A photovoltaic power generation prediction method and system based on the influence of snow

Abstract

The application relates to the field of information technology and discloses a photovoltaic power generation prediction method and system based on snow influence, which constructs a first historical constant-temperature aging efficiency curve by using historical power generation related data of photovoltaic power generation equipment in a target area under snow conditions, determines a snow state change area in the curve, determines first historical photovoltaic conversion efficiency data of the area based on snow distribution uniformity, determines equipment state data and snow adhesion intensity data of the photovoltaic power generation equipment based on the first historical photovoltaic conversion efficiency data and the snow state change area, further constructs a second historical constant-temperature aging efficiency curve, combines the obtained curve with corresponding historical temperature and humidity data, further determines second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment to calculate a power prediction value of the photovoltaic power generation equipment in a future preset period, accurately evaluates the influence of snow on photovoltaic power generation efficiency, and further accurately predicts the power of the photovoltaic power generation equipment under the influence of snow.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, in particular to a photovoltaic power generation prediction method and system based on snow influence. BACKGROUND

[0002] Since the photovoltaic power station is located outdoors, its power generation is easily affected by weather, especially rain, snow and ice. The accumulation of snow on the photovoltaic panel will significantly reduce the power generation of the photovoltaic component and seriously affect the power supply guarantee of the power grid. At present, the power generation prediction of photovoltaic components under snow cover has not received enough attention and in-depth research. Although a small amount of existing research has analyzed the influence of snow on the power generation of photovoltaic components from a macro perspective, these researches often require a large amount of data, and the accuracy of the prediction results is limited.

[0003] Therefore, how to improve the prediction result of the power generation of the photovoltaic component has become a technical problem to be solved by those skilled in the art. SUMMARY

[0004] The present application provides a photovoltaic power generation prediction method and system based on snow influence, which solves the problem of how to improve the prediction result of the power generation of the photovoltaic component.

[0005] To solve the above technical problems, the present application provides a photovoltaic power generation prediction method based on snow influence, comprising:

[0006] Obtaining historical power generation data, historical temperature and humidity data and historical light radiation intensity data of photovoltaic power generation equipment in a target area under snow conditions to construct a first historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snow conditions;

[0007] Determining a snow state change area in the first historical constant temperature aging efficiency curve, and determining first historical photovoltaic conversion efficiency data of the snow state change area based on snow distribution uniformity;

[0008] Based on the first historical photovoltaic conversion efficiency data, determining equipment state data and snow adhesion intensity data corresponding to the snow state change area to construct a second historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snow conditions;

[0009] According to the second historical constant temperature aging efficiency curve and its corresponding historical temperature and humidity data, determining second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment, and calculating a power prediction value of the photovoltaic power generation equipment in a future preset period according to the second historical photovoltaic conversion efficiency data to output.

[0010] The present application provides a photovoltaic power generation prediction system based on snow influence, comprising:

[0011] a first curve construction module, configured to acquire historical power generation data, historical temperature and humidity data and historical light radiation intensity data of a photovoltaic power generation device in a target area in a snow day condition, to construct a first historical constant-temperature aging efficiency curve of the photovoltaic power generation device in the snow day condition;

[0012] an efficiency data calculation module, configured to determine a snow accumulation state changing area in the first historical constant-temperature aging efficiency curve, and determine first historical photovoltaic conversion efficiency data of the snow accumulation state changing area based on snow accumulation uniformity;

[0013] a second curve construction module, configured to determine device state data and snow adhesion intensity data corresponding to the snow accumulation state changing area based on the first historical photovoltaic conversion efficiency data, to construct a second historical constant-temperature aging efficiency curve of the photovoltaic power generation device in the snow day condition;

[0014] a power generation power prediction module, configured to determine second historical photovoltaic conversion efficiency data of the photovoltaic power generation device according to the second historical constant-temperature aging efficiency curve and corresponding historical temperature and humidity data, and calculate a power prediction value of the photovoltaic power generation device in a future preset period according to the second historical photovoltaic conversion efficiency data to output.

[0015] Compared with the prior art, the beneficial effects of the embodiments of the present application are at least one of the following:

[0016] (1) The first historical constant-temperature aging efficiency curve is constructed by multi-factor coupling modeling to quantify the dynamic influence of snow coverage on photovoltaic conversion efficiency, thereby improving the efficiency evaluation precision; the superposition effect of device state and snow adhesion is considered to quantify the influence of snow on the power generation of the photovoltaic power generation device, thereby providing a basis for device selection and operation strategy optimization; the synergistic power prediction of multiple factors such as aging, snow, temperature and humidity is integrated to avoid the deviation of a single model;

[0017] (2) Through the whole-chain closed loop of data acquisition, efficiency modeling, state analysis and power prediction, the influence of snow on photovoltaic power generation efficiency can be accurately evaluated to provide effective support for photovoltaic power generation device management and power prediction in a snow day condition. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0019] Figure 1is a flow chart of a photovoltaic power generation prediction method based on the influence of accumulated snow provided by some embodiments of the present application;

[0020] Figure 2 is a structural diagram of a photovoltaic power generation prediction system based on the influence of accumulated snow provided by some embodiments of the present application;

[0021] Reference signs:

[0022] Wherein, 10, first curve construction module; 20, efficiency data calculation module; 30, second curve construction module; 40, power generation power prediction module. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings and embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. The purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0024] In the description of the present application, the terms "first", "second", "third" and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second", "third" and the like can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0025] In the description of the present application, it should be noted that, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting" should be understood in a broad sense, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication inside two elements. The terms "vertical", "horizontal", "left", "right", "up", "down" and similar expressions used in this paper are only for the purpose of description, and cannot be understood as indicating or implying that the system or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present application. The term "and / or" used in this paper includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0026] In the description of the present application, it is necessary to explain that, unless otherwise defined, all the technical and scientific terms used by the present application are the same as the meanings generally understood by the person skilled in the art. The terms used in the specification of the present application are only for describing specific embodiments, not intended to limit the present application, and the above-mentioned terms can be understood according to the specific meaning in the present application for the person skilled in the art.

[0027] In an embodiment, as shown in Figure 1 The first aspect of the present application provides a photovoltaic power generation prediction method based on the influence of snow, comprising:

[0028] S1, obtaining the historical power generation data, historical temperature and humidity data and historical light radiation intensity data of the photovoltaic power generation equipment in the target area under snow conditions, to construct the first historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snow conditions;

[0029] In an embodiment, the obtaining of the historical power generation data, historical temperature and humidity data and historical light radiation intensity data of the photovoltaic power generation equipment in the target area under snow conditions comprises:

[0030] Obtaining the first historical power generation data of the photovoltaic power generation equipment in the target area in the past preset period, and determining the period when the historical temperature and humidity data exceeds the preset temperature and humidity threshold and the falling rate of the historical light radiation intensity data reaches the preset falling rate threshold as the snow period; the first historical power generation data includes historical power generation data, historical temperature and humidity data and historical light radiation intensity data;

[0031] Taking the first historical power generation data of the snow period as the second historical power generation data, and determining the light attenuation degree data in the second historical power generation data by using the linear regression algorithm, to calculate the snow thickness data corresponding to the second historical power generation data according to the correlation model between the snow thickness and the light attenuation degree constructed in advance;

[0032] Based on the second historical power generation data, quantifying the distribution characteristics of the historical power generation data of the photovoltaic power generation equipment in the snow period, obtaining the difference between the devices, and using clustering algorithm to group the second historical power generation data and its corresponding snow thickness data, to obtain the characteristic description between the snow thickness and the light attenuation of the photovoltaic power generation equipment under the snow conditions.

[0033] Specifically, the target area in the present application can be a photovoltaic power generation base, or the range where the photovoltaic assembly installed in a personal home is located. The size of the target area is not limited, but the target area contains two or more photovoltaic power generation devices. The first historical power generation data (including historical power generation data, historical temperature and humidity data, and historical light radiation intensity data, historical photovoltaic power conversion data, etc., that is, operation data or weather data related to photovoltaic power generation) of the photovoltaic power generation device contained in the target area in the past preset period (such as one year) is obtained through database query. For example, assuming that a photovoltaic power station is deployed in a cold northern region, the database stores one year of power generation data, temperature data, and light radiation intensity data. The historical records of a certain month (such as December 2024) can be extracted from the database through SQL query, including hourly power generation (unit: kilowatt), temperature (unit: Celsius), and light radiation intensity (unit: watt / square meter) as the first historical power generation data, which provides a basis for subsequent analysis. Then, the period in which the historical light radiation intensity data exceeds the preset temperature and humidity threshold and the falling rate reaches the preset falling rate threshold is determined as the snow day period. The present application uses environmental characteristics to quickly screen the target period, laying a foundation for subsequent snow analysis.

[0034] Then, the first historical power generation data of the snow day period is taken as the second historical power generation data, that is, the historical power generation data, historical temperature and humidity data, and historical light radiation intensity data of the photovoltaic power generation device under snow day conditions. Based on the second historical power generation data, the light attenuation degree data in the second historical power generation data is calculated using a linear regression algorithm. For example, in a sunny day without snow, the power generation increases by 20 kilowatts when the light intensity increases by 100 watts / square meter. However, in a snowy day, the power generation only increases by 10 kilowatts under the same light increment, indicating that snow causes about 50% light attenuation. A linear regression model can be constructed according to this relationship to calculate the light attenuation degree data, and the historical snow thickness data corresponding to the historical snow day conditions can be calculated according to the pre-constructed correlation model between snow thickness and light attenuation degree. The mapping relationship (such as a quadratic function) between snow thickness and light attenuation degree can be established through experiments or historical data to obtain the correlation model, and the parameters in the model can be fitted and calibrated through the least squares method.

[0035] Finally, based on the second historical power generation data, the distribution characteristics of the power generation data during the snow period, such as the mean, variance, skewness, maximum power reduction rate, etc., are calculated to obtain the differences between devices, and the K-means or DBSCAN algorithm is used to group the second historical power generation data of the photovoltaic power generation device and the corresponding snow thickness data according to the differences between devices, to determine the correlation mode between each snow thickness data and the degree of light attenuation, and then generate a characteristic description of the snow thickness and light attenuation of the photovoltaic power generation device under snow conditions according to the correlation mode, wherein the characteristic description can be represented as: for every 1 cm increase in snow thickness, the light attenuation is about 20%, and the power reduction is 15%.

[0036] From snow determination, snow estimation to difference analysis, without manual intervention, the whole process is automatically processed, improving the data processing rate; through clustering grouping to realize differentiated strategy, avoiding "one-size-fits-all" management; considering the dynamic influence of device difference and snow thickness, the snow power prediction error can be significantly reduced; the present application quantifies the influence of snow on photovoltaic power generation through a data-driven method, providing a scientific basis for the operation and optimization of photovoltaic power stations.

[0037] In an embodiment, the constructing the first historical constant-temperature aging efficiency curve of the photovoltaic power generation device under the snow condition comprises:

[0038] According to the characteristic description, taking the light radiation intensity in the second historical power generation data and its corresponding snow thickness as independent variables, and taking the power generation power in the second historical power generation data as dependent variable, linear regression processing is performed by least square method to obtain trend data of power generation power change with light intensity;

[0039] Extracting the power generation efficiency data of the photovoltaic power generation device during the snow period, to quantify the correlation between power fluctuation and power generation efficiency based on the trend data by polynomial fitting method, to obtain a power generation efficiency model, and constructing a power generation efficiency curve of the photovoltaic power generation device under the snow condition according to the power generation efficiency model;

[0040] According to a preset temperature interval division standard, the second historical power generation data is divided into multiple temperature intervals and their corresponding historical data distribution diagrams, and combined with the corresponding snow thickness data to quantify the snow impact degree index;

[0041] Quantify the constant-temperature efficiency parameters through the snow impact degree index and the historical temperature and humidity data in the second historical power generation data, and based on the constant-temperature efficiency parameters and each temperature interval, a polynomial fitting method is used to process the power generation efficiency curve to generate an aging curve set;

[0042] calculate a standard deviation of each aging curve in the set of aging curves, and determine a first historical isothermal aging efficiency curve of the photovoltaic power generation device under the snow day condition according to the standard deviation.

[0043] Specifically, the application extracts key parameters such as light radiation intensity and power generation power from the second historical power generation data, and cleanses them together with the snow thickness data to ensure the accuracy and integrity of the data. Then, the light radiation intensity and the snow thickness (snow reflection) are taken as independent variables, and the power generation power is taken as a dependent variable. The least square method is used to calculate the regression coefficient, and then a linear regression model is established. According to the regression model, the trend data of the power generation power changing with the light intensity is output, which reflects the influence law of the light intensity on the power generation power under the corresponding snow thickness condition. The snow reflection can be identified by comparing the difference in light radiation intensity between sunny days and snowy days. For example, if the light radiation intensity at 10 o'clock in the morning is suddenly 20% higher than the average value of the same period and lasts for several hours, it means that there is snow reflection.

[0044] The power generation efficiency data of the photovoltaic power generation device in the snow day period is extracted. The power generation efficiency can be calculated by the ratio of the actual power generation power to the theoretical maximum power generation power. Based on the trend data output by the regression model, the correlation between power fluctuation and power generation efficiency is quantified using a quadratic polynomial fitting method to obtain a power generation efficiency model. This model can represent the functional relationship between power generation efficiency and parameters such as light intensity and snow thickness (snow reflection). Then, the power generation efficiency model is used to draw a power generation efficiency curve of the photovoltaic power generation device under the snow day condition, which can reflect the change law of the device's power generation efficiency under different light intensity and snow thickness conditions. Using this curve, the future power generation efficiency under similar weather conditions can be predicted. For the power generation efficiency model, if the mean square error with the historical data exceeds the preset error threshold, the gradient descent method is used to adjust the model parameters.

[0045] According to the preset temperature interval division standard (such as every 5°C as an interval), the second historical power generation data is divided into multiple temperature intervals, and the power generation data in each temperature interval is counted to draw a historical data distribution graph to show the distribution of parameters such as power generation power and power generation efficiency in different temperature intervals. Then, based on the historical data distribution graph corresponding to each temperature interval and the snow thickness data, the influence degree of snow day on the photovoltaic power generation device is quantified by calculating relevant statistical indicators (such as average power generation power reduction rate and power generation efficiency loss rate) to obtain a snow day influence degree index.

[0046] The snow temperature data in the second historical power generation data is extracted, and divided by the value of the snow day influence degree index, so that a constant temperature efficiency parameter is obtained, which reflects the ability of the device to maintain stable power generation efficiency under specific temperature conditions; based on the constant temperature efficiency parameter and each temperature interval, a third-order polynomial fitting method is used to process the power generation efficiency curve, the order and coefficient of the polynomial are adjusted to make the fitting curve better reflect the aging trend of the device under different temperature and snow day conditions, and the polynomial fitting processing is performed on each temperature interval respectively, so that a set of aging curves is obtained.

[0047] The standard deviation of each aging curve in the aging curve set is calculated to reflect the dispersion degree of the curve data, the larger the standard deviation, the greater the fluctuation of the curve data, and the more unstable the aging of the device; and the aging curve with the smallest standard deviation is selected as the first historical constant temperature aging efficiency curve of the photovoltaic power generation device under snow day conditions, so as to reflect the actual performance change of the photovoltaic device under different temperature and snow conditions to the greatest extent.

[0048] The present application comprehensively considers the influence of illumination intensity, snow thickness, temperature, humidity and other factors on the power generation efficiency of photovoltaic power generation equipment in snow days to construct the first historical constant temperature aging efficiency curve, and reveals the aging trend of photovoltaic power generation equipment under snow day conditions with time or use; based on the quantitative analysis of the snow day influence degree index and the constant temperature efficiency parameter, more reasonable device operation and maintenance strategies are formulated to improve the reliability and power generation efficiency of the device.

[0049] S2, determine the snow state change area in the first historical constant temperature aging efficiency curve, and determine the first historical photovoltaic conversion efficiency data of the snow state change area based on the snow distribution uniformity;

[0050] In an embodiment, the determination of the snow state change area in the first historical constant temperature aging efficiency curve comprises:

[0051] Extract the photovoltaic conversion efficiency value of the first historical constant temperature aging efficiency curve in the preset sampling period to obtain a plurality of time distribution points to quantify the conversion efficiency difference between the photovoltaic conversion efficiency values corresponding to adjacent time distribution points, and the time distribution points corresponding to the conversion efficiency difference exceeding the preset conversion efficiency change threshold are taken as snow state change candidate points;

[0052] The clustering algorithm is used to group the snow state change candidate points to obtain a preliminary snow state change area, and the change trend of the photovoltaic conversion efficiency in the preliminary snow state change area is analyzed to determine the final snow state change area based on the continuity of the change trend of the photovoltaic conversion efficiency.

[0053] Specifically, the present application extracts the photovoltaic conversion efficiency value (i.e. power generation efficiency data) from the first historical constant temperature aging efficiency curve according to a preset sampling period (such as every hour, every day) to form time series data, then performs difference calculation on the photovoltaic conversion efficiency values of adjacent time distribution points to obtain a conversion efficiency difference value sequence, and marks the time distribution points with a conversion efficiency difference value exceeding a preset conversion efficiency change threshold (such as 2%) as snow state change candidate points through a conditional judgment statement.

[0054] The K-means algorithm is used to group the snow state change candidate points, and the number of clusters can be set to 3, representing the snow accumulation, initial melting and accelerated melting stages respectively, to preliminarily divide the snow state change time region and obtain a preliminary snow state change region. The rationality of the region division is verified by analyzing the change trend of the photovoltaic conversion efficiency in the preliminary division region through NumPy. Finally, it is checked whether the change trend line of each preliminary snow state change region meets the physical law of snow cover (efficiency drop) or ablation (efficiency recovery), and the region that passes the continuity verification is retained as the final snow state change region.

[0055] The present application can accurately identify the key time points of snow cover and ablation by quantifying the change of photovoltaic conversion efficiency, combining clustering algorithm and trend analysis, avoiding the subjectivity of manual judgment. By using the time distribution points and conversion efficiency difference values in the historical data, the continuous efficiency curve is converted into discrete key points, reducing data redundancy while retaining key information. Through the process of quantitative analysis, clustering grouping and trend verification, combining data-driven (efficiency difference value, clustering) and physical law (trend continuity), the present application realizes the accurate identification of snow state change region, effectively improves the accuracy and automation of snow monitoring.

[0056] In an embodiment, the first historical photovoltaic conversion efficiency data for determining the snow state change region based on the snow distribution uniformity includes:

[0057] Real-time snow depth data and corresponding snow cover images of the snow state change region are obtained, and image processing technology is used to process the snow cover images to obtain snow contour data to determine the snow distribution uniformity;

[0058] Based on the real-time snow depth data and the snow distribution uniformity, a real-time time region of snow state change is divided through a preset snow depth change rate threshold and a preset uniformity change rate threshold;

[0059] extracting the forward historical photovoltaic conversion efficiency and the backward historical photovoltaic conversion efficiency at both ends of the period corresponding to the real-time time region in the snow state change region, and processing the forward historical photovoltaic conversion efficiency and the backward historical photovoltaic conversion efficiency by using a weighted average method based on the snow distribution uniformity to obtain a forward weighted historical photovoltaic conversion efficiency and a backward weighted historical photovoltaic conversion efficiency;

[0060] quantifying a weighted efficiency difference between the forward weighted historical photovoltaic conversion efficiency and the backward weighted historical photovoltaic conversion efficiency, and determining that the real-time time region is a significant change region when the weighted efficiency difference exceeds a preset efficiency difference threshold;

[0061] extracting the historical photovoltaic conversion efficiency at both ends of the period corresponding to the significant change region in the snow state change region as first historical photovoltaic conversion efficiency data.

[0062] Specifically, the present application uses a laser radar or an ultrasonic sensor to obtain real-time snow depth data of a snow state change region; at the same time, a real-time snow cover image is collected by a drone or a fixed camera, and the snow cover image is subjected to grayscale processing to convert a color image into a grayscale image, binarization processing is performed by setting a suitable threshold value to clearly separate the snow region and the non-snow region in the image, so as to effectively extract the contour information of the snow region and lay a foundation for subsequent analysis; then, the ratio of the pixel standard deviation to the pixel average value of the snow region is calculated as the snow distribution uniformity, and the lower the uniformity is, the more uniform the snow distribution is.

[0063] The real-time snow depth data and the snow distribution uniformity are subjected to time series analysis, the depth change rate and the uniformity change rate of adjacent time points are calculated, the depth change rate is compared with a preset snow depth change rate threshold (such as 0.5 cm per hour), and the uniformity change rate is compared with a preset uniformity change rate threshold (such as 0.1), if any threshold is broken, the time point is marked as a critical point of snow state change; then, the time series is divided into a plurality of real-time time regions according to the critical point, so as to accurately capture the dynamic change process of the snow state; wherein each region represents a snow state change event.

[0064] From the historical database of the photovoltaic power station (which contains power generation data of photovoltaic power generation equipment in the past preset time period and related weather data, environmental data, equipment operation data, etc.), the forward historical photovoltaic conversion efficiency (corresponding to the front end of the time period) and the backward historical photovoltaic conversion efficiency (corresponding to the back end of the time period) at both ends of the time period corresponding to the real-time time area are extracted respectively. The forward historical photovoltaic conversion efficiency and the backward historical photovoltaic conversion efficiency are weighted according to the snow distribution uniformity design, and the forward weighted historical photovoltaic conversion efficiency and the backward weighted historical photovoltaic conversion efficiency are obtained. Among them, the weight of the weighted average is the reciprocal of the snow distribution uniformity index.

[0065] The difference between the forward-weighted historical photovoltaic conversion efficiency and the backward-weighted historical photovoltaic conversion efficiency is used as the weighted efficiency difference value. This value is then compared with a preset efficiency difference threshold (e.g., 0.03). If the threshold is exceeded, the real-time time area is determined to be a region of significant change. The forward and backward historical photovoltaic conversion efficiency data corresponding to the time period of the region of significant change are then extracted from the historical database as the first historical photovoltaic conversion efficiency data output for subsequent analysis or modeling. Otherwise, the forward and backward historical photovoltaic conversion efficiency data at both ends of the corresponding time period are extracted from the historical database according to a preset fixed time period, such as daily, weekly, or monthly, to serve as the first historical photovoltaic conversion efficiency data.

[0066] This invention uses real-time snow depth data and snow distribution uniformity extracted through image processing technology to more accurately reflect the impact of snow condition changes on photovoltaic conversion efficiency, avoiding the limitations of a single data source. By combining real-time monitoring data (snow depth, images) and historical efficiency data, and using a weighted average method to process historical photovoltaic conversion efficiency, the weights of forward and backward efficiencies can be dynamically adjusted, more scientifically quantifying the efficiency differences before and after snow changes, and improving the accuracy of identifying areas of significant change.

[0067] S3. Based on the first historical photovoltaic conversion efficiency data, determine the equipment status data and snow adhesion intensity data corresponding to the snow condition change area, so as to construct the second historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions.

[0068] In one embodiment, step S3 includes:

[0069] Based on the first historical photovoltaic conversion efficiency data, the historical snow accumulation time and historical power generation efficiency data of the photovoltaic power generation equipment are obtained, so as to determine the first equipment status data and its corresponding equipment failure time area from the snow status change area;

[0070] According to the device failure time region, an actual constant temperature aging efficiency curve is extracted from the first historical constant temperature aging efficiency curve, and a random forest algorithm is used to smooth the actual constant temperature aging efficiency curve to obtain an optimized constant temperature aging efficiency curve.

[0071] First snow accumulation intensity data and first snow weather condition data of the photovoltaic power generation device in the device failure time region are extracted, and a first snow weather power generation duration of the photovoltaic power generation device is quantified according to the first snow weather condition data;

[0072] The first snow accumulation intensity data is compared with a preset intensity threshold, and when the first snow accumulation intensity data exceeds the preset intensity threshold, a first constant temperature aging parameter is quantified according to the first snow weather power generation duration;

[0073] The optimized constant temperature aging efficiency curve is adjusted according to the first constant temperature aging parameter to obtain a second historical constant temperature aging efficiency curve of the photovoltaic power generation device under the snow weather condition.

[0074] Specifically, the first historical photovoltaic conversion efficiency data in the application includes forward and backward historical photovoltaic conversion efficiency data. When the forward historical photovoltaic conversion efficiency data is greater than the backward historical photovoltaic conversion efficiency data, it indicates that the device has failed. Combined with historical meteorological data and historical device sensor data, the historical snow accumulation time on the surface of the device is determined, and the historical power generation efficiency data of the photovoltaic power generation device is extracted. Then, by setting a reasonable threshold and judgment rule, the first device state data (that is, the device failure state data, which may include abnormal parameters such as voltage, current, and temperature of the device) is selected from the snow state change region to determine the time interval of the device failure, that is, the device failure time region (for example, when the output power of the device continuously drops below a certain set value and lasts for a certain period of time, the time period is marked as the device failure time region).

[0075] According to the device failure time region, an actual constant temperature aging efficiency curve is extracted from the first historical constant temperature aging efficiency curve, and a random forest algorithm is used to smooth the actual constant temperature aging efficiency curve to obtain an optimized constant temperature aging efficiency curve.

[0076] The first snow accumulation intensity data and the first snow weather condition data of the photovoltaic power generation device in the device failure time area are extracted from the historical database; wherein, the snow accumulation intensity data can be measured by a pressure sensor or an optical sensor on the surface of the device, and the snow weather condition data includes snowfall, wind speed, temperature and humidity, snowfall time, etc.; then, the first snow weather power generation time of the photovoltaic power generation device is quantified according to the first snow weather condition data, such as snowfall and snowfall time; for example, when the snowfall is small and the wind speed is low, the device can still maintain a certain power generation capacity, at this time, the time of normal power generation of the device is counted into the first snow weather power generation time; when the snowfall is large or the wind speed is too high to cause the device to stop power generation, the time period is not counted into the power generation time; the snow weather power generation time can also be determined according to the snowfall and the temperature, that is, by constructing a relationship model with snowfall and temperature as input and power generation time as output according to historical data, for example, when the snowfall is 15 mm and the temperature is-5 degrees Celsius, the power generation time is shortened to 30% of the normal condition, that is, only 3 hours of power generation per day when the snow completely covers the device.

[0077] The first snow accumulation intensity data is compared with the preset intensity threshold value, when the first snow accumulation intensity data exceeds the preset intensity threshold value, it indicates that the influence of snow on the device is large, which may cause the acceleration of the aging of the device, and then the first constant temperature aging parameter is quantified according to the first snow weather power generation time. Wherein, the preset intensity threshold value is set according to the performance and design requirements of the device, which is used to judge the influence degree of snow on the device; the constant temperature aging parameter is a function related to the power generation time, when the power generation time is short, the first constant temperature aging parameter is large, which indicates that the aging speed of the device under the snow weather condition is accelerated.

[0078] The first constant temperature aging parameter is used as an adjustment factor to shift or scale the optimized constant temperature aging efficiency curve by using a linear or nonlinear method, so as to reflect the influence of snow accumulation intensity on the aging efficiency of the device; after adjustment, the second historical constant temperature aging efficiency curve of the photovoltaic power generation device under the snow weather condition is obtained, which comprehensively considers the factors such as snow state, device state and snow accumulation intensity, and can more accurately describe the aging rule and power generation efficiency change of the device under the snow weather condition.

[0079] This invention constructs a second historical isothermal aging efficiency curve by comprehensively considering factors such as snow accumulation, equipment condition, and snow adhesion intensity. This curve can more accurately reflect the actual aging of photovoltaic power generation equipment and changes in power generation efficiency under snowy conditions. By clarifying the equipment failure time zone and the isothermal aging parameters under different snow adhesion intensities, it helps maintenance personnel to formulate targeted maintenance plans in advance, rationally allocate maintenance resources, reduce the probability of equipment failure under snowy conditions, extend equipment life, and enable the photovoltaic power generation system to better adapt to complex environmental conditions such as snowy weather. By adjusting and optimizing the isothermal aging efficiency curve, the system can more rationally allocate power generation tasks according to actual conditions, thereby improving the stability and reliability of the entire photovoltaic power generation system.

[0080] In one embodiment, step S3 further includes:

[0081] Based on the first historical photovoltaic conversion efficiency data, historical snow humidity content and historical snowy weather equipment aging data of the photovoltaic power generation equipment are extracted to quantify the trend of snow condition change, and the second equipment status data and equipment replacement time corresponding to the snow condition change area are determined according to the trend of snow condition change.

[0082] Extract the second snow adhesion intensity data and the second snow weather condition data of the photovoltaic power generation equipment during the equipment replacement time, and quantify the second snow weather power generation duration of the photovoltaic power generation equipment based on the second snow weather condition data;

[0083] The second snow adhesion strength data is compared with a preset strength threshold, and when the second snow adhesion strength data exceeds the preset strength threshold, the second constant temperature aging parameter is quantified according to the second snowy day power generation duration.

[0084] Based on the second isothermal aging parameter, the least squares method is used to fit the isothermal aging trend curve of the photovoltaic power generation equipment to serve as the second historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions.

[0085] Specifically, this invention cleans and organizes the first historical photovoltaic conversion efficiency data, removing outliers and missing values ​​to ensure data accuracy and completeness. Then, it extracts snow-related features from the preprocessed data, such as snow moisture content, which can be indirectly obtained by analyzing changes in equipment conversion efficiency under different snow conditions, and which can affect snow adhesion. Simultaneously, it extracts historical snow-related equipment aging data from the historical database, such as equipment performance degradation rate and failure rate. Next, it uses time series analysis methods, such as moving average and exponential smoothing, to perform trend analysis on the extracted features, quantifying the trend of snow condition changes. Then, based on the snow condition change trend and the equipment's operating parameters and performance indicators, it determines the second equipment status data (i.e., equipment replacement status data, such as whether the equipment is aging, damaged, or in a situation requiring replacement) corresponding to the snow condition change area. Finally, based on the second equipment status data and historical maintenance records, it determines the equipment replacement time.

[0086] Data on the second snow adhesion intensity and second snowy weather conditions of photovoltaic power generation equipment within the equipment replacement period are obtained from historical databases. The collected data are preprocessed, including data filtering and normalization, to improve data quality and usability. Based on the second snowy weather conditions data and the operating characteristics of the equipment, the power generation duration of the second snowy weather is quantified. The quantification process can use a threshold judgment method. When the conditions such as light intensity and temperature meet the power generation requirements of the equipment, the timing starts and continues until the conditions are no longer met.

[0087] Based on the material properties, design requirements, and usage experience of the equipment, a preset strength threshold is set. The preset strength threshold should be able to reflect the safe operating range of the equipment under snow adhesion conditions. The second snow adhesion strength data is compared with the preset strength threshold to determine whether the snow adhesion strength exceeds the threshold. When the second snow adhesion strength data exceeds the preset strength threshold, the second constant temperature aging parameter is quantified based on the second snowy day power generation duration. The quantification of the second constant temperature aging parameter can also use empirical formulas or models to establish a correlation between the power generation duration and the constant temperature aging parameter, and then calculate the value of the second constant temperature aging parameter.

[0088] Using the second isothermal aging parameter as the independent variable and the historical photovoltaic conversion efficiency of the equipment as the dependent variable, the least squares method is used for fitting to construct the isothermal aging trend curve, which is then used as the second historical isothermal aging efficiency curve. During the curve fitting process, numerical calculation methods, such as the Gauss-Newton method, can be used to solve for the optimal solution of the isothermal aging trend curve parameters. Residual analysis, correlation coefficient calculation, and other methods are used to evaluate the fitted second historical isothermal aging efficiency curve to check the goodness of fit and rationality of the curve.

[0089] This invention quantifies the trend of snow cover changes by deeply analyzing historical photovoltaic conversion efficiency data, accurately determining the equipment status data and replacement time corresponding to the snow cover change area, providing a scientific basis for equipment maintenance and management; accurately quantifying snow adhesion intensity and power generation duration enables a comprehensive understanding of the equipment's operating status under snowy conditions, providing accurate data support for subsequent constant temperature aging parameter quantification; and scientifically constructing constant temperature aging efficiency curves provides an important reference for equipment performance evaluation and optimization.

[0090] S4. Based on the second historical constant temperature aging efficiency curve and its corresponding historical temperature and humidity data, determine the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment, and calculate the power prediction value of the photovoltaic power generation equipment in the future preset period based on the second historical photovoltaic conversion efficiency data and output it.

[0091] In one embodiment, determining the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment based on the second historical isothermal aging efficiency curve and its corresponding historical temperature and humidity data includes:

[0092] Based on the second historical isothermal aging efficiency curve and its corresponding historical temperature and humidity target data, an environmental impact factor model is constructed to extract the efficiency characteristics and aging characteristics of the photovoltaic power generation equipment.

[0093] The efficiency and aging characteristics are analyzed using a support vector machine algorithm to obtain the snow impact weight coefficient. When the snow impact weight coefficient exceeds a preset weight coefficient threshold, the historical photovoltaic conversion efficiency change trend data of the photovoltaic power generation equipment is extracted.

[0094] Based on the historical photovoltaic conversion efficiency change trend data, the photovoltaic conversion dynamic adjustment parameters of the photovoltaic power generation equipment under the influence of snow accumulation are determined, and the photovoltaic conversion dynamic adjustment parameters and the historical temperature and humidity target data are subjected to linear regression analysis to determine the conversion efficiency fluctuation range of the photovoltaic power generation equipment.

[0095] Historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment within the range of conversion efficiency fluctuation is extracted as the second historical photovoltaic conversion efficiency data.

[0096] Specifically, this invention cleans the second historical isothermal aging efficiency curve and its corresponding historical temperature and humidity target data to remove outliers and missing values, ensuring data quality and integrity. Feature extraction methods, such as principal component analysis and linear discriminant analysis, are used to analyze the relationship between the isothermal aging efficiency curve and the historical temperature and humidity target data. This extracts feature vectors that best represent the equipment efficiency and aging characteristics from the original curve data and temperature and humidity data, such as temperature, humidity, power conversion rate and its rate of change, and aging time. Based on the extracted feature vectors, an environmental impact factor model is constructed using regression models (such as linear regression and multinomial regression) and neural network models (such as BP neural networks and RBF neural networks). The environmental impact factor model is trained using historical data, and model parameters are adjusted to accurately fit the actual data, extracting the efficiency and aging characteristics of the photovoltaic power generation equipment.

[0097] Further analysis of the extracted efficiency and aging features identifies a subset of features relevant to the impact of snow cover. This subset, along with corresponding snow cover state labels (such as snow thickness and snow cover duration), is used as input data for the Support Vector Machine (SVM) algorithm. Appropriate kernel functions (such as linear, polynomial, and radial basis function kernels) and penalty parameters are selected to construct an SVM classification model. The SVM model is trained using training data, and the hyperplane of the model is solved using an optimization algorithm (such as Sequential Minimum Optimization, SMO) to enable the model to accurately distinguish equipment features under different snow cover states. Based on the trained SVM model, the contribution of each feature in the classification process is determined by analyzing the SVM's decision function, thus outputting the weight coefficients of each feature's impact on snow cover, i.e., the snow cover impact weight coefficients. Then, the weighting coefficient of snow impact is compared with the preset weighting coefficient threshold. When the weighting coefficient of snow impact exceeds the threshold, it can be considered that snow has a significant impact on photovoltaic conversion efficiency. The photovoltaic conversion efficiency data of photovoltaic power generation equipment corresponding to the second historical constant temperature aging efficiency curve is extracted from the historical database, and its changing trend is analyzed to obtain historical photovoltaic conversion efficiency changing trend data.

[0098] Data analysis methods (such as moving average and exponential smoothing) are used to process historical photovoltaic (PV) conversion efficiency trend data to determine dynamic adjustment parameters for PV power generation equipment under the influence of snow accumulation. These parameters may include the adjustment range and time interval of conversion efficiency. Then, linear regression analysis is performed on the dynamic adjustment parameters and historical temperature and humidity target data to establish a regression model. The regression coefficients are estimated using the least squares method to obtain the regression equation. Based on the regression equation, the influence of temperature and humidity target data and dynamic adjustment parameters on PV conversion efficiency is analyzed to determine the fluctuation range of PV power generation equipment conversion efficiency.

[0099] Finally, photovoltaic conversion efficiency data within the fluctuation range of conversion efficiency are extracted from the historical photovoltaic conversion efficiency database, and the extracted data is verified to check whether the data conforms to the actual operating conditions. Verification methods can include cross-validation with other model prediction results, and the verified data is used as the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment.

[0100] This invention, by combining the second historical isothermal aging efficiency curve with corresponding historical temperature and humidity data, can accurately determine the second historical photovoltaic conversion efficiency data of photovoltaic power generation equipment under complex environments (especially the influence of snow accumulation), providing a solid data foundation for subsequent equipment performance evaluation, optimization, and operation and maintenance decisions. Constructing an environmental impact factor model to extract efficiency and aging characteristics helps to deeply understand the operating patterns of photovoltaic power generation equipment under different environmental conditions, providing an important basis for equipment performance improvement and fault prediction. Using a support vector machine algorithm to analyze efficiency and aging characteristics to obtain the snow accumulation impact weight coefficient can scientifically quantify the degree of snow accumulation's influence on photovoltaic conversion efficiency, providing an accurate reference for subsequent dynamic adjustment parameter determination. Accurately determining the conversion efficiency fluctuation range through linear regression analysis can more accurately reflect the performance changes of the equipment in actual operation. Employing a series of data analysis methods and algorithms, with a high degree of automation, it can quickly and accurately process large amounts of data, improving the efficiency and accuracy of data analysis.

[0101] In one embodiment, the step of calculating and outputting a predicted power value of the photovoltaic power generation device for a future preset period based on the second historical photovoltaic conversion efficiency data includes:

[0102] Based on the second historical photovoltaic conversion efficiency data and its corresponding snowy weather conditions data, a conversion efficiency prediction model is constructed using machine learning algorithms.

[0103] The weather forecast data for the future preset period is obtained and input into the conversion efficiency prediction model for processing, and the conversion efficiency prediction value for each time point within the future preset period is output.

[0104] The system parameters of the photovoltaic power generation equipment are obtained and combined with the predicted conversion efficiency values ​​to quantify and output the predicted power value of the photovoltaic power generation equipment in the future preset period.

[0105] Specifically, this invention constructs sample data from historical photovoltaic conversion efficiency data and corresponding snow conditions (such as snowfall, snow thickness, temperature, solar irradiance, wind speed, humidity, etc.) for the corresponding time period, and cleans the data to ensure accuracy. Then, key features are extracted from the sample data, such as time features (hour, season, daytime), and weather features (current snow thickness, temperature, irradiance, etc.). Random forests, XGBoost, and other network models suitable for handling multi-feature regression are selected. A machine learning algorithm is used, with historical conversion efficiency, weather features, and time features as model inputs, and predicted conversion efficiency values ​​for a preset future time as model outputs, to iteratively train the model. Root mean square error is used as an evaluation metric during training until the number of iterations reaches a preset value, such as 10,000, or the model accuracy meets the requirements, thus obtaining the efficiency prediction model.

[0106] Weather forecast data for the target area in the future for a predetermined period can be obtained from professional meteorological agencies or professional meteorological service APIs, or from meteorological forecast models trained by machine learning algorithms using historical meteorological data of the target area. This includes, but is not limited to, temperature and humidity, snowfall, solar irradiance, and snow thickness. This weather forecast data is then input into the trained efficiency forecast model for processing, and the predicted conversion efficiency values ​​of the photovoltaic power generation equipment at each time point in the future predetermined period are output. The future predetermined period is generally 12 hours or 24 hours.

[0107] By obtaining the system parameters of the photovoltaic power generation equipment, such as the area of ​​the photovoltaic panel, the installation angle, and the geographical location, and by calculating the product of the predicted irradiance value at a future preset time and the predicted conversion efficiency value of the photovoltaic panel area at each time point, the power prediction value of the photovoltaic power generation equipment in the future preset period can be obtained.

[0108] This invention is based on the correlation analysis between second historical photovoltaic conversion efficiency data and snowy weather conditions. It uses a machine learning model to capture the nonlinear relationship between the two to reduce efficiency prediction errors. Future weather forecast data is input into the conversion efficiency prediction model, and then the short-term predicted power of the photovoltaic equipment is calculated based on the output predicted conversion efficiency results and the system parameters of the equipment, realizing the full-chain quantification of "weather-efficiency-power". By combining machine learning algorithms with the fusion of multi-source data, the power prediction accuracy of photovoltaic power generation equipment under snowy weather conditions is systematically improved.

[0109] This application proposes a photovoltaic power generation prediction method based on the impact of snow accumulation to address the problem of improving the prediction results of photovoltaic module power generation. This method constructs a first historical isothermal aging efficiency curve by combining historical power generation, temperature, humidity, and solar radiation intensity data to quantify the real-time impact of snow cover on photovoltaic conversion efficiency. It locates areas of snow condition change based on the correlation analysis between snow distribution uniformity and efficiency abrupt change points, providing accurate input for subsequent analysis. By using the first historical photovoltaic conversion efficiency data and snow distribution data, it quantifies the impact of snow on photovoltaic module power generation efficiency, providing a basis for equipment selection and operation and maintenance strategy optimization. A second historical isothermal aging efficiency curve is constructed based on equipment status and snow adhesion intensity data, and combined with future weather forecast data to achieve short-term power prediction of photovoltaic power generation equipment, providing precise support for grid dispatch and energy storage configuration. This invention can accurately assess the impact of snow on photovoltaic power generation efficiency, providing effective support for photovoltaic power generation equipment management and power prediction under snowy conditions.

[0110] It should be noted that although the steps in the flowchart above are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order requirement for the execution of these steps, and they can be executed in other orders.

[0111] In another embodiment, such as Figure 2 As shown, a second aspect of the present invention provides a photovoltaic power generation prediction system based on the effects of snow cover, comprising:

[0112] The first curve construction module 10 is used to obtain historical power generation data, historical temperature and humidity data and historical solar radiation intensity data of photovoltaic power generation equipment under snow conditions in the target area, so as to construct the first historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snow conditions.

[0113] Efficiency data calculation module 20 is used to determine the snow state change area in the first historical constant temperature aging efficiency curve, and to determine the first historical photovoltaic conversion efficiency data of the snow state change area based on the snow distribution uniformity.

[0114] The second curve construction module 30 is used to determine the equipment status data and snow adhesion intensity data corresponding to the snow condition change area based on the first historical photovoltaic conversion efficiency data, so as to construct the second historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions.

[0115] The power generation prediction module 40 is used to determine the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment based on the second historical constant temperature aging efficiency curve and its corresponding historical temperature and humidity data, and to calculate and output the power prediction value of the photovoltaic power generation equipment in a future preset period based on the second historical photovoltaic conversion efficiency data.

[0116] It should be noted that the modules in the aforementioned photovoltaic power generation prediction system based on snow cover effects can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module. For specific limitations regarding the photovoltaic power generation prediction system based on snow cover effects, please refer to the limitations of the photovoltaic power generation prediction method based on snow cover effects described above; both have the same function and role, and will not be repeated here.

[0117] In summary, this invention relates to the field of information technology and discloses a method and system for predicting photovoltaic power generation based on the impact of snow cover. It constructs a first historical isothermal aging efficiency curve using historical power generation data of photovoltaic power generation equipment in a target area under snowy conditions, and identifies areas of snow cover change within this curve. Based on the uniformity of snow distribution, it determines the first historical photovoltaic conversion efficiency data for these areas. Based on the first historical photovoltaic conversion efficiency data and the areas of snow cover change, it determines the equipment status data and snow adhesion intensity data of the photovoltaic power generation equipment, thereby constructing a second historical isothermal aging efficiency curve. The obtained curve is then combined with corresponding historical temperature and humidity data to determine the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment. This allows for the calculation of the predicted power value of the photovoltaic power generation equipment in a future preset period, accurately assessing the impact of snow cover on photovoltaic power generation efficiency, and thus precisely predicting the power of the photovoltaic power generation equipment under the influence of snow cover.

[0118] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts of the embodiments, refer to each other. Each embodiment focuses on its differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0119] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the technical principles of this invention, and these improvements and substitutions should also be considered within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the scope of the claims.

Claims

1. A photovoltaic power generation prediction method based on the impact of snow cover, characterized in that, include: Historical power generation data, historical temperature and humidity data, and historical solar radiation intensity data of photovoltaic power generation equipment in the target area under snowy conditions are obtained to construct the first historical isothermal aging efficiency curve of the photovoltaic power generation equipment under snowy conditions. The snow condition change region in the first historical isothermal aging efficiency curve is determined, and the first historical photovoltaic conversion efficiency data of the snow condition change region is determined based on the snow distribution uniformity. Based on the first historical photovoltaic conversion efficiency data, the equipment status data and snow adhesion intensity data corresponding to the snow condition change area are determined to construct the second historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions. Based on the second historical constant temperature aging efficiency curve and its corresponding historical temperature and humidity data, the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment is determined, and the power prediction value of the photovoltaic power generation equipment in the future preset period is calculated and output based on the second historical photovoltaic conversion efficiency data.

2. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The acquisition of historical power generation data, historical temperature and humidity data, and historical solar radiation intensity data of photovoltaic power generation equipment in the target area under snowy conditions includes: The system acquires the first historical power generation data of photovoltaic power generation equipment in the target area within a preset time period, and determines the time period in which the historical temperature and humidity data exceed the preset temperature and humidity threshold and the rate of decrease of the historical solar radiation intensity data reaches the preset rate of decrease threshold as a snowy time period; the first historical power generation data includes historical power generation data, historical temperature and humidity data and historical solar radiation intensity data; The first historical power generation data during the snowy period is used as the second historical power generation data, and a linear regression algorithm is used to determine the light attenuation data in the second historical power generation data. Based on the pre-built correlation model between snow thickness and light attenuation, the snow thickness data corresponding to the second historical power generation data is calculated. Based on the second historical power generation data, the distribution characteristics of the historical power generation data of the photovoltaic power generation equipment during the snowy period are quantified to obtain the differences between the equipment. Then, a clustering algorithm is used to group the second historical power generation data and its corresponding snow thickness data to obtain the characteristic description of the photovoltaic power generation equipment between snow thickness and light attenuation under the snowy conditions.

3. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 2, characterized in that, The construction of the first historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions includes: Based on the aforementioned feature description, using the solar radiation intensity and its corresponding snow thickness in the second historical power generation data as independent variables and the power generation in the second historical power generation data as dependent variables, linear regression is performed using the least squares method to obtain trend data of power generation with solar radiation intensity. Extract the power generation efficiency data of the photovoltaic power generation equipment during the snowy period, and based on the trend data, quantify the correlation between power fluctuation and power generation efficiency through a polynomial fitting method to obtain a power generation efficiency model, and construct the power generation efficiency curve of the photovoltaic power generation equipment under the snowy conditions based on the power generation efficiency model. The second historical power generation data is divided into intervals according to the preset temperature interval division standard to obtain multiple temperature intervals and their corresponding historical data distribution maps, and combined with the corresponding snow thickness data to quantify the degree of snow impact. The isothermal efficiency parameter is quantified by the snow impact index and the historical temperature and humidity data in the second historical power generation data. Based on the isothermal efficiency parameter and each temperature range, the power generation efficiency curve is processed by a polynomial fitting method to generate an aging curve set. Calculate the standard deviation of each aging curve in the set of aging curves, and determine the first historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions based on the standard deviation.

4. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The determination of the snow accumulation state change region in the first historical isothermal aging efficiency curve includes: Extract the photovoltaic conversion efficiency value of the first historical constant temperature aging efficiency curve within a preset sampling period to obtain multiple time distribution points to quantify the conversion efficiency difference between photovoltaic conversion efficiency values ​​corresponding to adjacent time distribution points, and take the time distribution point corresponding to the conversion efficiency difference exceeding the preset conversion efficiency change threshold as a candidate point for snow accumulation status change. Clustering algorithms are used to group the candidate points of snow cover state change to obtain preliminary snow cover state change regions. The changing trend of photovoltaic conversion efficiency within the preliminary snow cover state change regions is analyzed to determine the final snow cover state change regions based on the continuity of the changing trend of photovoltaic conversion efficiency.

5. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The first historical photovoltaic conversion efficiency data for determining the snow cover state change area based on snow cover distribution uniformity includes: The real-time snow depth data and corresponding snow cover image of the area where the snow condition changes are obtained, and the snow cover image is processed using image processing technology to obtain snow contour data to determine the uniformity of snow distribution. Based on the real-time snow depth data and the snow distribution uniformity, the real-time time region of snow state change is divided by a preset snow depth change rate threshold and a preset uniformity change rate threshold. Extract the forward historical photovoltaic conversion efficiency and the backward historical photovoltaic conversion efficiency at both ends of the time period corresponding to the real-time time region in the snow cover state change region, and process the forward historical photovoltaic conversion efficiency and the backward historical photovoltaic conversion efficiency using a weighted average method based on the snow cover distribution uniformity to obtain the forward weighted historical photovoltaic conversion efficiency and the backward weighted historical photovoltaic conversion efficiency. The weighted efficiency difference between the forward-weighted historical photovoltaic conversion efficiency and the backward-weighted historical photovoltaic conversion efficiency is quantified, and when the weighted efficiency difference exceeds a preset efficiency difference threshold, the real-time time region is determined to be a region of significant change. Extract the historical photovoltaic conversion efficiency at both ends of the time period corresponding to the area of ​​significant change in the snow cover status change region to output as the first historical photovoltaic conversion efficiency data.

6. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The step of determining the equipment status data and snow adhesion intensity data corresponding to the snow-covered state change area based on the first historical photovoltaic conversion efficiency data, in order to construct the second historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy conditions, includes: Based on the first historical photovoltaic conversion efficiency data, the historical snow accumulation time and historical power generation efficiency data of the photovoltaic power generation equipment are obtained, so as to determine the first equipment status data and its corresponding equipment failure time area from the snow status change area; Based on the equipment failure time range, the actual constant temperature aging efficiency curve is extracted from the first historical constant temperature aging efficiency curve, and the actual constant temperature aging efficiency curve is smoothed by the random forest algorithm to obtain the optimized constant temperature aging efficiency curve. Extract the first snow adhesion intensity data and the first snow day condition data of the photovoltaic power generation equipment within the equipment failure time area, and quantify the first snow day power generation duration of the photovoltaic power generation equipment based on the first snow day condition data; The first snow adhesion strength data is compared with a preset strength threshold, and when the first snow adhesion strength data exceeds the preset strength threshold, the first constant temperature aging parameter is quantified according to the first snowy day power generation duration. The optimized isotemperature aging efficiency curve is adjusted according to the first isotemperature aging parameters to obtain the second historical isotemperature aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions.

7. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The step of determining the equipment status data and snow adhesion intensity data corresponding to the snow-covered state change area based on the first historical photovoltaic conversion efficiency data, in order to construct the second historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions, further includes: Based on the first historical photovoltaic conversion efficiency data, historical snow humidity content and historical snowy weather equipment aging data of the photovoltaic power generation equipment are extracted to quantify the trend of snow condition change, and the second equipment status data and equipment replacement time corresponding to the snow condition change area are determined according to the trend of snow condition change. Extract the second snow adhesion intensity data and the second snow weather condition data of the photovoltaic power generation equipment during the equipment replacement time, and quantify the second snow weather power generation duration of the photovoltaic power generation equipment based on the second snow weather condition data; The second snow adhesion strength data is compared with a preset strength threshold, and when the second snow adhesion strength data exceeds the preset strength threshold, the second constant temperature aging parameter is quantified according to the second snowy day power generation duration. Based on the second isothermal aging parameter, the least squares method is used to fit the isothermal aging trend curve of the photovoltaic power generation equipment to serve as the second historical isothermal aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions.

8. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The step of determining the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment based on the second historical isothermal aging efficiency curve and its corresponding historical temperature and humidity data includes: Based on the second historical isothermal aging efficiency curve and its corresponding historical temperature and humidity target data, an environmental impact factor model is constructed to extract the efficiency characteristics and aging characteristics of the photovoltaic power generation equipment. The efficiency and aging characteristics are analyzed using a support vector machine algorithm to obtain the snow impact weight coefficient. When the snow impact weight coefficient exceeds a preset weight coefficient threshold, the historical photovoltaic conversion efficiency change trend data of the photovoltaic power generation equipment is extracted. Based on the historical photovoltaic conversion efficiency change trend data, the photovoltaic conversion dynamic adjustment parameters of the photovoltaic power generation equipment under the influence of snow accumulation are determined, and the photovoltaic conversion dynamic adjustment parameters and the historical temperature and humidity target data are subjected to linear regression analysis to determine the conversion efficiency fluctuation range of the photovoltaic power generation equipment. Historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment within the range of conversion efficiency fluctuation is extracted as the second historical photovoltaic conversion efficiency data.

9. The photovoltaic power generation prediction method based on the influence of snow cover according to claim 1, characterized in that, The step of calculating and outputting the predicted power value of the photovoltaic power generation equipment for a future preset period based on the second historical photovoltaic conversion efficiency data includes: Based on the second historical photovoltaic conversion efficiency data and its corresponding snowy weather conditions data, a conversion efficiency prediction model is constructed using machine learning algorithms. The weather forecast data for the future preset period is obtained and input into the conversion efficiency prediction model for processing, and the conversion efficiency prediction value for each time point within the future preset period is output. The system parameters of the photovoltaic power generation equipment are obtained and combined with the predicted conversion efficiency values ​​to quantify and output the predicted power value of the photovoltaic power generation equipment in the future preset period.

10. A photovoltaic power generation prediction system based on the effects of snow cover, characterized in that, include: The first curve construction module is used to obtain historical power generation data, historical temperature and humidity data and historical solar radiation intensity data of photovoltaic power generation equipment under snow conditions in the target area, so as to construct the first historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snow conditions. The efficiency data calculation module is used to determine the snow state change area in the first historical constant temperature aging efficiency curve, and to determine the first historical photovoltaic conversion efficiency data of the snow state change area based on the snow distribution uniformity. The second curve construction module is used to determine the equipment status data and snow adhesion intensity data corresponding to the snow condition change area based on the first historical photovoltaic conversion efficiency data, so as to construct the second historical constant temperature aging efficiency curve of the photovoltaic power generation equipment under the snowy weather conditions. The power generation prediction module is used to determine the second historical photovoltaic conversion efficiency data of the photovoltaic power generation equipment based on the second historical constant temperature aging efficiency curve and its corresponding historical temperature and humidity data, and to calculate and output the power prediction value of the photovoltaic power generation equipment in a future preset period based on the second historical photovoltaic conversion efficiency data.

Images