Distributed photovoltaic power generation load influence factor analysis processing method and system

Abstract

The invention provides a distributed photovoltaic power generation load influence factor analysis processing method and system, and relates to the technical field of photovoltaic power generation and artificial intelligence, and the method comprises the steps: determining factors influencing a distributed photovoltaic power generation load; acquiring related data of the distributed photovoltaic power generation system to form a sample data set according to the factors influencing the distributed photovoltaic power generation load; training by using the sample data set to obtain a photovoltaic power generation load prediction model; calculating an SHAP value of each influence feature in a to-be-detected data set collected in real time based on the model; on the basis of the SHAP value of each influence feature obtained through calculation, performing importance sorting on each factor influencing the distributed photovoltaic power generation load; and corresponding measures are taken according to the importance degree of each factor influencing the distributed photovoltaic power generation load, so that the energy utilization efficiency is improved. According to the scheme, the importance degree of the factors influencing the distributed photovoltaic power generation load can be analyzed, and then corresponding measures are taken in a targeted mode to improve the energy utilization efficiency.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation and artificial intelligence technology, and in particular to a method and system for analyzing and processing factors affecting distributed photovoltaic power generation load. Background Art

[0002] Today, as energy depletion and energy conservation and environmental protection issues become increasingly serious, distributed power generation has been used more and more widely. With the increasing penetration rate of distributed wind energy and photovoltaics and other new energy sources in the power grid in microgrids, while alleviating energy tension and environmental deterioration, the intermittent and unstable nature of wind and photovoltaic power generation also poses great challenges to the safe, reliable and economic operation of the power grid. Through accurate distributed photovoltaic power generation load forecasting, the future photovoltaic power generation load can be predicted, which can provide a scientific decision-making basis for automatic power generation control and grid dispatching, effectively improve the utilization efficiency of distributed power sources, effectively reduce the impact of high penetration of distributed photovoltaics on the power grid, improve power supply reliability, and ensure the safe, reliable and economic operation of the power grid.

[0003] At present, the power generation control and dispatching of the power grid are only based on the prediction results of photovoltaic power generation load. However, there are many factors that affect the distributed photovoltaic power generation load, such as natural factors, equipment factors, installation and maintenance factors, power grid factors, power grid access capacity factors, etc., and in actual control and dispatching, it is not possible to make every factor in a better or optimal state, and in different environments or at different times, the factors that have the main impact on the distributed power generation load are also different. This leads to the poor rationality of the existing power generation control and dispatching of the power grid based only on the prediction results of power generation load, and low energy utilization efficiency. Summary of the invention

[0004] In view of this, in view of the above factors, it is necessary to propose an analysis and processing method and system for the influencing factors of distributed photovoltaic power generation load, so as to analyze the importance of the factors affecting the distributed photovoltaic power generation load, and then take corresponding measures in a targeted manner to improve energy utilization efficiency.

[0005] In a first aspect, the present invention provides a method for analyzing and processing factors affecting distributed photovoltaic power generation load, comprising:

[0006] Determine the factors that affect the distributed photovoltaic power generation load; each influencing factor includes at least one influencing feature;

[0007] According to the factors affecting the distributed photovoltaic power generation load, relevant data of the distributed photovoltaic power generation system is obtained to form a sample data set; wherein the sample data set includes: a sample training set and a sample test set;

[0008] The photovoltaic power generation load prediction model is trained using the sample training set, and the model is tested and verified using the sample test set;

[0009] Based on the photovoltaic power generation load prediction model, the SHAP value of each influencing feature in the real-time collected test data set is calculated;

[0010] Based on the calculated SHAP value of each influencing feature, the factors affecting distributed photovoltaic power generation load are ranked in importance;

[0011] Take appropriate measures according to the importance of various factors affecting distributed photovoltaic power generation load to improve energy utilization efficiency.

[0012] Preferably, the factors affecting the distributed photovoltaic power generation load include: natural factors, equipment factors, installation and maintenance factors, power grid factors and user demand factors;

[0013] The influencing characteristics of the natural factors include: light intensity and temperature;

[0014] The influencing characteristics under the equipment factors include: PV panel quality and inverter efficiency;

[0015] The influencing features under the installation and maintenance factors include: the installation direction and angle of the photovoltaic panels, the cleaning of the photovoltaic panels, and the maintenance of the photovoltaic panels;

[0016] The influencing characteristics under the said power grid factors include: power grid stability and power grid access capacity;

[0017] The influencing characteristics of the user demand factors include: power load characteristics and energy management strategies.

[0018] Preferably, the obtaining of relevant data of the distributed photovoltaic power generation system includes:

[0019] Collect meteorological data, equipment parameter data, load data, grid performance data and energy strategy data of distributed photovoltaic power generation systems to obtain feature data sets;

[0020] Label the collected data to obtain a label data set;

[0021] The feature data set and the label data set are preprocessed to obtain a sample data set for training the model; wherein the sample data set is divided into a sample training set and a sample test set according to a certain ratio.

[0022] Preferably, the preprocessing of the feature data set and the label data set includes:

[0023] The feature data set and the label data set are cleaned and standardized.

[0024] Preferably, the photovoltaic power generation load prediction model is obtained by training the sample training set, including:

[0025] The photovoltaic power generation load prediction model is obtained by using the sample training set and adopting convolutional neural network or recurrent neural network training.

[0026] Preferably, the step of calculating the SHAP value of each influencing feature in the test data set collected in real time includes:

[0027] The SHAP value of each influencing feature is calculated using the following formula:

[0028]

[0029] Among them, λ i is used to characterize the SHAP value of the i-th influencing feature, Φ is used to characterize the subset of influencing features in the photovoltaic power generation load forecasting model, N is the total number of influencing features, f(Φ) is used to characterize the output value of the photovoltaic power generation load forecasting model on the influencing feature subset Φ, and f(Φ∪{i}) is used to characterize the output value of the photovoltaic power generation load forecasting model on the union of the influencing feature subset Φ and the i-th influencing feature. | Φ | Characterizes the size of Φ.

[0030] Preferably, taking corresponding measures according to the importance of various factors affecting the distributed photovoltaic power generation load includes:

[0031] Set the sorting threshold and perform the following operations for each factor affecting distributed photovoltaic power generation load:

[0032] If the ranking of the light intensity is before the ranking threshold, the automatic lighting angle adjustment device is controlled to start, so that the photovoltaic panel rotates with the sunlight; and the photovoltaic panel cleaning device is controlled to start, and the dirt and dust on the surface of the photovoltaic panel are cleaned;

[0033] If the temperature ranking is before the ranking threshold, a recommendation is issued that the load should be reduced or the equipment should be replaced due to overheating of the equipment;

[0034] If the ranking of the PV panel quality and the inverter efficiency is before the ranking threshold, a prompt is issued that the equipment factor affects the power generation load, so that the background can determine whether it is necessary to replace the appropriate PV panel and inverter;

[0035] If the installation direction and angle of the photovoltaic panels, or the ranking of the photovoltaic panel maintenance is before the ranking threshold, a suggestion for on-site maintenance of the photovoltaic panels is sent to the operation and maintenance personnel;

[0036] If the ranking of the photovoltaic panel cleaning is before the ranking threshold, the photovoltaic panel cleaning device is controlled to start to clean the dirt and dust on the surface of the photovoltaic panel;

[0037] If the ranking of the grid stability characteristics is before the ranking threshold, a recommendation is issued to install a grid stabilization device or optimize the grid structure;

[0038] If the ranking of the grid access capacity is before the ranking threshold, a suggestion is issued to reconfigure the capacity reasonably according to the access capacity of the local grid;

[0039] If the power load characteristics or energy management strategy are ahead of the sorting threshold, a recommendation is given to the user to install energy storage equipment or optimize the operating status of the power equipment.

[0040] In a second aspect, the present invention provides a system for analyzing and processing factors affecting distributed photovoltaic power generation load, the system comprising: a distributed photovoltaic power generation system background, an automatic lighting angle adjustment device, and a photovoltaic panel cleaning device; the distributed photovoltaic power generation system background is respectively communicated with the automatic lighting angle adjustment device and the photovoltaic panel cleaning device;

[0041] The distributed photovoltaic power generation system backend is configured to execute the analysis and processing method of the factors affecting the distributed photovoltaic power generation load as described in the first aspect; and is also configured to issue a start or stop instruction to the automatic lighting angle adjustment device and the photovoltaic panel cleaning device;

[0042] The automatic lighting angle adjustment device is configured to be activated according to the instruction issued by the background of the distributed photovoltaic power generation system, so that the photovoltaic panel rotates with the sunlight to increase the amount of light;

[0043] The photovoltaic panel cleaning device is configured to start according to the instructions issued by the background of the distributed photovoltaic power generation system to clean the dirt and dust on the surface of the photovoltaic panel.

[0044] Preferably, the system further comprises: a visualization module; the visualization module is used to intuitively and in real time display the ranking of the degree of influence of different factors on the power generation load, as well as the suggestions and measures given.

[0045] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute any of the methods described in the first aspect.

[0046] It can be seen from the above technical scheme that the present invention provides a method and system for analyzing and processing factors affecting distributed photovoltaic power generation load. In this scheme, when analyzing and processing factors affecting distributed photovoltaic power generation load, the factors affecting distributed photovoltaic power generation load are first determined, and then the relevant data of the distributed photovoltaic power generation system is obtained according to the determined factors affecting the distributed power generation load, forming a sample data set including a sample training set and a sample test set; further, the photovoltaic power generation load prediction model is obtained by training the sample data set; further, based on the photovoltaic power generation load prediction model obtained by training, the SHAP value of each influencing feature in the test data set collected in real time is calculated, so that the importance of each influencing factor can be determined according to the size of the SHAP value of each influencing feature, and then corresponding measures can be taken in a targeted manner to control and dispatch the power grid for power generation. It can be seen that this scheme can accurately and explainably analyze the impact of each influencing factor on the distributed photovoltaic power generation load by calculating the SHAP value of each influencing feature, and then take corresponding measures according to the importance of each influencing factor to reasonably control and dispatch the power grid for power generation, so that energy can be fully utilized and the efficiency of energy utilization can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flow chart of a method for analyzing and processing factors affecting distributed photovoltaic power generation load provided by an embodiment of the present invention.

[0048] Figure 2 A schematic diagram of a system for analyzing and processing factors affecting distributed photovoltaic power generation load provided by an embodiment of the present invention.

[0049] In the figure: a distributed photovoltaic power generation system background 10, an automatic lighting angle adjustment device 20, a photovoltaic panel cleaning device 30, and a visualization module 40. DETAILED DESCRIPTION

[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0051] like Figure 1 As shown, the present invention provides a method for analyzing and processing factors affecting distributed photovoltaic power generation load, the method comprising the following steps:

[0052] Step 101: determining factors affecting distributed photovoltaic power generation load; wherein each influencing factor includes at least one influencing feature;

[0053] Step 102: according to the factors affecting the distributed photovoltaic power generation load, obtain the relevant data of the distributed photovoltaic power generation system to form a sample data set; wherein the sample data set includes: a sample training set and a sample test set;

[0054] Step 103: train a photovoltaic power generation load prediction model using the sample training set, and test and verify the model using the sample test set;

[0055] Step 104: Based on the photovoltaic power generation load prediction model, the SHAP value of each influencing feature in the real-time collected test data set is calculated;

[0056] Step 105: based on the calculated SHAP value of each influencing feature, ranking the factors affecting the distributed photovoltaic power generation load in terms of importance;

[0057] Step 106: Take corresponding measures according to the importance of various factors affecting the distributed photovoltaic power generation load to improve energy utilization efficiency.

[0058] In this embodiment, when analyzing and processing the factors affecting the distributed photovoltaic power generation load, the factors affecting the distributed photovoltaic power generation load are first determined, and then the relevant data of the distributed photovoltaic power generation system is obtained according to the determined factors affecting the distributed power generation load, forming a sample data set including a sample training set and a sample test set; further, the photovoltaic power generation load prediction model is obtained by training using the sample data set; further, based on the photovoltaic power generation load prediction model obtained by training, the SHAP value of each influencing feature in the real-time collected test data set is calculated, so that the importance of each influencing factor can be determined according to the size of the SHAP value of each influencing feature, and then corresponding measures can be taken in a targeted manner to control and dispatch the power generation of the power grid. It can be seen that this scheme can accurately and explainably analyze the impact of each influencing factor on the distributed photovoltaic power generation load by calculating the SHAP value of each influencing feature, and then take corresponding measures according to the importance of each influencing factor to reasonably control and dispatch the power generation of the power grid, so that energy can be fully utilized and the efficiency of energy utilization can be improved. Moreover, in actual power generation control and scheduling, it is no longer necessary to adopt a fixed method to deal with changes in power generation load caused by various factors, such as optimizing only one or two factors, nor is it necessary to optimize each influencing factor. Instead, it is necessary to dynamically optimize and control the impact of various factors on photovoltaic power generation load. This can not only effectively solve the problems of grid voltage fluctuations, harmonic pollution, power factor reduction, voltage quality mutations, etc. caused by various influencing factors, but also the cost of the measures taken is lower.

[0059] For step 101, factors affecting distributed photovoltaic power generation load are determined; wherein each influencing factor includes at least one influencing feature;

[0060] In this step, the factors affecting distributed photovoltaic power generation load may mainly include: natural factors, equipment factors, installation and maintenance factors, power grid factors and user demand factors;

[0061] For the influencing characteristics under natural factors, they can mainly include light intensity and temperature. Light intensity is the most direct factor affecting distributed photovoltaic power generation. The stronger the light, the more electricity the photovoltaic panels generate. On sunny days, the light is sufficient and the photovoltaic power generation is usually high; on cloudy, cloudy or hazy weather conditions, the light intensity is weakened and the power generation will drop significantly. The light resources in different regions vary greatly. For example, the light resources in the northwest of my country are abundant, while those in the eastern region are relatively less. Therefore, when evaluating the distributed photovoltaic power generation load, it is necessary to consider the local light intensity and its changing law. At the same time, the power generation efficiency of photovoltaic panels is also affected by temperature. Generally speaking, an increase in temperature will cause the output power of photovoltaic cells to decrease. This is because as the temperature increases, the internal resistance of photovoltaic cells increases, thereby reducing the current and voltage. In practical applications, high temperatures in summer may reduce the photovoltaic power generation load, while lower temperatures in winter may have different degrees of impact on the photovoltaic power generation load. In addition, different types of photovoltaic cells have different sensitivities to temperature.

[0062] For the influencing characteristics under the equipment factor, it can mainly include the quality of photovoltaic panels and the efficiency of inverters. The quality of photovoltaic panels directly determines its power generation efficiency and reliability. High-quality photovoltaic panels have higher conversion efficiency, longer service life and better stability. When selecting photovoltaic panels, factors such as their brand, technical parameters, and quality certification need to be considered. The performance of photovoltaic panels produced by different manufacturers may vary. Therefore, when evaluating distributed photovoltaic power generation loads, the specific quality of the photovoltaic panels used needs to be considered. In addition, since the inverter is the key equipment for converting the direct current generated by photovoltaic panels into alternating current, the efficiency of the inverter will affect the output power of the entire photovoltaic power generation system. A high-efficiency inverter can convert more direct current into alternating current, thereby increasing the photovoltaic power generation load. The performance of the inverter is also affected by factors such as its capacity, input voltage range, and output waveform quality. When selecting an inverter, it is necessary to select a suitable inverter model and parameters based on the scale and requirements of the photovoltaic power generation system.

[0063] For the influencing characteristics under the installation and maintenance factors, they can mainly include the installation direction and angle of photovoltaic panels, photovoltaic panel cleaning and photovoltaic panel maintenance. First, the installation angle and direction of photovoltaic panels have a great influence on the intensity and time of light they receive. Generally speaking, photovoltaic panels should be installed in the south-facing direction, and the angle with the horizontal plane should be adjusted according to the local latitude to maximize the reception of sunlight. Unreasonable installation angles and directions will lead to a reduction in photovoltaic power generation load. For example, if the photovoltaic panel is installed at an angle that is too large or too small, the light receiving area may be reduced, thereby reducing the power generation. In addition, the deviation of the installation direction may also affect the photovoltaic power generation load. Secondly, regular cleaning of photovoltaic panels can maintain their good performance. Dust, dirt, leaves and other debris will cover the surface of photovoltaic panels, reducing their ability to receive light, thereby affecting the power generation. Furthermore, regular maintenance of photovoltaic panels can also help maintain the good performance of distributed photovoltaic systems. For example, check whether the cable connection is loose, check the operating status of equipment such as inverters, etc. Timely discovery and resolution of problems can ensure the stable operation of the photovoltaic power generation system and increase the photovoltaic power generation load.

[0064] For the influencing characteristics under the power grid factor, it can mainly include power grid stability and power grid access capacity. On the one hand, the distributed photovoltaic power generation system needs to be connected to the power grid to realize the output and use of electric energy. The stability of the power grid has an important impact on the photovoltaic power generation load. If the power grid voltage fluctuates greatly, the frequency is unstable, or there are harmonics, it may affect the normal operation of the photovoltaic power generation system and even cause equipment damage. In order to ensure the stable operation of the distributed photovoltaic power generation system, corresponding measures need to be taken, such as installing power grid stabilization devices, optimizing the power grid structure, etc., to improve the stability of the power grid. On the other hand, the power grid access capacity refers to the maximum capacity of the distributed photovoltaic power generation system that can be connected to the power grid. If the capacity of the photovoltaic power generation system exceeds the power grid access capacity, it may cause problems such as power grid overload, voltage increase or frequency instability. When planning and designing distributed photovoltaic power generation systems, it is necessary to consider the access capacity of the local power grid and make reasonable capacity configuration according to the actual situation to ensure the safe and stable operation of the photovoltaic power generation system.

[0065] The influencing characteristics under the user demand factor can mainly include power load characteristics and energy management strategies. On the one hand, the power load characteristics of users will affect the load of distributed photovoltaic power generation. If the power load of users matches the output power curve of photovoltaic power generation, a higher proportion of self-generation and self-use can be achieved, reducing dependence on the power grid. For example, users with large power consumption during the day, such as commercial buildings and factories, are more matched with the output power curve of distributed photovoltaic power generation, and can make better use of photovoltaic power generation. Users with large power consumption at night, such as residential households, may need to rely more on power grid power supply. On the other hand, reasonable energy management strategies can also improve the utilization efficiency of distributed photovoltaic power generation. Users can achieve effective utilization of photovoltaic power generation by installing energy storage equipment and optimizing the operating time of power consumption equipment. For example, installing energy storage equipment can store electricity when photovoltaic power generation is in excess, and release electricity during peak power consumption or when photovoltaic power generation is insufficient, thereby improving energy utilization efficiency. In addition, users can also use intelligent control systems to automatically adjust the operating status of power consumption equipment according to photovoltaic power generation and power demand to achieve optimal energy management.

[0066] Step 102, according to the factors affecting the distributed photovoltaic power generation load, obtain the relevant data of the distributed photovoltaic power generation system to form a sample data set; wherein the sample data set includes: a sample training set and a sample test set;

[0067] In this step, when obtaining the relevant data of the distributed photovoltaic power generation system to form a sample data set, it can be achieved in the following ways:

[0068] Collect meteorological data, equipment parameter data, load data, grid performance data and energy strategy data of distributed photovoltaic power generation systems to obtain feature data sets;

[0069] Label the collected data to obtain a label data set;

[0070] The feature data set and the label data set are preprocessed to obtain a sample data set for training the model; wherein the sample data set is divided into a sample training set and a sample test set according to a certain ratio.

[0071] In this embodiment, the relevant data of the distributed photovoltaic power generation system can be first collected, such as meteorological data containing information such as light intensity, temperature, humidity, etc., equipment parameter information containing parameters such as photovoltaic panel type, inverter type, photovoltaic panel installation angle, and grid performance data containing information such as grid stability and access capacity, power load characteristic data and energy management strategy data, and these data are used to form a feature data set. Then, each collected data is labeled to obtain a label data set. In this way, after preprocessing the feature data set and the label data set, a sample data set for training the model can be obtained. When preprocessing the data, the data can be cleaned first to eliminate the outliers in the data, thereby ensuring the reliability of the sample data set. After data cleaning, the data should be standardized, such as processing the data in a normalized manner to convert the data into a form that can be input into the model. After obtaining the sample data set, the sample data set is divided into a sample training set and a sample test set according to a certain ratio. For example, it is divided in a ratio of 6:4.

[0072] Step 103, using the sample training set to train a photovoltaic power generation load prediction model, and using the sample test set to test and verify the model;

[0073] In this step, consider building a photovoltaic power generation load prediction model based on a deep neural network. For example, using a sample training set, a convolutional neural network or a recurrent neural network is used to train a photovoltaic power generation load prediction model, and then the sample test set is used to test and verify the performance of the trained model to ensure the reliability of the trained model.

[0074] Step 104, based on the photovoltaic power generation load prediction model, calculate the SHAP value of each influencing feature in the real-time collected test data set;

[0075] In this step, the DeepSHAP method is considered to be applied to the trained photovoltaic power generation load prediction model to quantitatively analyze the contribution of each influencing feature to the photovoltaic power generation load. The SHAP value is based on the concept of Shapley value. Its core idea is to decompose the prediction result of the model into the sum of the contribution of each feature. By calculating the marginal contribution of each feature to the prediction result, its importance is determined, which can help understand why the model makes a specific prediction. That is, the prediction result of the model can be explained by assigning an importance value to each influencing feature, thereby realizing the analysis of the importance of each influencing factor.

[0076] In one embodiment, when calculating the SHAP value of each influencing feature in the test data set collected in real time, it can be calculated by the following calculation formula:

[0077]

[0078] Among them, λ i The SHAP value used to characterize the i-th influencing feature indicates the degree of influence of the influencing feature. Φ is used to characterize the subset of influencing features in the photovoltaic power generation load forecasting model. N is the total number of influencing features. f(Φ) is used to characterize the output value of the photovoltaic power generation load forecasting model on the influencing feature subset Φ. f(Φ∪{i}) is used to characterize the output value of the photovoltaic power generation load forecasting model on the union of the influencing feature subset Φ and the i-th influencing feature. | Φ | represents the size of Φ, ! represents the factorial, i = {1,…,N}.

[0079] Step 105, ranking the factors affecting the distributed photovoltaic power generation load in terms of importance based on the calculated SHAP value of each influencing feature;

[0080] In this step, the SHAP values ​​of each influencing feature are calculated, and the influencing features of each factor affecting the distributed photovoltaic power generation load are ranked in order of importance. For example, the ranking of the degree of influence of different influencing factors on the power generation load can be intuitively displayed by drawing a bar chart, scatter plot, list, etc. of the SHAP value, and the interaction relationship between the factors can be analyzed at the same time. For example, if the SHAP values ​​of two factors show a synergistic change trend in some cases, it means that there may be a synergistic relationship between them that affects the power generation load.

[0081] Step 106: according to the importance of various factors affecting the distributed photovoltaic power generation load, take corresponding measures to control and dispatch the power generation of the power grid.

[0082] In this step, consider taking corresponding measures according to the importance of each factor. For example, you can set a sorting threshold and perform the following operations for each factor that affects the distributed photovoltaic power generation load:

[0083] If the ranking of the light intensity is before the ranking threshold, the automatic lighting angle adjustment device is controlled to start, so that the photovoltaic panel rotates with the sunlight; and the photovoltaic panel cleaning device is controlled to start, and the dirt and dust on the surface of the photovoltaic panel are cleaned, so as to improve the photovoltaic power generation efficiency. Of course, in practice, the start-up cycle of the automatic lighting angle adjustment device and the photovoltaic panel cleaning device can also be optimized.

[0084] If the temperature ranking is before the ranking threshold, a recommendation is issued that the load should be reduced or the equipment should be replaced due to overheating of the equipment;

[0085] If the ranking of the quality of photovoltaic panels and the efficiency of inverters is before the ranking threshold, a reminder that the equipment factors affect the power generation load will be issued, so that the background can determine whether it is necessary to replace the appropriate photovoltaic panels and inverters; in this way, after the reminder is issued, personnel can continue to pay attention to the development of photovoltaic power generation technology, and improve the performance of photovoltaic power generation systems and their ability to resist meteorological factors by actively introducing and applying new technologies and products, such as new photovoltaic materials, efficient inverters and their technologies, and intelligent control systems.

[0086] If the installation direction and angle of the photovoltaic panels, or the ranking of the photovoltaic panel maintenance is before the ranking threshold, a suggestion for on-site maintenance of the photovoltaic panels is sent to the operation and maintenance personnel;

[0087] If the ranking of the photovoltaic panel cleaning is before the ranking threshold, the photovoltaic panel cleaning device is controlled to start to clean the dirt and dust on the surface of the photovoltaic panel;

[0088] If the ranking of the grid stability characteristics is before the ranking threshold, a recommendation is issued to install a grid stabilization device or optimize the grid structure;

[0089] If the ranking of the grid access capacity is before the ranking threshold, a suggestion is issued to reconfigure the capacity reasonably according to the access capacity of the local grid;

[0090] If the power load characteristics or energy management strategy are ahead of the sorting threshold, a recommendation is given to the user to install energy storage equipment or optimize the operating status of the power equipment.

[0091] It is easy to understand that for the multiple influencing factors before the sorting threshold, corresponding measures should be taken for each influencing factor to control and dispatch power generation of the power grid as reasonably as possible and improve energy utilization efficiency.

[0092] like Figure 2 As shown, the present invention also provides a distributed photovoltaic power generation load influencing factor analysis and processing system, the system comprises: a distributed photovoltaic power generation system background 10, an automatic lighting angle adjustment device 20 and a photovoltaic panel cleaning device 30; the distributed photovoltaic power generation system background 10 is respectively connected to the automatic lighting angle adjustment device 20 and the photovoltaic panel cleaning device 30 for communication;

[0093] The distributed photovoltaic power generation system backend 10 is configured to execute the analysis and processing method of the factors affecting the distributed photovoltaic power generation load as described in the above embodiments; and is also configured to issue a start or stop instruction to the automatic lighting angle adjustment device 20 and the photovoltaic panel cleaning device 30;

[0094] The automatic lighting angle adjustment device 20 is configured to be activated according to the instruction issued by the distributed photovoltaic power generation system backend 10, so that the photovoltaic panel rotates with the sunlight to increase the amount of light;

[0095] The photovoltaic panel cleaning device 30 is configured to start according to the instruction issued by the distributed photovoltaic power generation system background 10 to clean the dirt and dust on the surface of the photovoltaic panel.

[0096] In one embodiment, the system further includes: a visualization module 40; the visualization module 40 is used to intuitively and in real time display the ranking of the degree of influence of different factors on the power generation load, as well as the suggestions and measures given.

[0097] In summary, the method and system for analyzing and processing factors affecting distributed photovoltaic power generation load provided by the present invention have at least the following beneficial effects:

[0098] (1) Based on the SHAP value, the accuracy of the analysis of factors affecting distributed photovoltaic power generation load is improved, and it helps users understand the prediction results of the model.

[0099] (2) This scheme realizes accurate analysis of the factors affecting photovoltaic power generation load, can provide a scientific basis for energy system planning and operation, and help improve energy utilization efficiency.

[0100] (3) This solution effectively solves the problems of grid voltage fluctuation, harmonic pollution, power factor reduction, and voltage quality mutation caused by load influence.

[0101] (4) This solution can present the results of quantitative analysis in the form of intuitive charts, such as a bar chart showing the importance ranking of different influencing factors (SHAP value size), or a line chart showing the change of the SHAP value of a certain influencing factor over time or other variables.

[0102] (5) This solution has a deep understanding of the impact of factors; it can accurately quantify the contribution of each influencing factor to the distributed photovoltaic power generation load. For example, it can clearly indicate how much influence factors such as temperature, light intensity, and equipment parameters have on the power generation load under different circumstances, which helps researchers and engineers to gain a deeper understanding of the operating mechanism of the photovoltaic power generation system.

[0103] (6) Revealing complex relationships: It has a good ability to reveal the complex interactive relationships between multiple influencing factors. In actual distributed photovoltaic power generation scenarios, the relationships between various factors are not simple linear relationships. This solution can analyze the impact of the synergistic or antagonistic effects between light intensity and photovoltaic panel temperature on power generation load.

[0104] (7) This solution can optimize the distributed photovoltaic power generation system in a targeted manner by accurately identifying key influencing factors and their degree of influence. For example, if it is found that dust accumulation in a certain area has a greater impact on the power generation load, the cleaning cycle can be optimized; if it is a light blocking problem within a specific time period, the layout or installation angle of the photovoltaic panels can be adjusted to improve the overall power generation efficiency.

[0105] (8) Improved reliability: It helps to evaluate the impact of different factors on the stability of power generation load. After understanding the quantitative impact of factors such as extreme weather conditions and equipment aging on power generation load, measures can be taken in advance, such as strengthening equipment maintenance and equipping energy storage equipment, to improve the power supply reliability of distributed photovoltaic power generation systems.

[0106] (9) Planning and decision support: reasonable site selection and capacity planning; in the planning stage of distributed photovoltaic power generation projects, this quantitative analysis method can provide a basis for site selection and determination of installed capacity. For example, by analyzing the impact of factors in different geographical locations (such as local climate conditions, geographical environment, etc.) on power generation load, a more suitable construction site can be selected, a more reasonable photovoltaic power generation capacity can be determined, and over-investment or insufficient power generation can be avoided.

[0107] (10) Grid access and dispatching decisions: For grid operators, understanding the factors affecting distributed photovoltaic power generation load helps to better manage grid access and make dispatching decisions. Based on the prediction of power generation load under the influence of different factors, the dispatching plan of the grid can be reasonably arranged to ensure the stable operation of the grid.

[0108] The present specification also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed in a computer, the computer is caused to execute a method in any one of the embodiments in the specification.

[0109] The present specification also provides a computing device, including a memory and a processor, wherein executable codes are stored in the memory, and when the processor executes the executable codes, a method in any embodiment of the present specification is implemented.

[0110] Since the system embodiment provided by the present invention is based on the same inventive concept as the method embodiment of this specification, the specific content can be found in the description of the method embodiment of this specification, and will not be repeated here.

[0111] The modules or units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. The above disclosure is only the preferred embodiment of the present invention, and of course it cannot be used to limit the scope of the rights of the present invention. Those skilled in the art can understand that all or part of the processes of the above embodiment are implemented, and the equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for analyzing and processing factors affecting distributed photovoltaic power generation load, characterized in that: include: Determine the factors that affect the distributed photovoltaic power generation load; each influencing factor includes at least one influencing feature; According to the factors affecting the distributed photovoltaic power generation load, relevant data of the distributed photovoltaic power generation system is obtained to form a sample data set; wherein the sample data set includes: a sample training set and a sample test set; The photovoltaic power generation load prediction model is trained using the sample training set, and the model is tested and verified using the sample test set; Based on the photovoltaic power generation load prediction model, the SHAP value of each influencing feature in the real-time collected test data set is calculated; Based on the calculated SHAP value of each influencing feature, the factors affecting distributed photovoltaic power generation load are ranked in importance; Take corresponding measures according to the importance of each factor affecting distributed photovoltaic power generation load to improve energy utilization efficiency.

2. The method for analyzing and processing factors affecting distributed photovoltaic power generation load according to claim 1, characterized in that: The factors affecting distributed photovoltaic power generation load include: natural factors, equipment factors, installation and maintenance factors, power grid factors and user demand factors; The influencing characteristics of the natural factors include: light intensity and temperature; The influencing characteristics under the equipment factors include: PV panel quality and inverter efficiency; The influencing features under the installation and maintenance factors include: the installation direction and angle of the photovoltaic panels, the cleaning of the photovoltaic panels, and the maintenance of the photovoltaic panels; The influencing characteristics under the said power grid factors include: power grid stability and power grid access capacity; The influencing characteristics of the user demand factors include: power load characteristics and energy management strategies.

3. The method for analyzing and processing factors affecting distributed photovoltaic power generation load according to claim 1, characterized in that: The obtaining of the relevant data of the distributed photovoltaic power generation system includes: Collect meteorological data, equipment parameter data, load data, grid performance data and energy strategy data of distributed photovoltaic power generation systems to obtain feature data sets; Label the collected data to obtain a label data set; The feature data set and the label data set are preprocessed to obtain a sample data set for training the model; wherein the sample data set is divided into a sample training set and a sample test set according to a certain ratio.

4. The method for analyzing and processing the factors affecting the distributed photovoltaic power generation load according to claim 3 is characterized in that: The preprocessing of the feature data set and the label data set includes: The feature data set and the label data set are cleaned and standardized.

5. The method for analyzing and processing factors affecting distributed photovoltaic power generation load according to claim 1, characterized in that: The photovoltaic power generation load prediction model is obtained by training the sample training set, including: The photovoltaic power generation load prediction model is obtained by using the sample training set and adopting convolutional neural network or recurrent neural network training.

6. The method for analyzing and processing factors affecting distributed photovoltaic power generation load according to claim 1, characterized in that: The calculation of the SHAP value of each influencing feature in the test data set collected in real time includes: The SHAP value of each influencing feature is calculated using the following formula: Among them, λ i It is used to characterize the SHAP value of the i-th influencing feature, Φ is used to characterize the subset of influencing features in the photovoltaic power generation load forecasting model, N is the total number of influencing features, f(Φ) is used to characterize the output value of the photovoltaic power generation load forecasting model on the influencing feature subset Φ, f(Φ∪{i}) is used to characterize the output value of the photovoltaic power generation load forecasting model on the union of the influencing feature subset Φ and the i-th influencing feature, and |Φ| characterizes the size of Φ.

7. The method for analyzing and processing factors affecting distributed photovoltaic power generation load according to claim 2, characterized in that: The corresponding measures are taken according to the importance of various factors affecting the distributed photovoltaic power generation load, including: Set the sorting threshold and perform the following operations for each factor affecting distributed photovoltaic power generation load: If the ranking of the light intensity is before the ranking threshold, the automatic lighting angle adjustment device is controlled to start, so that the photovoltaic panel rotates with the sunlight; and the photovoltaic panel cleaning device is controlled to start, and the dirt and dust on the surface of the photovoltaic panel are cleaned; If the temperature ranking is before the ranking threshold, a recommendation is issued that the load should be reduced or the equipment should be replaced due to overheating of the equipment; If the ranking of the PV panel quality and the inverter efficiency is before the ranking threshold, a prompt is issued that the equipment factor affects the power generation load, so that the background can determine whether it is necessary to replace the appropriate PV panel and inverter; If the installation direction and angle of the photovoltaic panels, or the ranking of the photovoltaic panel maintenance is before the ranking threshold, a suggestion for on-site maintenance of the photovoltaic panels is sent to the operation and maintenance personnel; If the ranking of the photovoltaic panel cleaning is before the ranking threshold, the photovoltaic panel cleaning device is controlled to start to clean the dirt and dust on the surface of the photovoltaic panel; If the ranking of the grid stability characteristics is before the ranking threshold, a recommendation is issued to install a grid stabilization device or optimize the grid structure; If the ranking of the grid access capacity is before the ranking threshold, a suggestion is issued to reconfigure the capacity reasonably according to the access capacity of the local grid; If the power load characteristics or energy management strategy are ahead of the sorting threshold, a recommendation is given to the user to install energy storage equipment or optimize the operating status of the power equipment.

8. A system for analyzing and processing factors affecting distributed photovoltaic power generation load, characterized in that: The system includes: a distributed photovoltaic power generation system background, an automatic lighting angle adjustment device and a photovoltaic panel cleaning device; the distributed photovoltaic power generation system background is respectively connected to the automatic lighting angle adjustment device and the photovoltaic panel cleaning device for communication; The distributed photovoltaic power generation system backend is configured to execute the analysis and processing method of the factors affecting the distributed photovoltaic power generation load as described in claims 1-7; and is also configured to issue a start or stop instruction to the automatic lighting angle adjustment device and the photovoltaic panel cleaning device; The automatic lighting angle adjustment device is configured to start according to the instruction issued by the background of the distributed photovoltaic power generation system, so that the photovoltaic panel rotates with the sunlight to increase the amount of light; The photovoltaic panel cleaning device is configured to start according to the instructions issued by the background of the distributed photovoltaic power generation system to clean the dirt and dust on the surface of the photovoltaic panel.

9. The analysis and processing system for factors affecting distributed photovoltaic power generation load according to claim 8, characterized in that: The system also includes: a visualization module; the visualization module is used to intuitively and in real time display the ranking of the degree of influence of different factors on the power generation load, as well as the suggestions and measures given.

10. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method according to any one of claims 1 to 7.

Images