Photovoltaic power generation range prediction method considering photovoltaic output and meteorological factors

Abstract

The present invention provides a photovoltaic power generation interval prediction method considering photovoltaic output and meteorological elements, and relates to the technical field of photovoltaic power generation. The present invention collects multiple groups of photovoltaic components' own parameters and historical environmental power generation parameters, and collects the photovoltaic components' own parameters and environmental power generation parameters to be predicted in real time. Through the mapping and analysis of these data, it is ensured that the prediction model can fully reflect the actual operation status of the photovoltaic power generation system, and the meteorological data and power generation data in the reference data set and each entity data set are subjected to time series feature analysis and extreme event feature analysis, thereby realizing the effective fusion of multivariate data. The similarity between the reference data set and the entity data set can be quantified through the generated meteorological coincidence index and power generation trend coincidence index, and the reference entity most relevant to the photovoltaic components to be predicted is screened out. The power generation power prediction interval is generated by combining the photovoltaic components' own parameters to be predicted and the power generation data of the reference entity.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic power generation power interval prediction method taking photovoltaic output and meteorological factors into consideration. Background Art

[0002] In the context of global response to climate change and promotion of renewable energy development, photovoltaic power generation is developing rapidly as a clean and sustainable form of energy. Countries have increased their investment and application in photovoltaic power generation to reduce dependence on traditional fossil energy and slow down greenhouse gas emissions. However, the output power of photovoltaic power generation is affected by many external factors, especially meteorological conditions, such as light intensity, temperature, wind speed, etc. Changes in these factors will significantly affect the power generation efficiency of photovoltaic modules. Therefore, accurate prediction of photovoltaic power generation is of great significance for the stable operation and economic dispatch of power systems.

[0003] Existing technologies mainly use prediction models based on single meteorological factors or historical power generation data, such as linear regression and time series analysis. These methods often ignore the impact of photovoltaic module parameters and the interaction of multiple meteorological factors, which increases the risk of power dispatch.

[0004] In the existing technology, photovoltaic power generation prediction mainly relies on historical power generation data and simple meteorological parameters such as light intensity and temperature. These methods usually use statistical models or machine learning algorithms to predict future power generation, but often ignore the complexity and variability of meteorological conditions, especially the direct impact of extreme meteorological events on power generation. This one-sided data utilization method leads to a relatively rough prediction of photovoltaic power generation capacity, which often fails to reflect the real power generation potential and changing trends. It is impossible to provide accurate predictions. In addition, the existing technology lacks effective feature extraction and data fusion strategies when processing multivariate data, and the generalization ability in actual prediction is insufficient.

[0005] In summary, the existing technology has certain limitations in photovoltaic power generation prediction. A new method is needed that can comprehensively consider the parameters of photovoltaic modules themselves, meteorological elements and time series characteristics to improve the accuracy and reliability of prediction.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a photovoltaic power generation range prediction method taking into account photovoltaic output and meteorological factors, so as to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A photovoltaic power generation range prediction method considering photovoltaic output and meteorological factors, the specific steps include:

[0010] Step 1: Collect multiple sets of PV module parameters and historical environmental power generation parameters. The environmental power generation parameters include continuous power generation data and meteorological data with timestamps when the PV modules are generating electricity. The continuous historical power generation parameters and historical meteorological parameters are divided into equal time intervals. The data in each time interval is mapped with the parameters to form an entity data set.

[0011] Step 2: Collect the parameters of the photovoltaic module to be predicted and the environmental power generation parameters of the photovoltaic module in a period before the current moment in real time, and map the parameters of the photovoltaic module to be predicted and the environmental power generation parameters to form a reference data set, and number each entity data set;

[0012] Step 3: Perform time series feature analysis and extreme event feature analysis on the meteorological data in the reference dataset and each entity dataset, obtain meteorological sequence feature data and event feature data of the reference dataset and each entity dataset, perform time series feature analysis on the power generation data in the reference dataset and each entity dataset, and obtain power generation sequence feature data of the reference dataset and each entity dataset;

[0013] Step 4: Based on the meteorological sequence feature data, event feature data and power generation sequence feature data of the reference dataset and the entity dataset, all entity datasets are traversed to generate the meteorological coincidence index and power generation trend coincidence index between the reference dataset and each entity dataset;

[0014] Step 5: Filter out the entity data set with the smallest meteorological coincidence index error and the entity data set with the smallest power generation trend coincidence index with the reference data set respectively, and calibrate them as the first reference entity and the second reference entity. By collecting the own parameters of the photovoltaic component to be predicted and combining the own parameters of the first reference entity and the second reference entity, correct the power generation data of the first reference entity and the second reference entity to obtain the prediction interval of the power generation power of the photovoltaic component to be predicted.

[0015] Furthermore, the timestamps of the power generation data and meteorological data are synchronized accordingly, and the length of each time interval is the same. If the power generation data and meteorological data are not enough to be divided into one time interval, they will be discarded. Each time interval includes the power generation data fragment and the meteorological data fragment located in the time interval, and the timestamps of the power generation data fragment and the meteorological data fragment in each time interval are also synchronized accordingly, and are mapped with the static parameters of the corresponding photovoltaic components to form an entity data set.

[0016] Furthermore, the photovoltaic module's own parameters include the area of ​​the photovoltaic module's solar panel and the standard photoelectric conversion efficiency, the power generation data is the output power of the photovoltaic module, and the meteorological data includes the temperature and wind speed of the photovoltaic module's environment and the light intensity data on the photovoltaic module's surface.

[0017] Furthermore, the reference data set is:

[0018]

[0019] Among them, Ck s represents the reference data set, Yc g (zs) represents the parameters of the photovoltaic module to be predicted, represents the power generation data of the photovoltaic module to be predicted from time j-1 to time j, represents the meteorological data of the photovoltaic module to be predicted from time j-1 to time j, where time j is the current time, time j-1 is the time traced back to a period of time before the current time, and j-(j-1) = ω0, ω0 represents the length of the traced time period;

[0020] The entity dataset is:

[0021]

[0022] Among them, Sts i Represents the i-th entity data set, i represents the number of the entity data set, Stg i (zs) and represents the parameters of the PV module in the i-th entity data set, and They respectively represent the power generation data and meteorological data of the PV modules from time k-1 to time k in the ith entity data set. k-1 and k respectively represent the starting time value and the ending time value of the time interval corresponding to the entity data set. The time span of the time interval is ξ=k-(k-1), and ω≥ω0.

[0023] Furthermore, the method for obtaining the meteorological sequence characteristic data of the entity data set is based on:

[0024] After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set, the average and standard deviation of the normalized temperature data, the average and standard deviation of the wind speed data, and the average and standard deviation of the light intensity data are calculated, and then the meteorological sequence characteristic data is generated. The formula is as follows:

[0025]

[0026] Among them, Sqt irepresents the meteorological sequence feature data of the i-th entity data set, Pjg i and σg i Respectively represent the mean and standard deviation of the surface light intensity data of the photovoltaic module in the i-th entity data set, Pjw i and σw i Respectively represent the mean and standard deviation of the temperature data in the i-th entity data set, Pjf i and σf i They represent the mean and standard deviation of the wind speed data in the i-th entity data set, respectively. α, β, and γ represent the light intensity weight, temperature weight, and wind speed weight, respectively. α>β>γ>0, and α+β+γ=1;

[0027] The method for obtaining event feature data of entity dataset is as follows:

[0028] For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set after normalization, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows:

[0029]

[0030] Among them, Sst i Represents the event feature data of the i-th entity dataset, Jcg i , Jcw i and Jcf i They respectively represent the extreme values ​​of the light intensity data, temperature data and wind speed data on the surface of the photovoltaic module in the i-th entity data set.

[0031] Furthermore, the method for obtaining the power generation sequence characteristic data of the entity data set is:

[0032] The output power data of the photovoltaic modules in the entity data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated. The formula is as follows:

[0033]

[0034] Among them, Sgt. i represents the power generation sequence characteristic data of the i-th entity data set, Pjp i and σp i Respectively represent the mean and standard deviation of the output power data in the i-th entity data set.

[0035] Furthermore, the method for obtaining the meteorological sequence characteristic data of the reference data set is based on:

[0036] After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic modules in each reference data set, the mean and standard deviation of the normalized temperature data, the mean and standard deviation of the wind speed data, and the mean and standard deviation of the light intensity data are calculated, and the meteorological sequence characteristic data is generated based on the following formula:

[0037]

[0038] Among them, Sqt c Represents the meteorological sequence characteristic data of the reference data set, Pjg c and σg c Respectively represent the mean and standard deviation of the surface light intensity data of photovoltaic modules in the reference data set, Pjw c and σw c Respectively represent the mean and standard deviation of the temperature data in the reference data set, Pjf c and σf c Respectively represent the mean and standard deviation of wind speed data in the reference data set;

[0039] The method for obtaining event feature data of the reference data set is as follows:

[0040] For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in the normalized reference data set, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows:

[0041]

[0042] Among them, Sst c Represents the event feature data of the reference dataset, Jcg c , Jcw c and Jcf c Respectively represent the extreme values ​​of the light intensity data, temperature data and wind speed data on the surface of the photovoltaic module in the reference data set;

[0043] The method for obtaining the characteristic data of the power generation sequence of the reference data set is:

[0044] The output power data of the photovoltaic modules in the reference data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated based on the following formula:

[0045]

[0046] Among them, Sgt. c Represents the characteristic data of the power generation sequence of the reference data set, Pjp c and σpc Represent the mean and standard deviation of the output power data in the reference data set, respectively.

[0047] Furthermore, the formula for generating the meteorological coincidence index and the power generation trend coincidence index between the reference data set and the entity data set is:

[0048]

[0049] Among them, Qxc i represents the meteorological coincidence index between the reference dataset and the i-th entity dataset, Fdc i Represents the power generation trend coincidence index between the reference dataset and the i-th entity dataset.

[0050] Furthermore, the first reference entity is a physical data set corresponding to the minimum meteorological coincidence index between the reference data set, and the second reference entity is a physical data set corresponding to the minimum power generation trend coincidence index between the reference data set. The method for obtaining the prediction interval of the power generation power of the photovoltaic assembly to be predicted is:

[0051] According to the parameters of the first reference entity and the second reference entity, the output power data and the parameters of the photovoltaic module to be predicted, four sets of predicted power data are generated respectively, and the formulas are as follows:

[0052]

[0053] Among them, Yc1, Yc2, Yc3 and Yc4 represent four sets of predicted power data generated, A0, Respectively represent the area and standard photoelectric conversion efficiency of the solar panel of the photovoltaic module to be predicted, A1, They represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the first reference entity, A2, represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the second reference entity, Respectively represent output power data of photovoltaic modules corresponding to the first reference entity and the second reference entity;

[0054] The generated prediction interval of the power generation power of the photovoltaic assembly to be predicted is [min(Yc1, Yc2), max(Yc3, Yc4)].

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention collects multiple groups of photovoltaic module's own parameters and historical environmental power generation parameters, and collects the photovoltaic module's own parameters and environmental power generation parameters to be predicted in real time. Through the mapping and analysis of these data, it is ensured that the prediction model can fully reflect the actual operation status of the photovoltaic power generation system, and the meteorological data and power generation data in the reference data set and each entity data set are analyzed for time series characteristics and extreme event characteristics, realizing the effective fusion of multivariate data, which can better reflect the change trend and abnormal situation of meteorological data;

[0057] The generated meteorological coincidence index and power generation trend coincidence index can quantify the similarity between the reference data set and the entity data set, screen out the reference entity most relevant to the photovoltaic module to be predicted, and generate a more accurate power generation prediction interval by combining the parameters of the photovoltaic module to be predicted and the power generation data of the reference entity. This screening and correction mechanism based on the coincidence index significantly improves the accuracy of the prediction and provides more reliable data support for the scheduling and energy management of photovoltaic modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0059] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0060] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0061] Example:

[0062] See also Figure 1 , the present invention provides a technical solution:

[0063] A photovoltaic power generation range prediction method considering photovoltaic output and meteorological factors, the specific steps include:

[0064] Step 1: Collect multiple sets of PV module parameters and historical environmental power generation parameters. Environmental power generation parameters include continuous power generation data and meteorological data with timestamps when the PV modules are generating electricity. The continuous historical power generation parameters and historical meteorological parameters are divided into time intervals of equal length. The data in each time interval is mapped with the parameters to form an entity data set.

[0065] In this embodiment, the timestamps of the power generation data and meteorological data are synchronized accordingly, and the length of each time interval is the same. If the power generation data and meteorological data are not enough to be divided into one time interval, they will be discarded. This can ensure that the data in each time interval is complete and consistent, avoiding prediction errors caused by missing or inconsistent data. Each time interval includes power generation data segments and meteorological data segments located in the time interval, and the timestamps of the power generation data segments and the meteorological data segments in each time interval are also synchronized accordingly, and are mapped with the static parameters of the corresponding photovoltaic components to form an entity data set.

[0066] By ensuring the synchronization of the data's timestamp correspondence and the uniformity of the time interval, the quality and consistency of the data can be effectively improved, and the prediction errors caused by inconsistent data can be reduced. This embodiment ensures the high quality and consistency of the data through strict data processing steps, thereby improving the accuracy of the prediction.

[0067] Furthermore, the photovoltaic module's own parameters include the area of ​​the photovoltaic module's solar panel and the standard photoelectric conversion efficiency, the power generation data is the output power of the photovoltaic module, the meteorological data includes the temperature, wind speed of the photovoltaic module's environment and the light intensity data on the photovoltaic module's surface, the photovoltaic module's output power data, the photovoltaic module's environment temperature, wind speed and the light intensity data on the photovoltaic module's surface are all continuous analog data, so the output power, temperature, wind speed and light intensity data are continuously changing rather than discrete values.

[0068] The area of ​​solar panels refers to the total area of ​​solar panels in the photovoltaic module, measured in square meters. The standard photoelectric conversion efficiency refers to the efficiency of photovoltaic modules in converting solar energy into electrical energy under standard test conditions, which is provided by the manufacturer of photovoltaic modules. The output power refers to the actual electrical power output of the photovoltaic module at a specific moment, measured in watts. The output power changes with time. The temperature refers to the air temperature of the environment in which the photovoltaic module is located, measured in degrees Celsius. The temperature will affect the performance of the photovoltaic module because the efficiency of photovoltaic cells will decrease with increasing temperature. The wind speed refers to the wind speed around the photovoltaic module, measured in meters per second. The wind speed may affect the cooling effect of the photovoltaic module, thereby indirectly affecting its power generation efficiency. The light intensity refers to the intensity of sunlight irradiating the surface of the photovoltaic module, measured in watts per square meter. The light intensity directly affects the power generation efficiency of the photovoltaic module, because the basic principle of photovoltaic power generation is to convert light energy into electrical energy.

[0069] Step 2: Collect the parameters of the photovoltaic module to be predicted and the environmental power generation parameters of the photovoltaic module in a period before the current moment in real time, and map the parameters of the photovoltaic module to be predicted and the environmental power generation parameters to form a reference data set, and number each entity data set.

[0070] In this embodiment, the reference data set is:

[0071]

[0072] Among them, Ck s represents the reference data set, Yc g (zs) represents the parameters of the photovoltaic module to be predicted, represents the power generation data of the photovoltaic module to be predicted from time j-1 to time j, represents the meteorological data of the photovoltaic module to be predicted from time j-1 to time j, where time j is the current time, and time j-1 is the time traced back to a period of time before the current time, and j-(j-1) = ω0, ω0 represents the length of the traced time period, and the output power, temperature, wind speed and light intensity data of the reference data set are continuous analog data with the current time as the reference and the forward ω0 time length, Yc g (zs), and They are all data sets.

[0073] The entity dataset is:

[0074]

[0075] Among them, Sts i Represents the i-th entity data set, i represents the number of the entity data set, Stg i(zs) and represents the parameters of the PV module in the i-th entity data set, and They represent the power generation data and meteorological data of the photovoltaic module from time k-1 to time k in the i-th entity data set, respectively. k-1 and k represent the starting time and ending time of the time interval corresponding to the entity data set, respectively. The time span of the time interval is ω=k-(k-1), and ω≥ω0. The length of the time interval corresponding to each entity data set is the same. Stg i (zs), and are all collections of data, Stg i (zs) represents the set of the photovoltaic module's own parameters in the i-th entity data set, represents the collection of power generation data of photovoltaic modules in the i-th entity data set, Represents the collection of meteorological data of photovoltaic components in the i-th entity dataset.

[0076] ω≥ω0 ensures that the time span of the entity dataset is not less than the time span of the reference dataset. This ensures that the time range of the entity dataset can completely cover the time range of the reference dataset during comparison and analysis. This helps ensure that there is sufficient data for comparison and analysis at all time points, avoiding inaccurate analysis caused by missing data.

[0077] The construction of the reference dataset focuses on the data of the specific time period of the photovoltaic components to be predicted, including their own parameters, power generation data and meteorological data. This targeted dataset construction helps to accurately capture the real-time performance and environmental impact of the components to be predicted. At the same time, the construction of the entity dataset covers the historical data of multiple physical photovoltaic components, ensuring the comprehensiveness of the dataset.

[0078] Step 3: Perform time series feature analysis and extreme event feature analysis on the meteorological data in the reference dataset and each entity dataset, obtain the meteorological sequence feature data and event feature data of the reference dataset and each entity dataset, perform time series feature analysis on the power generation data in the reference dataset and each entity dataset, and obtain the power generation sequence feature data of the reference dataset and each entity dataset.

[0079] In this embodiment, the method for obtaining the meteorological sequence characteristic data of the entity data set is as follows:

[0080] After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set, the average and standard deviation of the normalized temperature data, the average and standard deviation of the wind speed data, and the average and standard deviation of the light intensity data are calculated, and then the meteorological sequence characteristic data is generated. The formula is as follows:

[0081]

[0082] Among them, Sqt i represents the meteorological sequence feature data of the i-th entity data set, Pjg i and σg i Respectively represent the mean and standard deviation of the surface light intensity data of the photovoltaic module in the i-th entity data set, Pjw i and σw i Respectively represent the mean and standard deviation of the temperature data in the i-th entity data set, Pjf i and σf i They represent the mean and standard deviation of the wind speed data in the i-th entity data set respectively. α, β and γ represent the light intensity weight, temperature weight and wind speed weight respectively. α>β>γ>0, and α+β+γ=1.

[0083] In photovoltaic power generation systems, light intensity is the most critical factor affecting power generation efficiency. Higher light intensity directly leads to higher power output. Therefore, when synthesizing meteorological sequence feature data, the light intensity weight is set to the highest to ensure that the model can fully capture the impact of light intensity on power generation performance. Temperature also has a significant impact on the efficiency of photovoltaic modules, but its impact is smaller than that of light intensity. High temperature may cause the efficiency of photovoltaic modules to decrease. Therefore, the temperature weight is set to the second highest to reflect its importance in meteorological sequence characteristics. Wind speed indirectly affects power generation efficiency mainly by affecting the cooling effect of the components. Moderate wind speed helps dissipate heat and may slightly improve efficiency, but its direct impact is not as significant as light intensity and temperature. Therefore, the wind speed weight is set to the lowest.

[0084] The meteorological sequence characteristic data reflects the expected power generation performance of photovoltaic modules under specific meteorological conditions, that is, the degree of impact on power generation. The larger the meteorological sequence characteristic data, the more suitable the current meteorological conditions are for photovoltaic modules to generate electricity, and the greater the power generation of photovoltaic modules should be. These characteristic data provide a quantitative indicator for evaluating and predicting the power generation performance of photovoltaic modules under different meteorological conditions by comprehensively considering the statistical characteristics of light intensity, temperature and wind speed.

[0085] A high average of light intensity means that there is sufficient light most of the time, and a small standard deviation means that the light intensity does not change much, that is, the light conditions are stable. Such conditions are conducive to the continuous and efficient power generation of photovoltaic modules. The efficiency of photovoltaic cells is usually higher at lower temperatures. A low average of temperature means that the ambient temperature is suitable and will not cause overheating of photovoltaic cells and reduce efficiency. A small standard deviation means that the temperature does not change much, that is, the temperature conditions are stable, which helps to keep the efficiency of photovoltaic modules stable. Moderate wind speed helps to dissipate heat and can reduce the temperature of photovoltaic modules, thereby improving efficiency. A large average of wind speed means that there is enough wind to help dissipate heat, and a large standard deviation means that the wind speed varies greatly, which may mean that there will be strong wind speeds when heat dissipation is needed, which helps to maintain a suitable operating temperature for photovoltaic modules.

[0086] The method for obtaining event feature data of entity dataset is as follows:

[0087] For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set after normalization, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows:

[0088]

[0089] Among them, Sst i Represents the event feature data of the i-th entity dataset, Jcg i , Jcw i and Jcf i They respectively represent the extreme values ​​of the light intensity data, temperature data and wind speed data on the surface of the photovoltaic module in the i-th entity data set.

[0090] Light intensity is the most direct factor affecting photovoltaic power generation. The range reflects the fluctuation range of light intensity, while the average represents the overall light level. By calculating the ratio of the range to the average, the stability of light intensity can be measured. A higher ratio indicates that the light intensity fluctuates greatly. Excessive fluctuations will introduce instability. A larger range and a lower average are not conducive to stable and efficient power generation by photovoltaic modules.

[0091] Temperature affects the efficiency of photovoltaic modules. The range reflects the fluctuation range of temperature, and the average indicates the overall temperature level. By calculating the ratio of the range to the average, the temperature change can be measured. A higher ratio indicates that the temperature fluctuation is large and the overall average is high, which may cause the risk of efficiency reduction due to overheating of the module. Wind speed mainly affects the heat dissipation performance of photovoltaic modules. The range reflects the fluctuation range of wind speed, and the average indicates the overall wind speed level. By calculating the ratio of the range to the average, the fluctuation of wind speed can be measured.

[0092] The event characteristic data reflects the fluctuation of temperature, wind speed and light intensity data on the surface of photovoltaic modules, that is, the situation of abnormal events. This characteristic data is calculated by comprehensively considering the ratio of the extreme value to the average of light intensity, temperature and wind speed, providing a quantitative indicator to evaluate the fluctuation of meteorological conditions. The greater the fluctuation of temperature, wind speed and light intensity data on the surface of photovoltaic modules, the higher the possibility of abnormal situation in event characteristic data.

[0093] Furthermore, the method for obtaining the power generation sequence characteristic data of the entity data set is:

[0094] The output power data of the photovoltaic modules in the entity data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated. The formula is as follows:

[0095]

[0096] Among them, Sgt. i represents the power generation sequence characteristic data of the i-th entity data set, Pjp i and σp i Respectively represent the mean and standard deviation of the output power data in the i-th entity data set.

[0097] When the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set are normalized respectively, the temperature, wind speed and light intensity data on the surface of the photovoltaic module are normalized respectively, and then the mean and standard deviation of the normalized temperature data, the mean and standard deviation of the wind speed data, and the mean and standard deviation of the light intensity data are calculated. In this embodiment, the temperature in the entity data set is used as an example for explanation. Other methods are similar to this method and will not be repeated here.

[0098] The logic for normalizing the temperature in the entity dataset is:

[0099]

[0100] Among them, Sw i (t) represents the continuous function representation of the temperature data in the i-th entity data set, t represents the time variable, and t∈[k-1,k], Sw i (t)' represents the normalized temperature data of the ith entity data set;

[0101] The logic for calculating the mean and standard deviation of normalized temperature data is:

[0102]

[0103] The specific operations of the mean and standard deviation of wind speed data, and the mean and standard deviation of light intensity data are the same as the above method.

[0104] The average output power reflects the average power generation capacity of the PV module within a certain period of time. A higher average indicates that the PV module can stably output a higher power most of the time. The standard deviation of the output power reflects the volatility of the PV module output power. The larger the standard deviation, the greater the fluctuation of the output power and the worse the stability. A lower standard deviation indicates that the output power of the PV module fluctuates less and the power generation process is more stable.

[0105] The power generation sequence characteristic data reflects the stability of the output power and the average power generation capacity of the photovoltaic module. The larger the power generation sequence characteristic data, the higher the power generation capacity of the photovoltaic module can output most of the time, and the smaller the output power fluctuation is, the more stable the power generation process is. The power generation sequence characteristic data is used to identify the performance of the photovoltaic module in a certain period of time.

[0106] The method for obtaining the meteorological sequence characteristic data of the reference data set is based on:

[0107] After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic modules in each reference data set, the mean and standard deviation of the normalized temperature data, the mean and standard deviation of the wind speed data, and the mean and standard deviation of the light intensity data are calculated, and the meteorological sequence characteristic data is generated based on the following formula:

[0108]

[0109] Among them, Sqt c Represents the meteorological sequence characteristic data of the reference data set, Pjg c and σg c Respectively represent the mean and standard deviation of the surface light intensity data of photovoltaic modules in the reference data set, Pjw c and σw c Respectively represent the mean and standard deviation of the temperature data in the reference data set, Pjf c and σf c Respectively represent the mean and standard deviation of wind speed data in the reference data set;

[0110] The method for obtaining event feature data of the reference data set is as follows:

[0111] For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in the normalized reference data set, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows:

[0112]

[0113] Among them, Sst c Represents the event feature data of the reference dataset, Jcg c , Jcw c and Jcf c Respectively represent the extreme values ​​of the light intensity data, temperature data and wind speed data on the surface of the photovoltaic modules in the reference data set.

[0114] The method for obtaining the characteristic data of the power generation sequence of the reference data set is:

[0115] The output power data of the photovoltaic modules in the reference data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated based on the following formula:

[0116]

[0117] Among them, Sgt. c Represents the characteristic data of the power generation sequence of the reference data set, Pjp c and σp c They respectively represent the mean and standard deviation of the output power data in the reference data set. In this embodiment, the construction logic of the meteorological sequence characteristic data, event characteristic data and power generation sequence characteristic data of the reference data set is the same as that of the reference data set, which will not be repeated here.

[0118] This embodiment collects multiple groups of photovoltaic components' own parameters and historical environmental power generation parameters, and collects the photovoltaic components' own parameters and environmental power generation parameters to be predicted in real time. Through the mapping and analysis of these data, it is ensured that the prediction model can fully reflect the actual operating status of the photovoltaic power generation system, and performs time series feature analysis and extreme event feature analysis on the meteorological data and power generation data within the reference data set and each entity data set, thereby realizing the effective fusion of multivariate data and better reflecting the changing trends and abnormal conditions of meteorological data.

[0119] Step 4: Based on the meteorological sequence characteristic data, event characteristic data and power generation sequence characteristic data of the reference dataset and the entity dataset, traverse all entity datasets to generate the meteorological coincidence index and power generation trend coincidence index between the reference dataset and each entity dataset.

[0120] In this embodiment, the formula for generating the meteorological coincidence index and the power generation trend coincidence index between the reference data set and the entity data set is:

[0121]

[0122] Among them, Qxc irepresents the meteorological coincidence index between the reference dataset and the i-th entity dataset, Fdc i Represents the power generation trend coincidence index between the reference dataset and the i-th entity dataset.

[0123] The meteorological coincidence index evaluates the similarity of meteorological conditions between the reference dataset and the entity dataset by calculating the differences between the meteorological sequence characteristic data and the event characteristic data. Meteorological conditions have a significant impact on photovoltaic power generation. By comparing the differences between the meteorological sequence characteristic data and the event characteristic data, the similarity of meteorological conditions between different datasets can be quantified, thereby evaluating their potential similarities in power generation performance.

[0124] The power generation trend coincidence index evaluates the similarity of the power generation trends of the reference dataset and the entity dataset by calculating the difference in the power generation sequence characteristic data between the two datasets. The power generation trend directly reflects the output power performance of the photovoltaic module. By comparing the difference in the power generation sequence characteristic data, the similarity of the power generation trends of different datasets can be quantified, thereby evaluating their potential consistency in power generation performance.

[0125] The meteorological overlap index reflects the similarity between the reference dataset and the entity dataset in meteorological conditions. The smaller the index, the smaller the difference between the two in meteorological conditions, and the higher the similarity between the two in meteorological conditions, which helps to predict their potential consistency in power generation performance. The power generation trend overlap index reflects the similarity between the reference dataset and the entity dataset in power generation trends. The smaller the index, the smaller the difference between the two in power generation trends, and the higher the similarity between the two in power generation trends, which helps to evaluate their potential consistency in power generation performance.

[0126] Step 5: Filter out the entity data set with the smallest meteorological coincidence index error and the entity data set with the smallest power generation trend coincidence index with the reference data set respectively, and calibrate them as the first reference entity and the second reference entity. By collecting the own parameters of the photovoltaic component to be predicted and combining the own parameters of the first reference entity and the second reference entity, correct the power generation data of the first reference entity and the second reference entity to obtain the prediction interval of the power generation power of the photovoltaic component to be predicted.

[0127] In this embodiment, the first reference entity is an entity data set corresponding to the minimum meteorological coincidence index between the reference data set, and the second reference entity is an entity data set corresponding to the minimum power generation trend coincidence index between the reference data set. The method for obtaining the prediction interval of the power generation power of the photovoltaic assembly to be predicted is:

[0128] According to the parameters of the first reference entity and the second reference entity, the output power data and the parameters of the photovoltaic module to be predicted, four sets of predicted power data are generated respectively, and the formulas are as follows:

[0129]

[0130] Among them, Yc1, Yc2, Yc3 and Yc4 represent four sets of predicted power data generated, A0, Respectively represent the area and standard photoelectric conversion efficiency of the solar panel of the photovoltaic module to be predicted, A1, They represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the first reference entity, A2, represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the second reference entity, Respectively represent the output power data of the photovoltaic components corresponding to the first reference entity and the second reference entity.

[0131] The area of ​​a solar panel directly affects the amount of solar radiation it receives. The larger the area, the more solar radiation it receives in theory, which may generate higher power. Photoelectric conversion efficiency refers to the efficiency of solar panels in converting received solar radiation into electrical energy. The higher the efficiency, the more electrical energy is generated under the same conditions.

[0132] By comparing the area and photoelectric conversion efficiency of the photovoltaic module to be predicted and the reference entity, the difference in their power generation capacity can be quantified. The difference in area and photoelectric conversion efficiency between the photovoltaic module to be predicted and the reference entity directly affects their power generation performance under the same conditions. By calculating the ratio of the two, a correction factor can be obtained to adjust the power generation data of the reference entity to make it closer to the actual power generation capacity of the photovoltaic module to be predicted. The correction factor is multiplied by the output power data of the reference entity to generate the predicted power data of the photovoltaic module to be predicted. By multiplying the output power data of the reference entity by the correction factor, a predicted value closer to the actual power generation capacity of the photovoltaic module to be predicted can be obtained. It helps to improve the accuracy of the prediction and provide a basis for determining the prediction interval.

[0133] The generated prediction interval of the power generation power of the photovoltaic module to be predicted is [min(Yc1, Yc2), max(Yc3, Yc4)]. By selecting the entity data set with the smallest meteorological coincidence index with the reference data set as the first reference entity, and the entity data set with the smallest power generation trend coincidence index as the second reference entity, it can be ensured that the photovoltaic module to be predicted has a high similarity with the reference entity in meteorological conditions and power generation trends. Meteorological conditions and power generation trends are key factors affecting the power generation performance of photovoltaic modules. The entity data set that is most similar to the reference data set in meteorological conditions and power generation trends is selected as the reference entity. According to the parameters of the photovoltaic module to be predicted and the reference entity, as well as the output power data of the reference entity, four sets of predicted power data are generated, which can quantify their differences in power generation capacity. Using the output power data of the reference entity, combined with the parameter differences, the predicted power data of the photovoltaic module to be predicted can be generated, thereby providing a basis for the prediction interval.

[0134] This embodiment determines the prediction interval of the power generation power of the photovoltaic module to be predicted by calculating the minimum and maximum values ​​of the four sets of predicted power data, and the prediction interval reflects the power generation capacity range of the photovoltaic module to be predicted under different conditions. By comprehensively considering the minimum and maximum values ​​of the four sets of predicted power data, a prediction interval containing the potential power generation capacity range can be provided, thereby providing a reliable reference for the power generation prediction of the photovoltaic module.

[0135] In summary, this embodiment selects the entity data set that is most similar to the reference data set in meteorological conditions and power generation trends as the reference entity, combines the parameters of the photovoltaic components to be predicted and the reference entity itself, generates predicted power data, and finally determines the prediction interval to improve the accuracy and reliability of photovoltaic component power generation prediction.

[0136] The generated meteorological coincidence index and power generation trend coincidence index can quantify the similarity between the reference data set and the entity data set, screen out the reference entity most relevant to the photovoltaic component to be predicted, and generate a more accurate power generation power prediction interval by combining the parameters of the photovoltaic component to be predicted and the power generation data of the reference entity. This screening and correction mechanism based on the coincidence index significantly improves the accuracy of the prediction and provides more reliable data support for the scheduling and energy management of photovoltaic components.

[0137] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0138] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0139] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0140] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A photovoltaic power generation interval prediction method considering photovoltaic output and meteorological factors, characterized in that: The specific steps include: Step 1: Collect multiple sets of PV module parameters and historical environmental power generation parameters. The environmental power generation parameters include continuous power generation data and meteorological data with timestamps when the PV modules are generating electricity. The continuous historical power generation parameters and historical meteorological parameters are divided into equal time intervals. The data in each time interval is mapped with the parameters to form an entity data set. Step 2: Collect the parameters of the photovoltaic module to be predicted and the environmental power generation parameters of the photovoltaic module in a period before the current moment in real time, and map the parameters of the photovoltaic module to be predicted and the environmental power generation parameters to form a reference data set, and number each entity data set; Step 3: Perform time series feature analysis and extreme event feature analysis on the meteorological data in the reference dataset and each entity dataset, obtain meteorological sequence feature data and event feature data of the reference dataset and each entity dataset, perform time series feature analysis on the power generation data in the reference dataset and each entity dataset, and obtain power generation sequence feature data of the reference dataset and each entity dataset; Step 4: Based on the meteorological sequence feature data, event feature data and power generation sequence feature data of the reference dataset and the entity dataset, all entity datasets are traversed to generate the meteorological coincidence index and power generation trend coincidence index between the reference dataset and each entity dataset; Step 5: Filter out the entity data set with the smallest meteorological coincidence index error and the entity data set with the smallest power generation trend coincidence index between the reference data set and the reference data set, and mark them as the first reference entity and the second reference entity. By collecting the parameters of the photovoltaic module to be predicted and combining the parameters of the first reference entity and the second reference entity, correct the power generation data of the first reference entity and the second reference entity to obtain the prediction interval of the power generation power of the photovoltaic module to be predicted. The photovoltaic module's own parameters include the area of ​​the photovoltaic module's solar panel and the standard photoelectric conversion efficiency, the power generation data is the photovoltaic module's output power, and the meteorological data includes the temperature and wind speed of the photovoltaic module's environment and the light intensity data on the photovoltaic module's surface; The reference dataset is: Among them, Ck s represents the reference data set, Yc g (zs) represents the parameters of the photovoltaic module to be predicted, represents the power generation data of the photovoltaic module to be predicted from time j-1 to time j, represents the meteorological data of the photovoltaic module to be predicted from time j-1 to time j, where time j is the current time, time j-1 is the time traced back to a period of time before the current time, and j-(j-1) = ω0, ω0 represents the length of the traced time period; The entity dataset is: Among them, Sts i Represents the i-th entity data set, i represents the number of the entity data set, Stg i (zs) and represents the parameters of the PV module in the i-th entity data set, and They represent the power generation data and meteorological data of the photovoltaic module from time k-1 to time k in the i-th entity data set, respectively. k-1 and k represent the starting time value and the ending time value of the time interval corresponding to the entity data set, respectively. The time span of the time interval is ω=k-(k-1), and ω≥ω0; The method for obtaining the meteorological sequence characteristic data of the entity data set is based on: After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set, the average and standard deviation of the normalized temperature data, the average and standard deviation of the wind speed data, and the average and standard deviation of the light intensity data are calculated, and then the meteorological sequence characteristic data is generated. The formula is as follows: Among them, Sqt i represents the meteorological sequence feature data of the i-th entity data set, Pjg i and σg i Respectively represent the mean and standard deviation of the surface light intensity data of the photovoltaic module in the i-th entity data set, Pjw i and σw i Respectively represent the mean and standard deviation of the temperature data in the i-th entity data set, Pjf i and σf i They represent the mean and standard deviation of the wind speed data in the i-th entity data set, respectively. α, β, and γ represent the light intensity weight, temperature weight, and wind speed weight, respectively. α>β>γ>0, and α+β+γ=1; The method for obtaining event feature data of entity dataset is as follows: For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in each entity data set after normalization, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows: Among them, Sst i Represents the event feature data of the i-th entity dataset, Jcg i , Jcw i and Jcf i Respectively represent the extreme values ​​of the surface light intensity data, temperature data and wind speed data of the photovoltaic module in the i-th entity data set; The method for obtaining the power generation sequence characteristic data of the entity data set is: The output power data of the photovoltaic modules in the entity data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated. The formula is as follows: Among them, Sgt. i represents the power generation sequence characteristic data of the i-th entity data set, Pjp i and σp i Respectively represent the mean and standard deviation of the output power data in the i-th entity data set; The method for obtaining the meteorological sequence characteristic data of the reference data set is based on: After normalizing the temperature, wind speed and light intensity data on the surface of the photovoltaic modules in each reference data set, the mean and standard deviation of the normalized temperature data, the mean and standard deviation of the wind speed data, and the mean and standard deviation of the light intensity data are calculated, and the meteorological sequence characteristic data is generated based on the following formula: Among them, Sqt c Represents the meteorological sequence characteristic data of the reference data set, Pjg c and σg c Respectively represent the mean and standard deviation of the surface light intensity data of photovoltaic modules in the reference data set, Pjw c and σw c Respectively represent the mean and standard deviation of the temperature data in the reference data set, Pjf c and σf c Respectively represent the mean and standard deviation of wind speed data in the reference data set; The method for obtaining event feature data of the reference data set is as follows: For the temperature, wind speed and light intensity data on the surface of the photovoltaic module in the normalized reference data set, the range values ​​of the normalized temperature data, wind speed data and light intensity data are calculated respectively, and then the event feature data is generated. The formula is as follows: Among them, Sst c Represents the event feature data of the reference dataset, Jcg c , Jcw c and Jcf c Respectively represent the extreme values ​​of the light intensity data, temperature data and wind speed data on the surface of the photovoltaic module in the reference data set; The method for obtaining the characteristic data of the power generation sequence of the reference data set is: The output power data of the photovoltaic modules in the reference data set are normalized, the mean and variance of the normalized output power data are calculated, and the power generation sequence characteristic data of the output power data are calculated based on the following formula: Among them, Sgt. c Represents the characteristic data of the power generation sequence of the reference data set, Pjp c and σp c Respectively represent the mean and standard deviation of the output power data in the reference data set; The formula for generating the meteorological coincidence index and power generation trend coincidence index between the reference dataset and the entity dataset is as follows: Among them, Qxc i represents the meteorological coincidence index between the reference dataset and the i-th entity dataset, Fdc i Represents the power generation trend coincidence index between the reference dataset and the i-th entity dataset.

2. The photovoltaic power generation interval prediction method considering photovoltaic output and meteorological factors according to claim 1 is characterized by: The timestamps of the power generation data and meteorological data are synchronized accordingly, and the length of each time interval is the same. If the power generation data and meteorological data are not enough to be divided into one time interval, they will be discarded. Each time interval includes the power generation data fragment and the meteorological data fragment located in the time interval, and the timestamps of the power generation data fragment and the meteorological data fragment in each time interval are also synchronized accordingly, and are mapped with the static parameters of the corresponding photovoltaic components to form an entity data set.

3. The photovoltaic power generation interval prediction method considering photovoltaic output and meteorological factors according to claim 1 is characterized by: The first reference entity is a physical data set corresponding to the minimum meteorological coincidence index between the reference data set, and the second reference entity is a physical data set corresponding to the minimum power generation trend coincidence index between the reference data set. The method for obtaining the prediction interval of the power generation power of the photovoltaic assembly to be predicted is: According to the parameters of the first reference entity and the second reference entity, the output power data and the parameters of the photovoltaic module to be predicted, four sets of predicted power data are generated respectively, and the formulas are as follows: Among them, Yc1, Yc2, Yc3 and Yc4 represent four sets of predicted power data generated, A0, Respectively represent the area and standard photoelectric conversion efficiency of the solar panel of the photovoltaic module to be predicted, A1, They represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the first reference entity, A2, represent the area and standard photoelectric conversion efficiency of the solar panel corresponding to the second reference entity, Respectively represent output power data of photovoltaic modules corresponding to the first reference entity and the second reference entity; The generated prediction interval of the power generation power of the photovoltaic assembly to be predicted is [min(Yc1, Yc2), max(Yc3, Yc4)].

Images