A cloud-shading-free identification method for a photovoltaic power station

Abstract

The application provides a cloud-shading identification method for a photovoltaic power station, which is based on meteorological collection data of the photovoltaic power station, can quickly and accurately identify cloud-shading conditions of the photovoltaic power station on a certain day, and thus determines external conditions for executing a photovoltaic module shading algorithm. The method comprises the following steps: S01: obtaining photovoltaic irradiance data I rr on a certain day; S02: identifying abnormal data in the irradiance data I rr , replacing the abnormal data by using an interpolation method considering boundary conditions, and obtaining new irradiance data IRR; S03: identifying the number of inflection points with a given amplitude in the irradiance data IRR and the number of data points in a peak-valley inflection point interval range; S04: calculating a data point proportion in the inflection point interval; S05: judging that the data point proportion in the inflection point interval is less than 2% as a quasi-cloudless day and greater than 2% as a cloudy day; S06: judging the irradiance data of the quasi-cloudless day, and if the average irradiance within 30 minutes before and after the midday time is greater than 700 W / m 2 , the day is judged as a cloudless shading day.

Description

Technical Field

[0001] This invention relates to the technical field of photovoltaic power generation, specifically to a method for identifying cloud-free conditions in photovoltaic power plants. Background Technology

[0002] In the intelligent diagnosis of photovoltaic power plants, there are currently two main methods for diagnosing shading of photovoltaic modules: image recognition and data analysis. Image recognition relies on drones to scan with visible and infrared light to detect shading issues, but the limited number of drone flights restricts its timeliness, and drones are relatively expensive. Data analysis, on the other hand, is widely studied due to its low cost and high timeliness. However, cloud cover can reduce the accuracy of data analysis, often requiring cloud exclusion. Traditional methods rely on weather forecasts, but accurately identifying cloud cover is difficult, and these methods require external data access, leading to high costs and cybersecurity risks.

[0003] The disadvantages of existing technologies are as follows:

[0004] (1) Image recognition methods are costly and have poor timeliness.

[0005] The photovoltaic module shading identification method based on image recognition requires the configuration of drones and visible light and infrared cameras, which increases the hardware cost; moreover, drones are limited by flight time, flight distance and weather conditions, so they cannot perform real-time diagnosis and have poor timeliness.

[0006] (2) Weather forecasting-based methods are costly and pose cybersecurity risks.

[0007] Methods based on weather forecast data require detailed weather data, which increases data access costs; moreover, the data needs to be obtained from the external network, which poses certain network security risks; and weather forecast data cannot effectively identify the presence of clouds. Summary of the Invention

[0008] To address the aforementioned issues, this invention provides a method for identifying cloud cover in photovoltaic power plants. Based on meteorological data collected from photovoltaic power plants, it can quickly and accurately identify the cloud cover situation of a photovoltaic power plant on a given day, thereby determining the external conditions for the execution of the photovoltaic module shading algorithm.

[0009] A method for identifying cloud-free conditions in photovoltaic power plants, characterized by comprising the following steps: S01:

[0010] Obtain photovoltaic irradiance data for a certain day I rr ;

[0011] S02: Identifying Irradiance Data rrAbnormal data in the data were identified, and an interpolation method considering boundary conditions was used to replace the abnormal data to obtain new irradiance data IRR.

[0012] S03: Identify the number of inflection points for a given amplitude in the irradiance data IRR, and the number of data points within the peak-valley inflection point interval;

[0013] S04: Calculate the proportion of data points within the inflection interval;

[0014] S05: Determine whether the proportion of data points within the inflection point interval is less than 2% as cloudless weather, and whether the proportion is greater than 2% as cloudy weather.

[0015] S06: Judge the irradiance data for cloudless weather. If the average irradiance within 30 minutes before and after midday is greater than 700W / ㎡, then judge it as a cloudless day.

[0016] Its further features are:

[0017] The specific steps for identifying and replacing abnormal data in step S02 are as follows:

[0018] S02-01: Compare the irradiance data I sequentially rr Is the value in I less than 0 or greater than I? max , where I max This represents the highest irradiance in the region where the photovoltaic power station is located in the past 10 years.

[0019] S02-02: Record irradiance data as less than 0 or greater than 1. max The data is indexed into a list of numbers;

[0020] S02-03: Identify the continuity of data in a sequence list to obtain single data points and continuous data intervals;

[0021] S02-04: Determine the location of the obtained single-point data and continuous data intervals. If the abnormal data is in I... rr For the starting position, substitute using the following formula:

[0022] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr The index in the table, where n is the number of data points in a single point or a continuous interval;

[0023] If abnormal data is in I rr If the last position is not specified, then replace it according to the following formula:

[0024] In the formula: S i For I rrAbnormal data in I, where i represents the abnormal data in I. rr Index in S -1 The first non-abnormal data point preceding a single point or a continuous interval;

[0025] If the data is not at the beginning or end, replace it according to method S02-05.

[0026] S02-05: If both single-point data and continuous interval data occur before midday, replace the abnormal data using the following formula:

[0027] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr Index in S -1 This is the first non-abnormal data point before a single point or continuous interval, where n is the number of data points in the single point or continuous interval.

[0028] If the data for a single point or a continuous interval is all after midday, replace the outlier data using the following formula:

[0029] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr Index in S -1 This represents the first non-abnormal data point before a single point or continuous interval, where n is the number of data points in the single point or continuous interval.

[0030] Furthermore, the midday time in steps S02-05 is calculated using the following formula:

[0031] In the formula: m is the index corresponding to midday time, and N is the index of I. rr The amount of data in the data.

[0032] The identification of the number of inflection points for a given amplitude in the irradiance data in step S03, and the number of data points within the peak-valley inflection point interval, are performed according to the following steps:

[0033] S03-01: Based on the amount of data in the IRR, determine the interval length step used to determine whether a point of data is an inflection point;

[0034] S03-02: Set a limit value std for filtering out minor noise points;

[0035] S03-03: Determine the inflection point for each point in the data IRR using the following method.

[0036] In step S04, the proportion of data points within the inflection interval is calculated using the following formula:

[0037]

[0038] Where: Num z Let N be the number of data points at each inflection point, and I be the number of data points at each inflection point. rr The amount of data in the data.

[0039] After adopting the above technical solution, cloudless weather identification based on irradiance data is possible without the need for external equipment or external data access, making it safe, reliable, and low-cost. The algorithm can identify short-term cloud cover with high efficiency and accuracy. Based on power plant operation data diagnosis, it is more practical. Its cloudless weather identification based on irradiance data can conveniently and quickly identify whether there is cloud cover in historical weather, providing technical support for photovoltaic module failure shading. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the operation process of the present invention;

[0041] Figure 2 This is a flowchart illustrating step S02 of the present invention;

[0042] Figure 3 This is a flowchart illustrating step S03 of the present invention;

[0043] Figure 4 Photovoltaic irradiance data I for a specific embodiment of the present invention rr Data curves;

[0044] Figure 5 This is an IRR data curve diagram of a specific embodiment of the present invention;

[0045] Figure 6 This is the IRR cloud obstruction point identification result of a specific embodiment of the present invention;

[0046] Figure 7 Photovoltaic irradiance data I for a specific embodiment of the present invention rr Data curves;

[0047] Figure 8 This is an IRR data curve diagram of a specific embodiment two of the present invention;

[0048] Figure 9 This is the IRR cloud occlusion point identification result of a specific embodiment two of the present invention. Detailed Implementation

[0049] A method for identifying cloud-free shading in photovoltaic power plants, see [link / reference]. Figures 1-3 It includes the following steps: S01: Obtain photovoltaic irradiance data for a certain day I rr .

[0050] S02: Identifying Irradiance Data rr Abnormal data in the data were identified, and an interpolation method considering boundary conditions was used to replace the abnormal data to obtain new irradiance data IRR.

[0051] The specific steps for identifying and replacing abnormal data in step S02 are as follows:

[0052] S02-01: Compare the irradiance data I sequentially rr Is the value in I less than 0 or greater than I? max , where I max This represents the highest irradiance in the region where the photovoltaic power station is located in the past 10 years.

[0053] S02-02: Record irradiance data as less than 0 or greater than 1. max The data is indexed into a list of numbers;

[0054] S02-03: Identify the continuity of data in a sequence list to obtain single data points and continuous data intervals;

[0055] S02-04: Determine the location of the obtained single-point data and continuous data intervals. If the abnormal data is in I... rr For the starting position, substitute using the following formula:

[0056] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr The index in the table, where n is the number of data points in a single point or a continuous interval;

[0057] If abnormal data is in I rr If the last position is not specified, then replace it according to the following formula:

[0058] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr Index in S -1 The first non-abnormal data point preceding a single point or a continuous interval;

[0059] If the data is not at the beginning or end, replace it according to method S02-05.

[0060] S02-05: If both single-point data and continuous interval data occur before midday, replace the abnormal data using the following formula:

[0061] In the formula: S i For I rrAbnormal data in I, where i represents the abnormal data in I. rr Index in S -1 This is the first non-abnormal data point before a single point or continuous interval, where n is the number of data points in the single point or continuous interval.

[0062] If the data for a single point or a continuous interval is all after midday, replace the outlier data using the following formula:

[0063] In the formula: S i For I rr Abnormal data in I, where i represents the abnormal data in I. rr Index in S -1 This represents the first non-abnormal data point before a single point or continuous interval, where n is the number of data points in the single point or continuous interval.

[0064] The midday time in steps S02-05 is calculated using the following formula:

[0065] In the formula: m is the index corresponding to midday time, and N is the index of I. rr The amount of data in the data.

[0066] S03: Identify the number of inflection points for a given amplitude in the irradiance data IRR, and the number of data points within the peak-valley inflection point interval;

[0067] The identification of the number of inflection points for a given amplitude in the irradiance data in step S03, as well as the number of data points within the peak-valley inflection point interval, is performed according to the following steps.

[0068] S03-01: Based on the amount of data in the IRR, determine the interval length step used to determine whether a point of data is an inflection point;

[0069] S03-02: Set a limit value std for filtering out minor noise points;

[0070] S03-03: Determine the inflection point for each point in the data IRR using the following method.

[0071] S04: Calculate the proportion of data points within the inflection point interval; the proportion of data points within the inflection point interval in step S04 is calculated using the following formula:

[0072] Where: Num z Let N be the number of data points at each inflection point, and I be the number of data points at each inflection point. rr The amount of data in the data.

[0073] S05: Determine whether the proportion of data points within the inflection point interval is less than 2% as cloudless weather, and greater than 2% as cloudy weather.

[0074] S06: Judge the irradiance data for cloudless weather. If the average irradiance within 30 minutes before and after midday is greater than 700W / ㎡, then judge it as a cloudless day.

[0075] Specific implementation steps for Embodiment 1 are as follows:

[0076] S01: Obtain photovoltaic irradiance data for a specific day. rr Data curves as follows Figure 4 As shown:

[0077] S02: Identifying Irradiance Data rr The abnormal data in the data were identified, and an interpolation method considering boundary conditions was used to replace the abnormal data to obtain new irradiance data (IRR). The obtained IRR data curve is shown in the figure. Figure 5 As shown;

[0078] S03: Identify the number of inflection points for a given amplitude in the irradiance data IRR, and the number of data points within the peak-to-valley inflection point interval; identify the shaded points, such as... Figure 6 As indicated by the red dot;

[0079] S04: The proportion of data points within the inflection interval is calculated to be 12.924%;

[0080] S05: If the proportion of data points within the inflection point interval is greater than 2%, it is judged as cloudy weather.

[0081] The specific implementation steps for Specific Embodiment Two are as follows:

[0082] S01: Obtain photovoltaic irradiance data for a specific day. rr Data curves as follows Figure 7 As shown:

[0083] S02: Identifying Irradiance Data rr The abnormal data in the data were identified, and an interpolation method considering boundary conditions was used to replace the abnormal data to obtain new irradiance data (IRR). The obtained IRR data curve is shown in the figure. Figure 8 As shown;

[0084] S03: Identify the number of inflection points for a given amplitude in the irradiance data IRR, and the number of data points within the peak-to-valley inflection point interval; identify the shaded points, such as... Figure 9 As indicated by the red dot;

[0085] S04: The proportion of data points within the inflection interval is calculated to be 1.54%;

[0086] S05: Cloudless weather is defined as when the proportion of data points within the inflection interval is less than 2%.

[0087] S06: Based on the irradiance data for cloudless weather, if the average irradiance within 30 minutes before and after midday is greater than 700W / ㎡, the weather for that day is determined to be cloudless.

[0088] In summary, it can achieve real-time and accurate fault diagnosis of photovoltaic strings without increasing costs, thereby improving the shading recognition rate of photovoltaic modules.

[0089] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0090] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for identifying cloud-free shading of a photovoltaic power plant, characterized in that, It comprises the following steps: S01: obtaining certain day photovoltaic irradiance data I rr ; S02: identifying abnormal data in the irradiance data I rr and replacing the abnormal data with an interpolation method considering boundary conditions to obtain new irradiance data IRR ; S03: identifying irradiance data IRR The number of inflection points and the number of data points in the peak-valley inflection point interval range are given in the middle. S04: Calculate the proportion of data points in the fold interval; S05: Determine whether the proportion of data points in the fold interval is less than 2% for clear weather, or greater than 2% for cloudy weather; S06: Determine the irradiance data for clear weather, if the average irradiance within 30 minutes before and after noon is greater than 700 W / m2, then determine it as a cloudless sheltered day.

2. The method for identifying cloud shadow of a photovoltaic power station according to claim 1, characterized in that, The specific steps of the abnormal data recognition and replacement of step S02 are as follows: S02-01: sequentially comparing the irradiance data I rr whether the value in the formula (1) is less than 0 or greater than I max wherein I max is the maximum irradiance in the region where the photovoltaic power station is located in the last 10 years. S02-02: record irradiance data less than 0 or greater than I max Data index to a list List; S02-03: Identify the continuity of data in the list to obtain single-point data and continuous data interval; S02-04: judging the position of the obtained single-point data and continuous data interval, if the abnormal data is in I rr If the start position, the following formula is used for replacement: (1) In the formula: S i is I rr abnormal data, i is the index of the abnormal data in I rr abnormal data, i is the index of the abnormal data in If the abnormal data is at the end position of I rr then it is replaced by the following formula: (2) wherein: S i is the index of the abnormal data in the abnormal data, I rr I rr is the index of the abnormal data in the abnormal data, S -1 is the first non-abnormal data before the single point or continuous interval.​ If the data is not at the beginning or end, replace it according to the method of S02-05 S02-05: If both single-point data or data of consecutive intervals are before noon, replace the abnormal data according to the following formula: (3) In the formula: S i is the index of the abnormal data in the abnormal data, I rr I rr is the index of the abnormal data in the abnormal data, S -1 is the index of the abnormal data in the abnormal data,​ If the single point data or the data of continuous interval are all after the daytime, the abnormal data are replaced by the following formula: (4) In the formula: S i is the index of the abnormal data in the abnormal data, I rr I rr is the index of the abnormal data in the abnormal data, S -1 is the index of the abnormal data in the abnormal data,​ 3. The method for identifying cloud shadow of a photovoltaic power station according to claim 2, characterized in that: Further, the time of day in step S02-05 is calculated according to the following formula: (5) In the formula: m is an index corresponding to a time of day, N is I rr the number of data in 4. The method for identifying cloud shadow of a photovoltaic power station according to claim 1, characterized in that, The identification of the fold point of a given amplitude in the irradiance data and the number of data points in the peak-valley fold interval range in step S03 is performed as follows: S03-01: determining an interval length step for judging whether a point data is a break point according to IRR a data amount; S03-02: Set a limit value std for filtering out small jitter points; S03-03: On each point in the data IRR a break point judgment is made as follows; The calculation of the proportion of data points in the fold interval in step S04 is performed according to the following formula: （5） In the formula: Num z is the amount of data at the fold point, N is I rr the amount of data in

Images