Online processing method of photovoltaic module energy efficiency data based on cloud computing

Abstract

The present invention relates to the field of data processing technology, and in particular to a cloud computing-based online processing method for photovoltaic component energy efficiency data. The method obtains energy efficiency data of any photovoltaic component and each reference photovoltaic component at each moment, and obtains an energy efficiency data sequence of any photovoltaic component on that day and a reference energy efficiency data sequence of each reference photovoltaic component on that day; if there is missing data in the energy efficiency data sequence, reference data and historical reference data of any missing data are obtained, and an initial reference value of any missing data is obtained based on data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data; a theoretical value and a fitted value of any missing data are obtained, and the initial reference value is corrected based on the difference between the theoretical value and the fitted value to obtain an optimized filling value, and the optimized filling value of each missing data in the energy efficiency data sequence is obtained, so as to accurately and completely obtain the energy efficiency data of any photovoltaic component.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an online processing method for photovoltaic module energy efficiency data based on cloud computing. Background Art

[0002] With the continuous growth of global energy demand and increasing environmental protection requirements, photovoltaic power generation, as a clean, renewable energy source, has gradually become an important energy source. As the core component of photovoltaic power generation systems, the energy efficiency of photovoltaic modules (i.e., solar panels) directly affects the overall efficiency of photovoltaic power generation systems. High-precision data from photovoltaic modules not only supports power generation forecasting, fault diagnosis, and intelligent operation and maintenance, reducing energy losses, but also assists in grid scheduling and energy storage configuration, improving renewable energy absorption capacity. Therefore, accurately and completely obtaining photovoltaic module energy efficiency data has become a key research direction for improving the performance of photovoltaic power generation systems and ensuring energy utilization efficiency.

[0003] When collecting the energy efficiency data of photovoltaic modules, the collected data may be missing due to failures in sensors, inverters or communication modules. Therefore, after collecting the energy efficiency data of photovoltaic modules, the missing data needs to be filled in to ensure the integrity of the data.

[0004] Traditionally, missing value filling typically uses mean filling, which involves taking the mean of the energy efficiency data from multiple PV modules of the same model and time to fill in the missing data. While mean filling can quickly and easily integrate similar data to fill in missing data, it gives equal weight to the energy efficiency data of each PV module. This can affect the accuracy of the filled data if some data is inherently unreliable. Furthermore, mean filling only fills in missing data based on other data from PV modules of the same model and time, without considering the inherent characteristics of the missing data. This results in poor consistency between the filled data and the original data, further reducing the accuracy of the filled data.

[0005] Therefore, how to improve the accuracy of filling data and accurately and completely obtain the energy efficiency data of photovoltaic modules has become an urgent problem that needs to be solved. Summary of the Invention

[0006] In view of this, an embodiment of the present invention provides a method for online processing of photovoltaic module energy efficiency data based on cloud computing to solve the problem of how to improve the accuracy of filling data and accurately and completely obtain the energy efficiency data of photovoltaic modules.

[0007] An embodiment of the present invention provides a method for online processing of photovoltaic module energy efficiency data based on cloud computing, the method comprising the following steps:

[0008] For any photovoltaic module, obtaining a preset number of reference photovoltaic modules of the any photovoltaic module, obtaining energy efficiency data of the any photovoltaic module and each reference photovoltaic module at each moment, and obtaining an energy efficiency data sequence of the any photovoltaic module on that day and a reference energy efficiency data sequence of each reference photovoltaic module on that day;

[0009] If missing data is detected in the energy efficiency data sequence, for any missing data, reference data of the missing data is obtained in each reference energy efficiency data sequence, historical reference data of each reference photovoltaic module is obtained within a preset number of historical days, and an initial reference value of the missing data is obtained based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs;

[0010] Obtaining a theoretical value of any missing data based on the environmental data and performance parameter data of any photovoltaic module, fitting the energy efficiency data sequence to obtain a fitted value of any missing data, and correcting the initial reference value based on a difference between the theoretical value and the fitted value of any missing data to obtain an optimized filling value of any missing data;

[0011] Obtain an optimized filling value for each missing data in the energy efficiency data sequence to obtain a complete energy efficiency data sequence of any photovoltaic component for transmission to a cloud computing platform for online processing.

[0012] Preferably, obtaining the initial reference value of any missing data according to the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the corresponding reference photovoltaic module includes:

[0013] For any reference data, obtaining the overall reliability of the reference data according to the data fluctuation characteristics of the reference energy efficiency data series to which the reference data belongs;

[0014] Taking any reference data as the center, construct a reference window of a preset length in the reference energy efficiency data sequence to which the any reference data belongs, and obtain the local stability of the any reference data based on the data fluctuation characteristics of the reference energy efficiency data within the reference window and the difference between the any reference data and the historical reference data of the reference photovoltaic module to which it belongs;

[0015] The overall reliability and local stability of each reference data are obtained, and the initial reference value of any missing data is obtained according to the overall reliability and local stability of each reference data.

[0016] Preferably, obtaining the overall reliability of any reference data according to the data fluctuation characteristics of the reference energy efficiency data sequence to which the any reference data belongs includes:

[0017] Recording a reference energy efficiency data sequence to which any reference data belongs as a reference sequence, obtaining the number of missing data in the reference sequence as the amount of missing data, obtaining the length of the reference sequence, and obtaining a ratio of the amount of missing data to the length of the reference sequence to obtain a degree of incompleteness of the reference sequence;

[0018] Obtaining rated energy efficiency data of a reference photovoltaic module to which any reference data belongs, obtaining the number of reference energy efficiency data in the reference sequence that is greater than the rated energy efficiency data, recording it as the amount of excess data, obtaining a ratio of the amount of excess data to the length of the reference sequence, and obtaining a degree of irrationality of the reference sequence;

[0019] The overall reliability of any reference data is obtained by substituting the inverse of the product of the incompleteness and irrationality of the reference sequence into an exponential function with a natural constant as the base.

[0020] Preferably, obtaining the local stability of any reference data according to the data fluctuation characteristics of the reference energy efficiency data within the reference window and the difference between any reference data and the historical reference data of the corresponding reference photovoltaic module includes:

[0021] Using an isolation forest algorithm, obtaining the number of abnormal data in the reference window, obtaining the ratio of the number of abnormal data to the preset length, and obtaining the abnormality degree of the reference window;

[0022] Obtaining an average of historical reference data of a reference photovoltaic module to which any reference data belongs, recording the average as a historical average, obtaining an absolute value of a difference between the historical average and any reference data, and obtaining a degree of deviation of the any reference data;

[0023] Obtain the coefficient of variation of the reference energy efficiency data in the reference window, obtain the product of the degree of abnormality of the reference window, the coefficient of variation and the degree of deviation of any reference data, substitute the negative of the product into an exponential function with a natural constant as the base, and obtain the local stability of any reference data.

[0024] Preferably, obtaining the initial reference value of any missing data according to the overall reliability and local stability of each reference data includes:

[0025] For any reference data, obtain the product of the overall reliability and the local stability of the any reference data to obtain the reference level of the any reference data, obtain the reference level of each reference data, obtain a corresponding reference level cumulative value, obtain the ratio of the reference level of the any reference data to the reference level cumulative value, and obtain a weight coefficient for the any reference data;

[0026] A weight coefficient for each reference data is obtained, and a weighted sum is performed on all reference data to obtain an initial reference value for any missing data.

[0027] Preferably, obtaining the theoretical value of any missing data based on the environmental data and performance parameter data of any photovoltaic module includes:

[0028] The environmental data of any photovoltaic module include irradiance and battery temperature, and the performance parameter data of any photovoltaic module include effective area of ​​photovoltaic module, efficiency under standard test conditions, power temperature coefficient, optimal operating temperature, and annual attenuation rate;

[0029] Recording the time at which any missing data occurs as a target time, obtaining the product of the irradiance of any photovoltaic module at the target time, the effective area of ​​the photovoltaic module, and the efficiency under the standard test condition, to obtain the theoretical output power of any photovoltaic module at the target time;

[0030] Obtaining an absolute value of a difference between a battery temperature of any photovoltaic module at the target time and the optimal operating temperature to obtain a temperature difference, obtaining a product of the power temperature coefficient and the temperature difference and adding a constant 1 to obtain a temperature correction factor of any photovoltaic module at the target time;

[0031] Obtaining the service life of any photovoltaic module, obtaining a difference between a constant 1 and the annual attenuation rate, substituting the service life into an exponential function with the difference as the base, and obtaining the operating efficiency of any photovoltaic module at the target time;

[0032] The product of the theoretical output power, the temperature correction factor and the working efficiency is obtained to obtain a theoretical value of any missing data.

[0033] Preferably, the step of correcting the initial reference value according to the difference between the theoretical value and the fitted value of any missing data to obtain the optimized filling value of any missing data includes:

[0034] Obtain the difference between the initial reference value and the fitting value, record it as the fitting difference, obtain the difference between the initial reference value and the theoretical value, record it as the theoretical difference, obtain the average between the fitting difference and the theoretical difference, record it as the correction value, obtain the addition result between the initial reference value and the correction value, and obtain the optimized filling value of any missing data.

[0035] Preferably, for any missing data, obtaining reference data of the missing data in each reference energy efficiency data sequence includes:

[0036] Recording the time at which any missing data occurs as a target time, and for any reference energy efficiency data sequence, obtaining reference energy efficiency data at the target time in the any reference energy efficiency data sequence as reference data for any missing data;

[0037] If the reference energy efficiency data at the target moment in any of the reference energy efficiency data sequences is missing data, obtaining the average of the reference energy efficiency data at the previous moment and the next moment before the target moment in any of the reference energy efficiency data sequences as the reference data for any missing data;

[0038] If at least one of the reference energy efficiency data at the previous moment and the next moment of the target moment in any reference energy efficiency data sequence is missing data, a constant 0 is used as the reference data of any missing data.

[0039] Preferably, the obtaining of a preset number of reference photovoltaic assemblies of any photovoltaic assembly includes:

[0040] A photovoltaic assembly of the same model and located in the same region as any one of the photovoltaic assemblies is obtained as a reference photovoltaic assembly for the any one of the photovoltaic assemblies.

[0041] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0042] The present invention obtains a preset number of reference photovoltaic modules for any photovoltaic module, obtains energy efficiency data of any photovoltaic module and each reference photovoltaic module at each moment, and obtains an energy efficiency data sequence of any photovoltaic module on that day and a reference energy efficiency data sequence of each reference photovoltaic module on that day; if missing data is detected in the energy efficiency data sequence, then, for any missing data, reference data of the missing data is obtained in each reference energy efficiency data sequence, historical reference data of each reference photovoltaic module is obtained within a preset number of historical days, and an initial reference value of the missing data is obtained based on data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the corresponding reference photovoltaic module; theoretical values ​​of the missing data are obtained based on environmental data and performance parameter data of the missing data, the energy efficiency data sequence is fitted to obtain a fitted value of the missing data, and the initial reference value is corrected based on the difference between the theoretical value and the fitted value of the missing data to obtain an optimized filled value of the missing data; the optimized filled value of each missing data in the energy efficiency data sequence is obtained, and a complete energy efficiency data sequence of the photovoltaic module is obtained for transmission to a cloud computing platform for online processing. Among them, first, according to the data fluctuation characteristics of each reference energy efficiency data series and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs, the initial reference value of any missing data is obtained, and then according to the difference between the theoretical value and the fitted value of any missing data, the initial reference value is corrected to obtain the optimized filling value of any missing data. It not only integrates the reference data of each reference photovoltaic module of the same type, but also combines the local characteristics of the energy efficiency data series of any photovoltaic module, ensuring data integrity while improving the accuracy of the filled data, and accurately and completely obtaining the energy efficiency data of the photovoltaic module. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 This is a method flow chart of a method for online processing of photovoltaic module energy efficiency data based on cloud computing provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0045] The embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.

[0046] It should be noted that the terms "first," "second," and the like in the specification of the present disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0047] In order to illustrate the technical solution of the present invention, specific embodiments are provided below.

[0048] See also Figure 1 , is a method flow chart of a method for online processing of photovoltaic module energy efficiency data based on cloud computing provided by the first embodiment of the present invention, such as Figure 1 As shown, the method may include:

[0049] Step S101: For any photovoltaic component, a preset number of reference photovoltaic components of the any photovoltaic component are obtained, and the energy efficiency data of the any photovoltaic component and each reference photovoltaic component at each moment are obtained to obtain the energy efficiency data sequence of the any photovoltaic component on that day and the reference energy efficiency data sequence of each reference photovoltaic component on that day.

[0050] Photovoltaic modules (i.e., solar panels) are core components of photovoltaic power generation systems. Their energy efficiency directly impacts the overall efficiency of these systems. Therefore, accurately and completely acquiring energy efficiency data from these modules has become a crucial research direction for improving system performance and ensuring energy efficiency. Failures in sensors, inverters, or communication modules can result in missing data. Therefore, after collecting energy efficiency data from these modules, it is necessary to fill in the missing data to ensure data integrity.

[0051] First, energy efficiency data for each PV module at each sampling moment is obtained. Output power, typically measured in watts (W), is the actual output of a PV module after converting solar energy into electrical energy. It is the most intuitive indicator of PV module performance. Therefore, in this embodiment, the output power of each PV module at each moment is obtained as energy efficiency data based on its output voltage and output current. In this embodiment, the output power collection frequency is set to once per second, but this is not a limitation and can be set based on specific implementation scenarios. The acquisition of output power is conventional technology and is not detailed here.

[0052] The present invention obtains, for any photovoltaic module, a preset number of photovoltaic modules of the same model and located in the same region as any photovoltaic module as reference photovoltaic modules for any photovoltaic module. In this embodiment, the preset number of reference photovoltaic modules is set to 5, which is not limited here. It can be set according to the specific implementation scenario to obtain the energy efficiency data sequence of any photovoltaic module on that day and the reference energy efficiency data sequence of each reference photovoltaic module on that day.

[0053] Traditionally, missing value filling typically uses mean filling, which involves taking the mean of the energy efficiency data from multiple PV modules of the same model and time to fill in the missing data. While mean filling can quickly and easily integrate similar data to fill in missing data, it gives equal weight to the energy efficiency data of each PV module. This can affect the accuracy of the filled data if some data is inherently unreliable. Furthermore, mean filling only fills in missing data based on other data from PV modules of the same model and time, without considering the inherent characteristics of the missing data. This results in poor consistency between the filled data and the original data, further reducing the accuracy of the filled data.

[0054] Therefore, after detecting the existence of missing data, the embodiment of the present invention first obtains the initial reference value of the missing data based on the data fluctuation characteristics of the reference energy efficiency data sequence of each reference photovoltaic module, and then corrects the initial reference value in combination with the fluctuation characteristics of the energy efficiency data sequence of any photovoltaic module to obtain the optimized filling value of the missing data, thereby improving the accuracy of the filling data and accurately and completely obtaining the energy efficiency data of the photovoltaic module.

[0055] Step S102: If missing data is detected in the energy efficiency data sequence, for any missing data, reference data of the missing data is obtained in each reference energy efficiency data sequence, and historical reference data of each reference photovoltaic module is obtained within a preset number of historical days. Based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs, an initial reference value of the missing data is obtained.

[0056] After obtaining the energy efficiency data sequence for any photovoltaic module on that day, the energy efficiency data sequence is first checked for missing data. In this embodiment, a direct judgment method is used to perform missing data detection on the energy efficiency data sequence based on the characteristic representation of missing values ​​in the data (such as NaN, None, an empty string, etc.) to determine whether the energy efficiency data is missing. This is not limited here and can be set according to specific implementation scenarios. The direct judgment method is a conventional technique and is not described in detail here. If missing data is detected in the energy efficiency data sequence for any photovoltaic module on that day, the missing data in the energy efficiency data sequence needs to be filled. Because different photovoltaic modules have different data collection conditions, namely, different interference conditions and module aging conditions, the credibility of the reference energy efficiency data of each reference photovoltaic module varies, which in turn affects the accuracy of the filled data. Therefore, it is necessary to analyze the credibility of the reference energy efficiency data of each reference photovoltaic module based on the fluctuation of the reference energy efficiency data of each reference photovoltaic module and the historical data of the reference photovoltaic module to obtain more accurate filled data.

[0057] Therefore, for any missing data in the energy efficiency data sequence, reference data of any missing data is obtained in each reference energy efficiency data sequence, and historical reference data of each reference photovoltaic module is obtained within a preset historical number of days. Based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs, an initial reference value of any missing data is obtained. In this embodiment, the reference value of the preset historical number of days is 7 days (i.e., one week), which is not limited here and can be set according to the specific implementation scenario.

[0058] The method for obtaining the reference data of any missing data in each reference energy efficiency data sequence is as follows:

[0059] Recording the time at which any missing data occurs as a target time, and for any reference energy efficiency data sequence, obtaining reference energy efficiency data at the target time in the any reference energy efficiency data sequence as reference data for any missing data;

[0060] If the reference energy efficiency data at the target moment in any of the reference energy efficiency data sequences is missing data, obtaining the average of the reference energy efficiency data at the previous moment and the next moment before the target moment in any of the reference energy efficiency data sequences as the reference data for any missing data;

[0061] If at least one of the reference energy efficiency data at the previous moment and the next moment of the target moment in any reference energy efficiency data sequence is missing data, a constant 0 is used as the reference data of any missing data.

[0062] Furthermore, historical reference energy efficiency data of each reference photovoltaic module at the same target time every day in the previous week is obtained to obtain the historical reference data of each reference photovoltaic module every day in the previous week. For example, assuming that the target time (i.e., the time at which any missing data is located) is 2:00:00 p.m., then 5 reference data are obtained in the reference energy efficiency data sequence of 5 reference photovoltaic modules. Taking the first reference data as an example, the historical reference energy efficiency data of the reference photovoltaic module to which the first reference data belongs at 2:00:00 p.m. every day in the previous week is obtained as the historical reference data of the reference photovoltaic module to which the first reference data belongs every day in the previous week.

[0063] Finally, based on the data fluctuation characteristics of each reference energy efficiency data series and the difference between each reference data and the historical reference data of the corresponding reference PV module, the initial reference value of any missing data is obtained. The method for obtaining the initial reference value of any missing data is as follows:

[0064] (1) For any reference data, obtain the overall reliability of any reference data.

[0065] Since frequent data loss may be caused by sensor failure, communication interruption or equipment abnormality, and in real environment, the output power of photovoltaic modules cannot exceed the theoretical maximum value unless there is a sensor failure or data error, the overall reliability of any reference data can be obtained based on the data fluctuation characteristics of the reference energy efficiency data series to which the reference data belongs.

[0066] Specifically, the reference energy efficiency data sequence to which any reference data belongs is recorded as a reference sequence, the number of missing data in the reference sequence is obtained and recorded as the amount of missing data, the length of the reference sequence is obtained, and the ratio of the amount of missing data to the length of the reference sequence is obtained to obtain the degree of incompleteness of the reference sequence;

[0067] Obtaining rated energy efficiency data of a reference photovoltaic module to which any reference data belongs, obtaining the number of reference energy efficiency data in the reference sequence that is greater than the rated energy efficiency data, recording it as the amount of excess data, obtaining a ratio of the amount of excess data to the length of the reference sequence, and obtaining a degree of irrationality of the reference sequence;

[0068] The overall reliability of any reference data is obtained by substituting the inverse of the product of the incompleteness and irrationality of the reference sequence into an exponential function with a natural constant as the base.

[0069] In one embodiment, taking the i-th reference data as an example, the calculation formula for the overall reliability of the i-th reference data is:

[0070]

[0071] Among them, Ai is the overall reliability of the i-th reference data; Qi is the amount of missing data; Zi is the length of the reference sequence; Pi is the amount of excess data; It is an exponential function with a natural constant as base, used for inverse proportional normalization.

[0072] It should be noted that is the incompleteness of the i-th reference data. The larger Qi is, the more missing data there are in the reference sequence (i.e., the reference energy efficiency data sequence to which the i-th reference data belongs), and the more incomplete the reference sequence is. The larger it is, the less credible the i-th reference data is, and the smaller Ai is; is the unreasonable degree of the i-th reference data. The larger Pi is, the more data in the reference sequence (i.e., the reference energy efficiency data sequence to which the i-th reference data belongs) exceeds the rated energy efficiency. The sensor may be faulty or the data may be wrong. The more unreasonable the reference sequence is, The larger it is, the less credible the i-th reference data is, and the smaller Ai is.

[0073] (2) Obtain the local stability of any reference data.

[0074] The credibility of the reference data is not only related to the degree of fluctuation of the reference energy efficiency data sequence to which the reference data belongs, but also to the local fluctuation of the reference data and the consistency with the historical data. Therefore, with any reference data as the center, a reference window with a preset length of 29 is constructed in the reference energy efficiency data sequence to which any reference data belongs. There is no restriction here and it can be set according to the specific implementation scenario. According to the data fluctuation characteristics of the reference energy efficiency data in the reference window and the difference between any reference data and the historical reference data of the reference photovoltaic module to which it belongs, the local stability of any reference data is obtained.

[0075] Specifically, the isolation forest algorithm is used to obtain the number of abnormal data in the reference window, obtain the ratio of the number of abnormal data to the preset length, and obtain the abnormality degree of the reference window. The isolation forest algorithm belongs to the existing technology and is not described in detail here.

[0076] Obtaining an average of historical reference data of a reference photovoltaic module to which any reference data belongs, recording the average as a historical average, obtaining an absolute value of a difference between the historical average and any reference data, and obtaining a degree of deviation of the any reference data;

[0077] Obtain the coefficient of variation of the reference energy efficiency data in the reference window. The coefficient of variation belongs to the prior art and will not be described in detail here. Obtain the product of the degree of abnormality of the reference window, the coefficient of variation and the degree of deviation of any reference data, substitute the opposite of the product into an exponential function with a natural constant as the base, and obtain the local stability of any reference data.

[0078] In one embodiment, taking the i-th reference data as an example, the calculation formula for the local stability of the i-th reference data is:

[0079]

[0080] Where Bi is the local stability of the i-th reference data; Yi is the number of abnormal data in the reference window; Vi is the preset length (i.e. the length of the reference window); Ci is the coefficient of variation; Ji is the historical mean; Hi is the i-th reference data; It is an exponential function with a natural constant as the base, used for inverse proportional normalization; is the absolute value symbol.

[0081] It should be noted that The abnormality degree of the reference window. The larger Yi is, the more abnormal data there are in the reference window. The larger is , the smaller is the credibility of the i-th reference data, and the smaller is Bi; the larger is Ci, the greater is the volatility of the reference energy efficiency data in the reference window, that is, the greater is the degree of noise interference, the smaller is the credibility of the i-th reference data, and the smaller is Bi; is the deviation degree of the i-th reference data, The larger it is, the less consistent the i-th reference data is with the historical reference data. That is, the more it deviates from the historical reference data, the lower the credibility of the i-th reference data is, and the smaller Bi is.

[0082] (3) Obtain the overall reliability and local stability of each reference data, and obtain the initial reference value of any missing data based on the overall reliability and local stability of each reference data.

[0083] Specifically, for any reference data, the product of the overall reliability and the local stability of the any reference data is obtained to obtain the reference degree of the any reference data, the reference degree of each reference data is obtained, and a corresponding reference degree cumulative value is obtained, and the ratio of the reference degree of the any reference data to the reference degree cumulative value is obtained to obtain the weight coefficient of the any reference data;

[0084] A weight coefficient for each reference data is obtained, and a weighted sum is performed on all reference data to obtain an initial reference value for any missing data.

[0085] In one embodiment, taking the dth missing data as an example, the calculation formula for the initial reference value of the dth missing data is:

[0086]

[0087] Wherein, Td is the initial reference value of the d-th missing data; Ai is the overall reliability of the i-th reference data of the d-th missing data; Bi is the local stability of the i-th reference data of the d-th missing data; Hi is the i-th reference data of the d-th missing data; and m is the number of reference data of the d-th missing data.

[0088] It should be noted that is the weight coefficient of the i-th reference data of the d-th missing data. The greater the overall reliability and local temperature of the i-th reference data of the d-th missing data, the greater the credibility of the i-th reference data of the d-th missing data, the greater the weight coefficient, and the greater the accuracy of the initial reference value of the d-th missing data.

[0089] At this point, the initial reference value of any missing data is obtained.

[0090] Step S103: Obtain a theoretical value of any missing data based on the environmental data and performance parameter data of any photovoltaic component, fit the energy efficiency data sequence to obtain a fitted value of any missing data, and correct the initial reference value based on the difference between the theoretical value and the fitted value of any missing data to obtain an optimized filling value of any missing data.

[0091] After obtaining the initial reference value of any missing data, the initial reference value needs to be corrected to ensure the consistency of the filled data with the original data (i.e., the missing data). Considering the coherence of any missing data in the energy efficiency data sequence, the energy efficiency data sequence is fitted to obtain the fitted value of any missing data, and the initial reference value is corrected. Taking into account the influence of the environment in which the photovoltaic module is located, that is, the greater the solar irradiance, the higher the working efficiency of the photovoltaic module. At the same time, the working efficiency of the photovoltaic module is the highest when the battery temperature in the photovoltaic module reaches the optimal operating temperature. Therefore, based on the environmental data and performance parameter data of any photovoltaic module, the theoretical value of any missing data is obtained. Based on the difference between the theoretical value and the fitted value of any missing data, the initial reference value is corrected to obtain the optimized filling value of any missing data.

[0092] First, in this embodiment, the autoregressive integrated moving average (ARIMA) model is used to fit the energy efficiency data series to obtain the fitting value of any missing data. The autoregressive integrated moving average (ARIMA) model belongs to the existing technology and will not be described in detail here.

[0093] Secondly, according to the environmental data and performance parameter data of any photovoltaic module, the theoretical value of any missing data is obtained.

[0094] Specifically, the environmental data of any photovoltaic module includes irradiance and battery temperature. In this embodiment, a temperature sensor is used to obtain the battery temperature of any photovoltaic module, and a solar radiation sensor or data from the website of the local meteorological bureau is used to obtain the irradiance of any photovoltaic module. There is no limitation here and it can be set according to the specific implementation scenario; the performance parameter data of any photovoltaic module includes the effective area of ​​the photovoltaic module, efficiency under standard test conditions, power temperature coefficient, optimal operating temperature, and annual attenuation rate. In this embodiment, the performance parameter data of any photovoltaic module is obtained according to the nameplate and product specification of any photovoltaic module;

[0095] Recording the time at which any missing data occurs as a target time, obtaining the product of the irradiance of any photovoltaic module at the target time, the effective area of ​​the photovoltaic module, and the efficiency under the standard test condition, to obtain the theoretical output power of any photovoltaic module at the target time;

[0096] Obtaining an absolute value of a difference between a battery temperature of any photovoltaic module at the target time and the optimal operating temperature to obtain a temperature difference, obtaining a product of the power temperature coefficient and the temperature difference and adding a constant 1 to obtain a temperature correction factor of any photovoltaic module at the target time;

[0097] Obtaining the service life of any photovoltaic module, obtaining a difference between a constant 1 and the annual attenuation rate, substituting the service life into an exponential function with the difference as the base, and obtaining the operating efficiency of any photovoltaic module at the target time;

[0098] The product of the theoretical output power, the temperature correction factor and the working efficiency is obtained to obtain a theoretical value of any missing data.

[0099] In one embodiment, taking the dth missing data as an example, the calculation formula for the theoretical value of the dth missing data is:

[0100]

[0101] Where Sd is the theoretical value of the dth missing data; G is the irradiance of any PV module at the target time; F is the effective area of ​​the PV module of any PV module; The standard test condition efficiency of any photovoltaic module; is the power temperature coefficient of any photovoltaic module; is the battery temperature of any photovoltaic module at the target time; t is the optimal operating temperature of any photovoltaic module; is the annual degradation rate of any photovoltaic module; The service life of any photovoltaic module; is an absolute value symbol; in this embodiment, the reference value of the optimal operating temperature is 25°C, the reference value of the power temperature coefficient is -0.0045 / °C, and the reference value of the annual attenuation rate is 0.005. There is no restriction here and it can be set according to the specific implementation scenario.

[0102] It should be noted that is the theoretical output power of any photovoltaic module at the target time, The larger it is, the larger the theoretical value of the dth missing data is; is the temperature correction factor of any photovoltaic module at the target time, The larger the value is, the more the battery temperature of any photovoltaic module at the target time deviates from the optimal operating temperature. The smaller it is, the smaller the theoretical value of the dth missing data is; is the working efficiency of any photovoltaic module at the target time, and the service life of any photovoltaic module The larger , the more serious the aging of any photovoltaic module, the lower the working efficiency, and the smaller the theoretical value of the dth missing data.

[0103] Finally, according to the difference between the theoretical value and the fitted value of any missing data, the initial reference value is corrected to obtain the optimized filling value of any missing data.

[0104] Specifically, the difference between the initial reference value and the fitting value is obtained, recorded as the fitting difference, the difference between the initial reference value and the theoretical value is obtained, recorded as the theoretical difference, the average between the fitting difference and the theoretical difference is obtained, recorded as the correction value, the addition result between the initial reference value and the correction value is obtained, and the optimized filling value of any missing data is obtained.

[0105] In one embodiment, taking the dth missing data as an example, the calculation formula for the optimized filling value of the dth missing data is:

[0106]

[0107] in, is the optimized filling value of the d-th missing data; Td is the initial reference value of the d-th missing data; Nd is the fitted value of the d-th missing data; Sd is the theoretical value of the d-th missing data.

[0108] It should be noted that the greater the difference between the initial reference value of the d-th missing data and the fitted value or theoretical value, the more the initial reference value of the d-th missing data deviates from the original data (i.e., missing data), and the more the initial reference value of the d-th missing data needs to be corrected.

[0109] At this point, the optimized filling value of any missing data is obtained.

[0110] Step S104 , obtaining an optimized filling value for each missing data in the energy efficiency data sequence, and obtaining a complete energy efficiency data sequence of any photovoltaic component for transmission to a cloud computing platform for online processing.

[0111] According to the above method for obtaining the optimized filling value of any missing data, the optimized filling value of each missing data in the energy efficiency data sequence is obtained to obtain a complete energy efficiency data sequence of any photovoltaic component.

[0112] According to the method for obtaining the complete energy efficiency data sequence of any photovoltaic module, the complete energy efficiency data sequence of each photovoltaic module is obtained. Using the Internet of Things technology, the complete energy efficiency data sequence of each photovoltaic module is transmitted to the gateway device. The gateway device performs preliminary processing and integration on the received data, and then sends the data to the cloud computing platform via the Internet. The cloud computing platform's batch processing technology (such as Apache Hadoop and Apache Spark) is used to batch process and analyze the energy efficiency data of each photovoltaic module, calculate various indicators of the photovoltaic module, and evaluate the performance of the photovoltaic module. At the same time, the historical operating data of the photovoltaic module can be analyzed through machine learning algorithms to establish a fault diagnosis model. When new data is input, the fault diagnosis model can determine whether there is a fault in the photovoltaic module and the type of fault.

[0113] The focus of the present invention is to obtain accurate and complete energy efficiency data, use the Internet of Things technology to transmit the complete energy efficiency data sequence to the gateway device, the gateway device performs preliminary processing and integration of the received data, and uses the batch processing technology of the cloud computing platform to batch process and analyze the energy efficiency data of each photovoltaic module. This is existing technology and will not be repeated here.

[0114] In an embodiment of the present invention, for any photovoltaic component, a preset number of reference photovoltaic components of the any photovoltaic component are obtained, and energy efficiency data of the any photovoltaic component and each reference photovoltaic component at each moment are obtained to obtain an energy efficiency data sequence of the any photovoltaic component on that day and a reference energy efficiency data sequence of each reference photovoltaic component on that day; if missing data is detected in the energy efficiency data sequence, reference data of the any missing data is obtained in each reference energy efficiency data sequence, and historical reference data of each reference photovoltaic component is obtained within a preset number of historical days. According to the data fluctuation characteristics of each reference energy efficiency data sequence and each reference The method comprises the following steps: first, obtaining an initial reference value for any missing data based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs; then, obtaining a theoretical value for any missing data based on the environmental data and performance parameter data of any photovoltaic module, fitting the energy efficiency data sequence to obtain a fitted value for any missing data, and correcting the initial reference value based on the difference between the theoretical value and the fitted value for any missing data to obtain an optimized filling value for any missing data; and obtaining an optimized filling value for each missing data in the energy efficiency data sequence to obtain a complete energy efficiency data sequence for any photovoltaic module for transmission to a cloud computing platform for online processing. The method comprises the following steps: first, obtaining an initial reference value for any missing data based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs; then, correcting the initial reference value based on the difference between the theoretical value and the fitted value for any missing data to obtain an optimized filling value for any missing data. This method not only integrates the reference data of each reference photovoltaic module of the same type, but also combines the local characteristics of the energy efficiency data sequence of any photovoltaic module, thereby ensuring data integrity while improving the accuracy of the filled data, and accurately and completely obtaining the energy efficiency data of the photovoltaic module.

[0115] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A cloud computing-based online processing method for photovoltaic module energy efficiency data, characterized in that: The method comprises: For any photovoltaic module, obtaining a preset number of reference photovoltaic modules of the any photovoltaic module, obtaining energy efficiency data of the any photovoltaic module and each reference photovoltaic module at each moment, and obtaining an energy efficiency data sequence of the any photovoltaic module on that day and a reference energy efficiency data sequence of each reference photovoltaic module on that day; If missing data is detected in the energy efficiency data sequence, for any missing data, reference data of the missing data is obtained in each reference energy efficiency data sequence, historical reference data of each reference photovoltaic module is obtained within a preset number of historical days, and an initial reference value of the missing data is obtained based on the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the reference photovoltaic module to which it belongs; Obtaining, based on the environmental data and performance parameter data of any photovoltaic module, a theoretical value of any missing data, fitting the energy efficiency data sequence to obtain a fitted value of any missing data, obtaining a difference between the initial reference value and the fitted value, recording it as a fitted difference, obtaining a difference between the initial reference value and the theoretical value, recording it as a theoretical difference, obtaining an average between the fitted difference and the theoretical difference, recording it as a corrected value, and correcting the initial reference value using the corrected value to obtain an optimized filled value for any missing data; Obtaining an optimized filling value for each missing data in the energy efficiency data sequence to obtain a complete energy efficiency data sequence of any photovoltaic component for transmission to a cloud computing platform for online processing; The obtaining of the initial reference value of any missing data according to the data fluctuation characteristics of each reference energy efficiency data sequence and the difference between each reference data and the historical reference data of the corresponding reference photovoltaic module includes: For any reference data, substitute the inverse of the product of the degree of incompleteness and the degree of irrationality of the reference energy efficiency data series to which the reference data belongs into an exponential function with a natural constant as the base to obtain the overall reliability of the reference data; Taking any reference data as the center, construct a reference window of a preset length in the reference energy efficiency data sequence to which the any reference data belongs, obtain the product of the degree of abnormality of the reference window, the coefficient of variation of the reference energy efficiency data in the reference window, and the degree of deviation of the any reference data, substitute the inverse of the product into an exponential function with a natural constant as the base, and obtain the degree of local stability of the any reference data; The overall reliability and local stability of each reference data are obtained, and the initial reference value of any missing data is obtained according to the overall reliability and local stability of each reference data.

2. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The method for obtaining the degree of incompleteness and unreasonableness of the reference energy efficiency data sequence to which any reference data belongs includes: Recording a reference energy efficiency data sequence to which any reference data belongs as a reference sequence, obtaining the number of missing data in the reference sequence as the amount of missing data, obtaining the length of the reference sequence, and obtaining a ratio of the amount of missing data to the length of the reference sequence to obtain a degree of incompleteness of the reference sequence; Obtain the rated energy efficiency data of the reference photovoltaic module to which any reference data belongs, obtain the number of reference energy efficiency data in the reference sequence that is greater than the rated energy efficiency data, record it as the amount of excess data, obtain the ratio of the amount of excess data to the length of the reference sequence, and obtain the degree of irrationality of the reference sequence.

3. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 2, characterized in that: The method for obtaining the abnormality degree of the reference window and the deviation degree of any reference data includes: Using an isolation forest algorithm, obtaining the number of abnormal data in the reference window, obtaining the ratio of the number of abnormal data to the preset length, and obtaining the abnormality degree of the reference window; The average value of historical reference data of the reference photovoltaic assembly to which any reference data belongs is obtained, recorded as the historical average value, and the absolute value of the difference between the historical average value and any reference data is obtained to obtain the degree of deviation of any reference data.

4. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The obtaining of the initial reference value of any missing data according to the overall reliability and local stability of each reference data includes: For any reference data, obtain the product of the overall reliability and the local stability of the any reference data to obtain the reference level of the any reference data, obtain the reference level of each reference data, obtain a corresponding reference level cumulative value, obtain the ratio of the reference level of the any reference data to the reference level cumulative value, and obtain a weight coefficient for the any reference data; A weight coefficient for each reference data is obtained, and a weighted sum is performed on all reference data to obtain an initial reference value for any missing data.

5. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The obtaining of a theoretical value of any missing data according to the environmental data and performance parameter data of any photovoltaic module includes: The environmental data of any photovoltaic module include irradiance and battery temperature, and the performance parameter data of any photovoltaic module include effective area of ​​photovoltaic module, efficiency under standard test conditions, power temperature coefficient, optimal operating temperature, and annual attenuation rate; Recording the time at which any missing data occurs as a target time, obtaining the product of the irradiance of any photovoltaic module at the target time, the effective area of ​​the photovoltaic module, and the efficiency under the standard test condition, to obtain the theoretical output power of any photovoltaic module at the target time; Obtaining an absolute value of a difference between a battery temperature of any photovoltaic module at the target time and the optimal operating temperature to obtain a temperature difference, obtaining a product of the power temperature coefficient and the temperature difference and adding a constant 1 to obtain a temperature correction factor of any photovoltaic module at the target time; Obtaining the service life of any photovoltaic module, obtaining a difference between a constant 1 and the annual attenuation rate, substituting the service life into an exponential function with the difference as the base, and obtaining the operating efficiency of any photovoltaic module at the target time; The product of the theoretical output power, the temperature correction factor and the working efficiency is obtained to obtain a theoretical value of any missing data.

6. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The using the correction value to correct the initial reference value to obtain the optimized filling value of any missing data includes: The sum of the initial reference value and the correction value is obtained to obtain an optimized filling value of any missing data.

7. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The step of obtaining reference data of any missing data in each reference energy efficiency data sequence includes: Recording the time at which any missing data occurs as a target time, and for any reference energy efficiency data sequence, obtaining reference energy efficiency data at the target time in the any reference energy efficiency data sequence as reference data for any missing data; If the reference energy efficiency data at the target moment in any of the reference energy efficiency data sequences is missing data, obtaining the average of the reference energy efficiency data at the previous moment and the next moment before the target moment in any of the reference energy efficiency data sequences as the reference data for any missing data; If at least one of the reference energy efficiency data at the previous moment and the next moment of the target moment in any reference energy efficiency data sequence is missing data, a constant 0 is used as the reference data of any missing data.

8. The method for online processing of photovoltaic module energy efficiency data based on cloud computing according to claim 1, characterized in that: The obtaining of a preset number of reference photovoltaic assemblies of any photovoltaic assembly includes: A photovoltaic assembly of the same model and located in the same region as any one of the photovoltaic assemblies is obtained as a reference photovoltaic assembly for the any one of the photovoltaic assemblies.

Images