the selection of training sample data is critical. Training sample data is essentially PV data for a specific time period. However, current methods often fail to screen the training sample data from large amounts of historical data, arbitrarily using PV data from multiple time periods directly as training samples. If the time period corresponding to the training sample data has a low correlation with PV power generation efficiency (i.e., the training sample data is PV data with a low correlation to PV power generation efficiency), using this training sample data to train the machine learning model will affect the accuracy of the model, thus impacting the accuracy of PV module output power prediction. Summary of the Invention

[0003] To address the technical problem of inaccurate selection of training sample data when using network models to predict the output power of photovoltaic modules, the present invention aims to provide a method for predicting the output power of photovoltaic modules for economic benefit assessment of photovoltaic power plants. The specific technical solution adopted is as follows:

[0004] This invention provides a method for predicting the output power of photovoltaic modules for economic benefit assessment of photovoltaic power plants, comprising:

[0005] By combining the daily irradiance during the target historical sunshine period and the positive impact of environmental factors on the output power of photovoltaic modules, the daily energy conversion efficiency characteristics are obtained.

[0006] Based on the difference between the daily air dew point temperature and the photovoltaic module panel temperature, combined with the daily temperature range and the energy conversion efficiency characteristics, the power generation efficiency characteristics of the target historical sunshine period are obtained.

[0007] Based on the electricity price and photovoltaic power generation during each electricity price period of the day, the characteristics of the electricity demand during the target historical sunshine period are obtained;

[0008] By integrating the power generation efficiency characteristics and power demand characteristics of the target historical sunshine period, the sensitivity of photoelectric conversion efficiency of the target historical sunshine period is obtained. The sensitivity of photoelectric conversion efficiency is used to indicate the screening of the target historical sunshine period.

[0009] In one exemplary embodiment, the photovoltaic module output power prediction method further includes:

[0010] Obtain the inflection point in the daily output power curve of photovoltaic modules;

[0011] Based on the time corresponding to the inflection point, each day is divided into several core time periods.

[0012] In one exemplary embodiment, the process of obtaining the positive impact of daily irradiance on the output power of photovoltaic modules includes:

[0013] Determine the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power at each moment during the core time period;

[0014] The power characteristics at each moment are obtained based on the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power. The power characteristics are directly proportional to the solar altitude angle, irradiance, and photovoltaic module output power, and inversely proportional to the atmospheric particulate matter concentration.

[0015] By integrating the power characteristics at various times and combining the correlation between irradiance and photovoltaic module output power during the core time period, a first positive impact index for the core time period is obtained; the first positive impact index characterizes the positive impact of irradiance during the core time period on photovoltaic module output power.

[0016] In one exemplary embodiment, the correlation between the change in irradiance and the output power of the photovoltaic module is the Pearson correlation coefficient between irradiance and the output power of the photovoltaic module.

[0017] In one exemplary embodiment, the process of obtaining the positive impact of daily environmental factors on the output power of photovoltaic modules includes:

[0018] Determine the ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at each moment during the core time period.

[0019] By integrating ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at various times, a second positive impact index for the core time period is obtained. The second positive impact index is directly proportional to ambient wind speed and inversely proportional to atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature. The second positive impact index characterizes the impact of environmental factors on the output power of photovoltaic modules during the core time period.

[0020] In one exemplary embodiment, the daily energy conversion efficiency characteristics include energy conversion efficiency characteristics for each core time period of the day;

[0021] The process of obtaining the power generation efficiency characteristics includes:

[0022] Based on the energy conversion efficiency characteristics of the core time period and the difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period, the energy conversion factor for the core time period is obtained.

[0023] By integrating the energy conversion factors of each core time period of the day, the daily power generation efficiency sub-characteristics are obtained;

[0024] By integrating the daily power generation efficiency sub-features and the daily temperature range, the power generation efficiency characteristics of the target historical sunshine period are obtained; the power generation efficiency characteristics are both proportional to the power generation efficiency sub-features and the daily temperature range.

[0025] In an exemplary embodiment, the difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period is a temperature difference feature. If the average air dew point temperature during the core time period is greater than the average photovoltaic module panel temperature, then the temperature difference feature is a first temperature difference feature, which is proportional to the difference between the average air dew point temperature and the average photovoltaic module panel temperature. If the average air dew point temperature during the core time period is less than or equal to the average photovoltaic module panel temperature, then the temperature difference feature is a set value that is less than the first temperature difference feature.

[0026] The energy conversion factor is obtained from the energy conversion efficiency characteristics and temperature difference characteristics of the core time period. The energy conversion factor is directly proportional to the energy conversion efficiency characteristics and inversely proportional to the temperature difference characteristics.

[0027] In an exemplary embodiment, the process of obtaining the electricity demand level characteristics includes:

[0028] The proportion of photovoltaic power generation in each electricity price period of the day is determined, and the electricity price in each electricity price period of the day is combined to obtain the electricity demand in each electricity price period of the day; the electricity demand in each period is directly proportional to both the proportion of photovoltaic power generation and the electricity price.

[0029] By integrating the electricity demand for each electricity price period during the day within the target historical sunshine period, the characteristics of the electricity demand level during the target historical sunshine period are obtained.

[0030] In an exemplary embodiment, the photovoltaic module output power prediction method further includes: obtaining the photoelectric conversion efficiency sensitivity of multiple target historical illumination periods, and retaining the target historical illumination periods corresponding to the photoelectric conversion efficiency sensitivity that meet preset conditions as sample historical illumination periods.

[0031] In an exemplary embodiment, the process of acquiring the target historical illumination period includes:

[0032] Obtain the daily irradiation time period within a preset historical time period, wherein the irradiation time period is the time period in which the irradiance is greater than 0;

[0033] Obtain the average irradiance during the irradiation period;

[0034] Based on the daily average irradiance, the preset historical time periods are clustered to obtain several clusters, with each cluster representing a specific historical irradiance period.

[0035] This invention has the following beneficial effects: First, it analyzes the power supply aspect by analyzing relevant data of photovoltaic modules to obtain the power generation efficiency characteristics of the target historical sunshine period. Then, it analyzes the power generation demand aspect by combining the electricity price and photovoltaic power generation during different time periods of the day to obtain the power demand characteristics of the target historical sunshine period. By integrating these two aspects of data from the target historical sunshine period, the sensitivity of photoelectric conversion efficiency is obtained. The higher the sensitivity of photoelectric conversion efficiency, the stronger the correlation with photovoltaic power generation efficiency. Therefore, the target historical sunshine period can be screened based on the sensitivity of photoelectric conversion efficiency to obtain the target historical sunshine period with the required sensitivity. Then, training sample data can be determined based on the screened target historical sunshine period, improving the accuracy of training sample data selection and thus improving the accuracy of photovoltaic module output power prediction. Attached Figure Description

[0036] Figure 1 This is a flowchart of the process for obtaining the target historical illumination period according to an embodiment of the present invention;

[0037] Figure 2 This is a graph showing the variation of output power and solar irradiance of a photovoltaic module provided in one embodiment of the present invention;

[0038] Figure 3 This is a flowchart of a photovoltaic module output power prediction method for evaluating the economic benefits of photovoltaic power plants, provided by an embodiment of the present invention.

[0039] Figure 4 This is a daily variation curve trend of the output power of a photovoltaic module provided in different seasons according to an embodiment of the present invention;

[0040] Figure 5 This is a flowchart illustrating the positive impact of irradiation conditions on the output power of photovoltaic modules, provided in one embodiment of the present invention.

[0041] Figure 6 This is a flowchart illustrating the positive impact of environmental factors on the output power of photovoltaic modules, provided in one embodiment of the present invention.

[0042] Figure 7 This is a scatter plot of the daily temperature range and the output power of photovoltaic modules provided in one embodiment of the present invention;

[0043] Figure 8 This is a flowchart illustrating the acquisition of power generation efficiency characteristics provided in one embodiment of the present invention;

[0044] Figure 9 This is a flowchart illustrating the process of obtaining electrical energy demand characteristics according to an embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. All data and information collected in this application have been obtained with full consent.

[0047] This embodiment provides a method for predicting the output power of photovoltaic (PV) modules for assessing the economic benefits of PV power plants. The main idea is to analyze the historical output power of PV modules, consider meteorological interference factors, and predict the output power of PV modules based on the power generation efficiency of the PV power plant. To predict the output power of PV modules in a PV power plant, it is necessary to analyze the historical input, output, and other influencing data of the PV modules and evaluate them in conjunction with real-time monitoring data. For the process of predicting the output power of PV modules, the economic benefits of a PV power plant are mainly reflected in its power generation efficiency.

[0048] First, it's necessary to collect historical, multi-dimensional data related to the photovoltaic (PV) modules, such as irradiance and PV module panel temperature, as well as environmental data, such as ambient temperature, atmospheric particulate matter (TPM) concentration, ambient wind speed, and air dew point temperature. It should be understood that each of these data dimensions can be detected by its corresponding sensor. For example, an irradiance sensor placed next to the PV module can detect the solar irradiance. Specifically, the irradiance sensor can be a thermopile sensor, which works by using the thermoelectric effect to convert absorbed solar radiation into heat energy, then generating a micro-voltage signal through a thermopile (a series thermocouple). The voltage magnitude is proportional to the radiation intensity. The PV module panel temperature can be detected by a temperature sensor placed on the PV module panel. Ambient temperature can be detected by a temperature sensor placed next to the PV module. Atmospheric particulate matter (TPM) concentration can be specifically measured as TSP (Total Suspended Particulates) concentration, detected by a TSP concentration monitor. Ambient wind speed can be detected by a wind speed sensor placed next to the PV module. It should be understood that all sensors have the same sampling frequency and are sampled synchronously. The sampling frequency can be set according to actual needs, such as once every 10 seconds.

[0049] This embodiment pre-defines a historical time period, which includes multiple days. It should be understood that, in order to ensure the reliability of the target historical illumination period selection, the number of days included in the historical time period can be larger, and the specific number of days is set according to actual needs.

[0050] It should be understood that the time-series data of each dimension obtained in this embodiment can be normalized to eliminate the influence of dimensions and facilitate subsequent data processing. The normalization method can be a maximum / minimum value normalization method. For any dimension of time-series data, such as ambient temperature, the ambient temperature values ​​at various times within a historical time period are obtained, the maximum and minimum ambient temperature values ​​are determined, and the maximum / minimum value normalization method is used to normalize the ambient temperature values ​​at various times within the historical time period.

[0051] The output power of photovoltaic (PV) modules is a multivariate time-series problem with strong time dependence. Influencing factors (such as irradiance, temperature, TSP concentration, and cloud cover) exhibit significant non-stationarity and periodic fluctuations, with power output patterns differing across time periods. Directly analyzing all historical data, with its mixture of data under varying illumination and weather conditions, can easily obscure local patterns, affecting prediction stability and the accuracy of power generation efficiency assessments. Therefore, it is necessary to segment historical time periods, taking into account the influencing factors of PV module output power.

[0052] The output power of photovoltaic (PV) modules is primarily influenced by sunlight conditions. Therefore, historical time periods can be segmented based on irradiance data. Considering the significant differences in sunlight conditions across seasons and at different times of the day within the same season, the historical time periods are first coarsely segmented based on the overall irradiance levels of different dates, resulting in multiple target historical sunlight periods, also known as coarse-grained periods. Then, to differentiate between different sunlight conditions such as seasons and weather, fine-grained segmentation is performed within each coarse-grained period based on the intraday irradiance variation trend, further characterizing the daily variation pattern of PV module output power. This segmentation method helps improve the adaptability and prediction accuracy of power prediction models to complex sunlight variation scenarios.

[0053] In one exemplary embodiment, such as Figure 1 As shown, the following is a specific process for obtaining the target's historical illumination period:

[0054] Step S100: Obtain the daily irradiance time period in the preset historical time period. The irradiance time period is the time period when the irradiance is greater than 0.

[0055] Obtain the daily irradiance curves for a historical time period. Irradiance is 0 in the absence of light (nighttime) and greater than 0 in the presence of light (daytime). Extract the time periods with irradiance greater than 0 from the daily irradiance curves, defining these as the daily irradiance period (which can be understood as daytime within a day). The irradiance period includes irradiance data from multiple points in time.

[0056] Step S200: Obtain the average irradiance over the irradiation period.

[0057] For any given day, calculate the average irradiance of each irradiance period during that day, and use this average as the irradiance mean for that day's irradiance period. This yields the irradiance mean for each day's irradiance period.

[0058] Step S300: Based on the daily average irradiance, cluster the preset historical time periods to obtain several clusters, with each cluster representing a target historical irradiance period.

[0059] Based on the average daily irradiance over a historical time period, a two-dimensional array is created by combining the date and the average irradiance of each day. The Euclidean distance between any two two-dimensional arrays is used as the clustering relationship to cluster all days within a preset historical time period. This embodiment employs the K-means clustering algorithm. The value of the clustering parameter k can be set manually or obtained using the elbow method. This results in k clusters, each representing a specific historical irradiance period, i.e., a coarse-grained period.

[0060] By segmenting historical time periods using average irradiance, the daily irradiance due to seasonal or meteorological factors within the same target historical irradiance period becomes similar. Therefore, the output power difference of photovoltaic modules within the same target historical irradiance period is affected by other secondary factors. Thus, each target historical irradiance period can be further segmented into finer granular segments based on the output power of photovoltaic modules. Figure 2 This is a curve showing the changes in output power and solar irradiance of a photovoltaic module. The horizontal axis represents time, the left vertical axis represents the output power of the photovoltaic module (in MW), and the right vertical axis represents irradiance (in MJ / m²). 2 ).

[0061] In an exemplary embodiment, the photovoltaic module output power prediction method provided in this embodiment further includes the following fine-grained segmentation process: For any day in any target historical sunshine period, the photovoltaic module output power curve for that day is obtained. The photovoltaic module output power curve is obtained by curve fitting of the photovoltaic module output power at each moment, for example, by using the least squares method for curve fitting. The photovoltaic module output power at each moment can be detected by a power detection device installed at the power output terminal of the photovoltaic module. It should be understood that this embodiment normalizes the photovoltaic module output power at each moment. Specifically, the maximum and minimum values ​​of the photovoltaic module output power at each moment within the historical time period are obtained, and then the maximum and minimum value normalization method is used to normalize the photovoltaic module output power at each moment. The photovoltaic module output power involved in this embodiment is the normalized result.

[0062] The process involves identifying inflection points in the photovoltaic (PV) module's output power curve. An inflection point is the location where the curvature changes, representing a significant shift in the rate of change of the PV module's output power, thus indicating a transition in the PV module's operating state. The corresponding timestamps for each inflection point are also acquired. Finally, these timestamps are used as dividing points to segment the day's irradiance period into several core time periods. This yields several core time periods for each day within the target historical irradiance period. Each core time period is then used as a fine-grained time period for that day.

[0063] like Figure 3 As shown in the figure, the photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants provided in this embodiment includes the following steps:

[0064] Step S1: Combine the daily irradiance during the target historical sunshine period with the positive impact of environmental factors on the output power of photovoltaic modules to obtain the daily energy conversion efficiency characteristics;

[0065] Step S2: Based on the difference between the daily air dew point temperature and the photovoltaic module panel temperature, combined with the daily temperature range and energy conversion efficiency characteristics, the power generation efficiency characteristics of the target historical sunshine period are obtained.

[0066] Step S3: Based on the electricity price and photovoltaic power generation during each electricity price period of the day, obtain the characteristics of the electricity demand during the target historical sunshine period;

[0067] Step S4: Integrate the power generation efficiency characteristics and power demand characteristics of the target historical illumination period to obtain the sensitivity of photoelectric conversion efficiency during the target historical illumination period.

[0068] The following is a detailed explanation of each step.

[0069] Step S1: Combine the daily irradiance during the target historical sunshine period with the positive impact of environmental factors on the output power of photovoltaic modules to obtain the daily energy conversion efficiency characteristics.

[0070] The economic benefits assessment of photovoltaic (PV) power plants are influenced by both electricity supply and demand. PV module output power prediction primarily targets the electricity supply side. On the supply side, influencing factors include the energy conversion efficiency characteristics of PV modules, environmental meteorological conditions, and the operating status of the PV modules. Among these, energy conversion efficiency, characterizing photoelectric conversion efficiency, is the core indicator measuring the ability of PV modules to convert incident light energy into electrical energy, directly affecting the power generation capacity per unit area. Therefore, it is necessary to assess the energy conversion efficiency characteristics of PV modules daily during the target historical sunshine period.

[0071] The positive impact of daily irradiance during the target historical sunshine period on the output power of photovoltaic modules, as well as the positive impact of environmental factors on the output power of photovoltaic modules, are obtained by combining these two positive impacts to obtain the daily energy conversion efficiency characteristics.

[0072] Photovoltaic modules (taking mainstream crystalline silicon modules as an example) are based on the photoelectric effect, converting photon energy into electron-hole pairs. Different wavelengths of light produce different photon energies, resulting in significant differences in the output power of photovoltaic modules. It is worth noting that, under the same solar irradiance, if the spectral distribution is biased towards the longer wavelength (800-1100 nm) region, the average photon energy per unit irradiance decreases. In some wavelengths, photons cannot effectively excite electron-hole pairs, leading to a decline in the actual photoelectric conversion efficiency of the photovoltaic module.

[0073] In cloudy or hazy weather, suspended particulate matter in the atmosphere (such as dust, aerosols, haze, and water vapor particles) scatters mid- to long-wavelength (500-1000 nm) light, causing the overall spectrum to smooth out and reducing the proportion of short-wavelength light. With a fixed total irradiance, fewer short-wavelength photons and fewer high-energy photons result in lower photoelectric conversion efficiency.

[0074] Furthermore, while the irradiance on a photovoltaic panel may be the same at different solar altitude angles, the atmospheric scattering of sunlight varies, leading to differences in conversion efficiency. For example, at lower solar altitude angles, blue / ultraviolet light is strongly scattered and absorbed by the atmosphere, while the infrared proportion increases, resulting in a significantly lower average effective usable photon energy compared to scenarios with higher solar altitude angles. Therefore, analysis considering the solar altitude angle at different times is necessary. Figure 4 The figure shows the daily variation curves of the output power of photovoltaic modules in different seasons. The horizontal axis represents time, and the vertical axis represents the output power of the photovoltaic modules (in MW).

[0075] like Figure 5 As shown, the following is a specific process for obtaining the positive impact of daily irradiance on the output power of photovoltaic modules:

[0076] Step S1-1: Based on the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power at each moment in the core time period, obtain the power characteristics at each moment.

[0077] For any core time period within any day of the target historical sunshine period, obtain the solar altitude angle at each moment within that core time period. The solar altitude angle is related to each moment of the day; it varies at different times. The method for obtaining the solar altitude angle is existing technology and will not be elaborated further. In an exemplary embodiment, the solar altitude angle is also normalized. Specifically, the maximum and minimum values ​​of the solar altitude angle at each moment within the historical time period are obtained, and then the maximum and minimum value normalization method is used to normalize the solar altitude angle at each moment. The solar altitude angles involved in subsequent data processing are all normalized solar altitude angles. The larger the solar altitude angle, the greater the solar radiance.

[0078] The power characteristics at each moment are obtained based on the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power. As analyzed above, a larger solar altitude angle corresponds to greater solar irradiance, which in turn requires a larger photovoltaic module output power, resulting in a larger power characteristic. Conversely, a higher atmospheric particulate matter concentration corresponds to lower solar irradiance, which in turn requires a smaller photovoltaic module output power, resulting in a smaller power characteristic. Irradiance and photovoltaic module output power directly characterize the power characteristics. Therefore, the power characteristics are directly proportional to the solar altitude angle, irradiance, and photovoltaic module output power, and inversely proportional to the atmospheric particulate matter concentration.

[0079] In one exemplary embodiment, a method for quantifying power characteristics is given below:

[0080] ;

[0081] in, This represents the power characteristic at time j. Represents the irradiance at time j. This represents the solar altitude angle at time j. This represents the output power of the photovoltaic module at time j. This represents the concentration of atmospheric particulate matter at time j. It should be understood that all input parameters in this formula are normalized parameters.

[0082] Step S1-2: Integrate the power characteristics at each time point, and combine the correlation between the irradiance and the output power of the photovoltaic module during the core time period to obtain the first positive impact indicator of the core time period.

[0083] The irradiance curve and the photovoltaic module output power curve for the core time period are obtained. Since the photovoltaic module output power normally changes with irradiance, the stronger the correlation between the two curves, the higher the photoelectric conversion efficiency. Therefore, the correlation between the irradiance curve and the photovoltaic module output power curve for the core time period is obtained. In an exemplary embodiment, the Pearson correlation coefficient between the irradiance curve and the photovoltaic module output power curve for the core time period is obtained, and then the Pearson correlation coefficient is normalized to obtain the correlation. It should be understood that since the Pearson correlation coefficient ranges from [-1, 1], the normalization method is: (Pearson correlation coefficient + 1) / 2.

[0084] The power characteristics of each moment in the core time period are integrated. Specifically, the average power characteristics of each moment in the core time period are calculated and defined as the average power characteristics of the core time period.

[0085] The product of the correlation between irradiance and photovoltaic module output power during the core time period and the average power characteristic is calculated as the first positive impact indicator for the core time period. This first positive impact indicator characterizes the positive influence of irradiance conditions during the core time period on the photovoltaic module output power, specifically the positive impact of the light wavelength distribution on the photovoltaic module output power. This yields the first positive impact indicator for each core time period within each day of the target historical irradiance period.

[0086] The actual power generation efficiency of photovoltaic (PV) modules is also affected by several environmental factors. In this embodiment, these factors include atmospheric particulate matter concentration, cloud cover, and wind speed. Increased irradiance may lead to a rise in the temperature of the PV module panel. When the panel temperature increases, the semiconductor bandgap decreases, resulting in a decrease in energy conversion efficiency. Atmospheric particulate matter concentration and cloud cover affect the irradiance reaching the PV module; higher concentrations and cloud cover result in lower irradiance, impacting energy conversion efficiency. Furthermore, higher wind speeds can improve heat dissipation of the PV module to some extent, thereby increasing energy conversion efficiency.

[0087] Therefore, by combining the above factors, the positive impact of daily environmental factors on the output power of photovoltaic modules is determined. In an exemplary embodiment, such as... Figure 6 As shown, the following is a specific process for obtaining information on the positive impact of environmental factors on the output power of photovoltaic modules:

[0088] Steps S1-3: Determine the ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at each moment during the core time period.

[0089] For any given moment within a given core time period, acquire the ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature. Cloud cover is obtained from meteorological data. The cloud cover value ranges from 0% to 100%.

[0090] Steps S1-4: Integrate ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at each time point to obtain the second positive impact indicator for the core time period.

[0091] First, based on the ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at that moment, a positive impact sub-indicator is obtained for that moment. This positive impact sub-indicator is directly proportional to the ambient wind speed at that moment and inversely proportional to the atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at that moment. In an exemplary embodiment, a specific quantification method for the positive impact sub-indicator is given below:

[0092] ;

[0093] in, This represents the sub-indicator of positive impact at time j. This represents the ambient wind speed at time j. This represents the temperature of the photovoltaic module panel at time j. This represents the cloud cover at time j. It should be understood that all input parameters in this formula are normalized.

[0094] The positive impact sub-indicators at each moment within the core time period are integrated. Specifically, the average of the positive impact sub-indicators at each moment within the core time period is calculated as the second positive impact indicator for the core time period. The second positive impact indicator characterizes the impact of environmental factors during the core time period on the output power of photovoltaic modules. The larger the second positive impact indicator, the more positive the impact of environmental conditions on the output power of photovoltaic modules.

[0095] By integrating the first and second positive impact indicators of the core time period, the resulting energy conversion efficiency characteristics of the core time period are presented as follows:

[0096] ;

[0097] in, This indicates the energy conversion efficiency characteristics during the core time period. The primary indicator of positive impact during the core time period. This represents the second positive impact indicator for the core time period. The calculation method essentially combines the first and second positive impact indicators with a weighted sum, each with a weight of 0.5.

[0098] Using the above method, the energy conversion efficiency characteristics of each core time period in each day of the target historical sunshine period are obtained. The larger the value, the higher the energy conversion efficiency of the photovoltaic module during the core time period.

[0099] Step S2: Based on the difference between the daily air dew point temperature and the photovoltaic module panel temperature, combined with the daily temperature range and energy conversion efficiency characteristics, the power generation efficiency characteristics of the target historical sunshine period are obtained.

[0100] In addition to analysis at the core time period scale, assessing the power generation efficiency of a specific core time period also requires analyzing the energy conversion characteristics over a longer time scale using monitoring data for that day. Among these analyses, the output power of photovoltaic (PV) modules shows a significant correlation with the diurnal temperature range. Large diurnal temperature ranges are typically accompanied by clear skies and few clouds, which is conducive to the Earth's surface receiving high-intensity solar radiation with a reasonable wavelength distribution, thereby increasing the output power of PV modules. Furthermore, under climatic conditions with large diurnal temperature ranges, the operating temperature of PV modules fluctuates significantly, modulating the overall daily output power curve. For example... Figure 7 The image shows a scatter plot of the relationship between diurnal temperature range and photovoltaic module output power. The horizontal axis represents the diurnal temperature range (in degrees Celsius), and the vertical axis represents the photovoltaic module output power (in MW). Figure 7 The straight line in the figure is the fitted line.

[0101] To facilitate processing, the maximum and minimum values ​​of the daily temperature range for each day in the historical time period are obtained. The daily temperature range for each day in the historical time period is normalized using the maximum and minimum value normalization method. The daily temperature ranges mentioned below are the normalized daily temperature ranges.

[0102] However, when there is a large daily temperature range and the air humidity is high, if the surface temperature of the photovoltaic module drops below the air dew point temperature, moisture in the air may condense on the surface of the photovoltaic module, forming a continuous water film that covers the sensor surface. This can lead to a decrease in transmittance due to impurity scattering or absorption by the water film, reducing the effective light-illuminated area and lowering the utilization rate of the photovoltaic module for these photons.

[0103] like Figure 8 As shown, the following is a specific process for obtaining power generation efficiency characteristics:

[0104] Step S2-1: Based on the energy conversion efficiency characteristics of the core time period and the difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period, the energy conversion factor for the core time period is obtained.

[0105] The difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period is determined and defined as the temperature difference feature. The process for obtaining the temperature difference feature is as follows: It should be noted that in this data processing, because the difference between the air dew point temperature and the photovoltaic module panel temperature needs to be calculated, the air dew point temperature and the photovoltaic module panel temperature here are not normalized and are the actual temperatures before normalization. The air dew point temperature at each moment during the core time period is obtained. The air dew point temperature can be obtained using specialized air dew point temperature detection equipment. The average air dew point temperature at each moment during the core time period is calculated to obtain the mean air dew point temperature for the core time period. Similarly, the photovoltaic module panel temperature at each moment during the core time period is obtained, and then the average value is calculated to obtain the mean photovoltaic module panel temperature for the core time period. The average air dew point temperature during the core time period is subtracted from the average photovoltaic module panel temperature to obtain the difference. If this difference is greater than 0, it indicates that the average air dew point temperature during the core time period is greater than the average photovoltaic module panel temperature. In this case, moisture droplets will condense on the photovoltaic module surface, forming a continuous water film that affects energy conversion efficiency. This temperature difference characteristic is the first temperature difference characteristic, which is directly proportional to the difference between the average air dew point temperature and the average photovoltaic module panel temperature; the larger the difference, the larger the first temperature difference characteristic. If the difference is less than or equal to 0, meaning the average air dew point temperature during the core time period is less than or equal to the average photovoltaic module panel temperature, then the temperature difference characteristic is a set value less than the first temperature difference characteristic. A specific quantification method is given below:

[0106] ;

[0107] Where x represents the difference between the average air dew point temperature and the average temperature of the photovoltaic module panel during the core time period. This represents the temperature difference characteristic. Therefore, the value set for the condition less than the first temperature difference characteristic is 1.

[0108] The energy conversion factor for the core time period is derived from the energy conversion efficiency characteristics and temperature difference characteristics of that period. A higher energy conversion efficiency characteristic results in a higher energy conversion factor; a smaller temperature difference characteristic indicates a lower degree of continuous water film formation on the photovoltaic module surface, thus having a lower impact on energy conversion efficiency, or even no continuous water film formation at all, meaning it has no impact on energy conversion efficiency, resulting in a higher energy conversion factor. Therefore, the energy conversion factor is directly proportional to the energy conversion efficiency characteristic and inversely proportional to the temperature difference characteristic. A specific quantification method is given below:

[0109] ;

[0110] in, This represents the energy conversion factor for the i-th core time period within the v-th day of the target historical illumination period. This represents the energy conversion efficiency characteristic of the i-th core time period in day v. This represents the difference between the average air dew point temperature and the average temperature of the photovoltaic module panel during the i-th core time period on day v. This represents the temperature difference characteristics of the i-th core time period in day v.

[0111] Step S2-2: Integrate the energy conversion factors of each core time period of the day to obtain the daily power generation efficiency sub-characteristic.

[0112] The average energy conversion factor of each core time period within day v of the target historical sunshine period is calculated to obtain the power generation efficiency sub-characteristic of day v of the target historical sunshine period. This leads to the daily power generation efficiency sub-characteristic of the target historical sunshine period.

[0113] Step S2-3: Integrate the daily power generation efficiency sub-features and the daily temperature range to obtain the power generation efficiency features for the target historical sunshine period.

[0114] The larger the daily power generation efficiency sub-feature, the larger the power generation efficiency feature of the target historical sunshine period; the larger the daily temperature range, the larger the power generation efficiency feature of the target historical sunshine period. Therefore, by fusing the daily power generation efficiency sub-feature and the daily temperature range, the power generation efficiency feature of the target historical sunshine period is obtained, and the power generation efficiency feature is directly proportional to both the power generation efficiency sub-feature and the daily temperature range. In an exemplary embodiment, a specific quantification method is given below:

[0115] ;

[0116] in, This represents the power generation efficiency characteristics of the target historical sunshine period, where u represents the number of days included in the target historical sunshine period. This represents the diurnal temperature range on day v during the target's historical sunshine period. This represents the power generation efficiency sub-characteristic for day v of the target historical sunshine period. The higher the power generation efficiency characteristic for the target historical sunshine period, the higher the power generation efficiency for the target historical sunshine period, and thus the higher the photoelectric conversion efficiency of the photovoltaic modules in the photovoltaic power station.

[0117] Step S3: Based on the electricity price and photovoltaic power generation during each electricity price period of the day, obtain the characteristics of the electricity demand during the target historical sunshine period.

[0118] On the demand side, it is necessary to analyze the output characteristics of different electricity price periods in historical timeframes. Based on the electricity price and photovoltaic power generation during each electricity price period of the day, the characteristics of electricity demand during the target historical sunshine period can be obtained.

[0119] In one exemplary embodiment, such as Figure 9 As shown, the following is a specific process for obtaining the characteristics of electricity demand:

[0120] Step S3-1: Determine the proportion of photovoltaic power generation in each electricity price period of the day, and combine it with the electricity price in each electricity price period of the day to obtain the electricity demand in each electricity price period of the day.

[0121] For any day within the target historical sunshine period, obtain the electricity price periods for that day, along with the electricity price for each period. Specifically, consecutive moments with the same electricity price within a day (essentially the sunshine period of a day) constitute the electricity price periods, with different prices for different periods. It should be understood that the maximum and minimum electricity prices for each period within the historical period are obtained, and then normalized using a maximum / minimum value normalization method. Subsequent electricity prices refer to the normalized prices.

[0122] Obtain the photovoltaic power generation for each electricity price period, and the total photovoltaic power generation for that day (essentially the total photovoltaic power generation during the irradiation period of that day). Calculate the ratio of photovoltaic power generation for each electricity price period to the total photovoltaic power generation for that day, and use this ratio as the proportion of photovoltaic power generation for each electricity price period on that day.

[0123] Based on the proportion of photovoltaic power generation in each electricity price period of the day and the electricity price in each electricity price period of the day, the time-based electricity demand for each electricity price period is obtained. The time-based electricity demand is directly proportional to both the proportion of photovoltaic power generation and the electricity price. In an exemplary embodiment, the product of the proportion of photovoltaic power generation in each electricity price period and the electricity price is taken as the time-based electricity demand for that electricity price period.

[0124] Step S3-2: Integrate the time-of-day electricity demand for each electricity price period during the target historical sunshine period to obtain the electricity demand characteristics of the target historical sunshine period.

[0125] For any day within the target historical sunshine period, calculate the average of the electricity demand for each electricity price period on that day to obtain the daily electricity demand for that day.

[0126] Then, the average daily electricity demand during the target historical sunshine period is calculated, and the result is the electricity demand characteristic of the target historical sunshine period.

[0127] Step S4: Integrate the power generation efficiency characteristics and power demand characteristics of the target historical illumination period to obtain the sensitivity of photoelectric conversion efficiency during the target historical illumination period.

[0128] The power generation efficiency characteristics during the target historical sunshine period are analyzed from the perspective of electricity supply, while the electricity demand characteristics during the target historical sunshine period are analyzed from the perspective of power generation demand. By integrating the data from these two aspects, the sensitivity of photoelectric conversion efficiency during the target historical sunshine period is obtained.

[0129] In an exemplary embodiment, the power generation efficiency characteristics and power demand characteristics of the target historical sunshine period are weighted and summed, with each weight being 0.5. The result is the sensitivity of photoelectric conversion efficiency of the target historical sunshine period. Therefore, the essence is to calculate the average value of the power generation efficiency characteristics and power demand characteristics of the target historical sunshine period as the sensitivity of photoelectric conversion efficiency of the target historical sunshine period.

[0130] Using the above process, the sensitivity of photoelectric conversion efficiency for each target's historical illumination period is obtained. The higher the sensitivity of photoelectric conversion efficiency, the stronger the correlation with photovoltaic power generation efficiency. Therefore, the target's historical illumination period can be screened based on the sensitivity of photoelectric conversion efficiency for each target's historical illumination period.

[0131] In one exemplary embodiment, the photoelectric conversion efficiency sensitivity of each target historical illumination period is selected, and the target historical illumination periods corresponding to the photoelectric conversion efficiency sensitivity that meet preset conditions are retained. In one exemplary embodiment, a preset photoelectric conversion efficiency sensitivity threshold is used to filter target historical illumination periods with higher photoelectric conversion efficiency sensitivity. The value range of the photoelectric conversion efficiency sensitivity threshold is 0-1, and the specific value is set according to actual judgment needs. If the screening of target historical illumination periods is more stringent, the photoelectric conversion efficiency sensitivity threshold can be set larger. In this embodiment, 0.5 is used as an example. The photoelectric conversion efficiency sensitivity of each target historical illumination period is compared with the photoelectric conversion efficiency sensitivity threshold, and the target historical illumination periods corresponding to the photoelectric conversion efficiency sensitivity greater than the photoelectric conversion efficiency sensitivity threshold are retained. The retained target historical illumination periods are defined as sample historical illumination periods.

[0132] The photovoltaic (PV) data from the historical sunshine duration is used as training data for the model. It should be understood that the specific selection of PV data from the historical sunshine duration is determined by the actual training needs. For example, the input might be environmental data from the historical sunshine duration (including ambient wind speed, irradiance, atmospheric particulate matter concentration, PV module panel temperature, air dew point temperature, cloud cover, etc.), and the output might be the PV module output power during the historical sunshine duration. This results in a large amount of training data. This training data is used to train the machine learning model to predict the PV module output power for the next time period, such as the output power for the shortest target historical sunshine duration. The model training process is as follows: The training sample data is divided into a training set (70%), a validation set (15%), and a test set (15%). The training set is input into the XGBoost model for training: the initial predicted value is set to the average value; the current predicted residual is calculated; a regression tree is fitted to the residual (each tree fits the residual); the predicted value is updated; this iteration is repeated until the residual decreases or the maximum number of trees is reached; RMSE is used as the evaluation metric for model iteration: RMSE changes are monitored on the validation set; if the validation set error does not decrease for several consecutive rounds, training is stopped to avoid overfitting; the model with the smallest validation set RMSE is retained, completing the model training.

[0133] It should be understood that this invention is not limited to the specific selection of photovoltaic data from historical illumination periods of the samples; the data information included can be determined according to actual training needs. Furthermore, this invention is not limited to the network model used or the specific training process; implementers can flexibly choose according to actual needs.

[0134] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0135] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for predicting the output power of photovoltaic modules for economic benefit assessment of photovoltaic power plants, characterized in that, include: By combining the daily irradiance during the target historical sunshine period and the positive impact of environmental factors on the output power of photovoltaic modules, the daily energy conversion efficiency characteristics are obtained. Based on the difference between the daily air dew point temperature and the photovoltaic module panel temperature, combined with the daily temperature range and the energy conversion efficiency characteristics, the power generation efficiency characteristics of the target historical sunshine period are obtained. Based on the electricity price and photovoltaic power generation during each electricity price period of the day, the characteristics of the electricity demand during the target historical sunshine period are obtained; By integrating the power generation efficiency characteristics and power demand characteristics of the target historical sunshine period, the photoelectric conversion efficiency sensitivity of the target historical sunshine period is obtained. The photoelectric conversion efficiency sensitivity is used to indicate the screening of the target historical sunshine period. The process of obtaining the target historical illumination period includes: Obtain the daily irradiation time period within a preset historical time period, wherein the irradiation time period is the time period in which the irradiance is greater than 0; Obtain the average irradiance during the irradiation period; Based on the daily average irradiance, the preset historical time periods are clustered to obtain several clusters, and each cluster represents a target historical irradiance period. The photovoltaic module output power prediction method also includes: Obtain the inflection point in the daily output power curve of photovoltaic modules; Based on the time corresponding to the inflection point, each day is divided into several core time periods; The daily energy conversion efficiency characteristics include the energy conversion efficiency characteristics of each core time period within each day; The process of obtaining the power generation efficiency characteristics includes: Based on the energy conversion efficiency characteristics of the core time period and the difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period, the energy conversion factor for the core time period is obtained. By integrating the energy conversion factors of each core time period of the day, the daily power generation efficiency sub-characteristics are obtained; By integrating the daily power generation efficiency sub-features and the daily temperature range, the power generation efficiency characteristics of the target historical sunshine period are obtained; the power generation efficiency characteristics are both proportional to the power generation efficiency sub-features and the daily temperature range.

2. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 1, characterized in that, The process of obtaining the positive impact of daily irradiance on the output power of photovoltaic modules includes: Determine the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power at each moment during the core time period; The power characteristics at each moment are obtained based on the solar altitude angle, atmospheric particulate matter concentration, irradiance, and photovoltaic module output power. The power characteristics are directly proportional to the solar altitude angle, irradiance, and photovoltaic module output power, and inversely proportional to the atmospheric particulate matter concentration. By integrating the power characteristics at various times and combining the correlation between irradiance and photovoltaic module output power during the core time period, a first positive impact index for the core time period is obtained; the first positive impact index characterizes the positive impact of irradiance during the core time period on photovoltaic module output power.

3. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 2, characterized in that, The correlation between the change in irradiance and the output power of the photovoltaic module is the Pearson correlation coefficient between irradiance and the output power of the photovoltaic module.

4. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 1, characterized in that, The process of obtaining the positive impact of daily environmental factors on the output power of photovoltaic modules includes: Determine the ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at each moment during the core time period. By integrating ambient wind speed, atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature at various times, a second positive impact index for the core time period is obtained. The second positive impact index is directly proportional to ambient wind speed and inversely proportional to atmospheric particulate matter concentration, cloud cover, and photovoltaic module panel temperature. The second positive impact index characterizes the impact of environmental factors on the output power of photovoltaic modules during the core time period.

5. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 1, characterized in that, The difference between the air dew point temperature and the photovoltaic module panel temperature during the core time period is a temperature difference feature. If the average air dew point temperature during the core time period is greater than the average photovoltaic module panel temperature, then the temperature difference feature is a first temperature difference feature, which is proportional to the difference between the average air dew point temperature and the average photovoltaic module panel temperature. If the average air dew point temperature during the core time period is less than or equal to the average photovoltaic module panel temperature, then the temperature difference feature is a set value that is less than the first temperature difference feature. The energy conversion factor is obtained from the energy conversion efficiency characteristics and temperature difference characteristics of the core time period. The energy conversion factor is directly proportional to the energy conversion efficiency characteristics and inversely proportional to the temperature difference characteristics.

6. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 1, characterized in that, The process of obtaining the characteristics of the degree of electricity demand includes: The proportion of photovoltaic power generation in each electricity price period of the day is determined, and the electricity price in each electricity price period of the day is combined to obtain the electricity demand in each electricity price period of the day; the electricity demand in each period is directly proportional to both the proportion of photovoltaic power generation and the electricity price. By integrating the electricity demand for each electricity price period during the day within the target historical sunshine period, the characteristics of the electricity demand level during the target historical sunshine period are obtained.

7. The photovoltaic module output power prediction method for economic benefit assessment of photovoltaic power plants as described in claim 1, characterized in that, The photovoltaic module output power prediction method further includes: obtaining the photoelectric conversion efficiency sensitivity of multiple target historical illumination periods, and retaining the target historical illumination periods corresponding to the photoelectric conversion efficiency sensitivity that meet preset conditions as sample historical illumination periods.

Images