A method, apparatus, and medium for anomaly detection of a solar panel

Abstract

The application provides an abnormality detection method, medium and device of a solar panel, relates to the technical field of data processing, and comprises the following steps: acquiring current characteristics DA and voltage characteristics DU of a to-be-detected solar panel within a target time window; obtaining an initial abnormality degree value corresponding to the to-be-detected solar panel according to the DA, the DU and an abnormality detection model; if the initial abnormality degree value is greater than a first threshold value and smaller than a second threshold value, obtaining a key solar panel list G according to the DA and the DU; if the n / a is greater than a preset same-area proportion threshold value, acquiring a third threshold value; and if the initial abnormality degree value is greater than the third threshold value, determining that the to-be-detected solar panel is abnormal. The application can more accurately distinguish whether the fluctuation of the initial abnormality degree value is caused by self failure or external environmental influence, gradually eliminates the interference of the external environment on the current and voltage characteristics, and finally improves the accuracy of the abnormality detection result.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, medium, and device for detecting anomalies in solar panels. Background Technology

[0002] In the global transition to clean energy, solar energy, as an inexhaustible and renewable energy source, has become an important component of national energy strategies. As the core equipment of a solar power generation system, the stability and reliability of solar panels directly determine power generation efficiency and system lifespan. Therefore, real-time and accurate anomaly detection of solar panels, and timely identification and troubleshooting of faults (such as component aging, partial shading, and short circuits), are crucial for ensuring the efficient operation of solar power generation systems.

[0003] Currently, most anomaly detection methods for solar panels in the industry employ monitoring methods based on electrical parameters. The core of this approach is to collect the output current and voltage characteristics of the solar panel and compare them with preset normal thresholds or standard curves to determine if any abnormalities exist. This method is widely used in practical applications due to its simple principle, ease of implementation, and low cost.

[0004] However, solar panels operate in a highly outdoor environment, making their operation extremely susceptible to interference from external environmental factors. Specifically, drastic fluctuations in sunlight intensity (such as cloud cover, changes in light intensity during sunrise and sunset, rain, and snow) directly affect the photoelectric conversion efficiency of solar panels, leading to nonlinear changes in output current. Increases or decreases in ambient temperature alter the electrical properties of semiconductor materials, causing deviations in output voltage. Furthermore, weather conditions such as rain, snow, and dust storms not only weaken sunlight intensity but can also cause surface contamination of solar panels, further exacerbating abnormal fluctuations in current and voltage. Under these complex and ever-changing external environmental influences, existing methods for anomaly detection based solely on current and voltage characteristics have significant drawbacks: when current or voltage fluctuates, the system struggles to distinguish whether the fluctuation is a normal phenomenon caused by changes in the external environment or an abnormal state caused by a fault in the solar panel itself. This confusion directly leads to a significant decrease in the accuracy of the detection results. Therefore, effectively eliminating the interference of the external environment on current and voltage characteristics and improving the accuracy of solar panel anomaly detection has become a pressing technical problem to be solved in this field. Summary of the Invention

[0005] To address the aforementioned technical problems, this application provides a method, device, and medium for detecting anomalies in solar panels, which at least partially solves the problems existing in the prior art.

[0006] In a first aspect of this application, an anomaly detection method for a solar panel is provided, the method comprising:

[0007] Obtain the current characteristic DA and voltage characteristic DU of the solar panel under test within the target time window; where the end time of the target time window is the current time.

[0008] Based on DA, DU, and the anomaly detection model, the initial anomaly level value corresponding to the solar panel under test is obtained; wherein, the anomaly detection model has a first threshold and a second threshold; the second threshold is greater than the first threshold;

[0009] If the initial anomaly level is greater than the first threshold and less than the second threshold, then based on DA and DU, the list of critical solar panels G = (G1, G2, ..., G...) is obtained. i , ..., G n ); i = 1, 2, ..., n; where n is the number of critical solar panels; G i Let be the i-th key solar panel; the key solar panel is located within the area where the solar panel under test is located, and the matching degree between the corresponding key power generation characteristics and the power generation characteristics of the solar panel under test is greater than the preset matching degree threshold; the key power generation characteristics include the current characteristics and voltage characteristics of the corresponding key solar panel within the target time window; the power generation characteristics of the solar panel under test include DA and DU;

[0010] If n / a is greater than the preset threshold for the proportion of the same area, then a third threshold is obtained; wherein, the third threshold is greater than the first threshold, and the third anomaly value is less than or equal to the second threshold; the third threshold is determined based on the first threshold and n / a; a is the number of solar panels contained in the area where the solar panel to be tested is located;

[0011] If the initial anomaly level value is greater than the third threshold, the solar panel under test is determined to be abnormal.

[0012] In a second aspect of this application, a non-transitory computer-readable storage medium is provided, wherein at least one instruction or at least one program is stored in the storage medium, and the at least one instruction or at least one program is loaded and executed by a processor to implement the aforementioned method for detecting anomalies in solar panels.

[0013] In a third aspect of this application, an electronic device is provided, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0014] This application has at least the following beneficial effects:

[0015] The anomaly detection method for solar panels provided in this application first acquires the current characteristics (DA) and voltage characteristics (DU) of the solar panel under test within a target time window ending at the current time. Then, based on these characteristics and an anomaly detection model, an initial anomaly level value is obtained. A preliminary classification is made using the model's first and second thresholds. If the initial value is between the two thresholds, it indicates that relying solely on the current and voltage characteristics of the solar panel under test is insufficient to determine whether the fluctuation is a genuine anomaly or caused by external environmental interference, requiring further verification. Next, when the initial value is between the first and second thresholds, several key solar panels within the same region whose power generation characteristics (current and voltage) match the solar panel under test higher than a preset threshold are selected. Since solar panels in the same region are likely affected by essentially the same external environment, a high characteristic matching degree means that their power generation characteristics are similar under similar environmental interference conditions. The characteristics of these key solar panels then become a reference for judging whether the characteristics of the solar panel under test are affected by the environment, introducing the concept of environmental characteristics in the same region. The system takes into account consistency. Furthermore, when the ratio (n / a) of the number of key solar panels in the region exceeds a preset threshold for the same region, it indicates that most solar panels in the region match the power generation characteristics of the solar panel under test. This means the fluctuations in the characteristics of the solar panel under test are more likely caused by shared environmental factors in the region. At this point, a third threshold is determined based on the first threshold and n / a, falling between the first and second thresholds. This allows for dynamic adjustment of the judgment criteria according to the degree of environmental consistency within the same region, making the threshold more closely aligned with the normal fluctuation range under environmental influence. Finally, when the initial anomaly level value exceeds the third threshold, the solar panel under test is determined to be abnormal. Since the third threshold is adjusted based on the proportion of key solar panels in the same region, it considers both the characteristics of the solar panel under test itself and the characteristic matching of solar panels in the same region affected by the same environment. This allows for a more accurate differentiation between fluctuations in the initial anomaly level value caused by its own fault or external environmental influence, gradually eliminating interference from the external environment on current and voltage characteristics, ultimately improving the accuracy of anomaly detection results. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart of an anomaly detection method for solar panels provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0020] It should be noted that the following description covers various aspects of embodiments within the scope of the appended claims. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0021] Please refer to Figure 1 As shown, an embodiment of this application provides a method for detecting anomalies in a solar panel, the method comprising:

[0022] S100, acquire the current characteristic DA and voltage characteristic DU of the solar panel under test within the target time window; where the end time of the target time window is the current time.

[0023] Specifically, the target time window refers to a continuous time period ending at the current time, used to collect time-correlated current and voltage data. For example, the target time window can be 5-15 minutes long; if the current time is 10:00, the window would be 9:45-10:00. This setting is based on the environmental response characteristics of the photovoltaic system. A window shorter than 5 minutes will be affected by transient interference, while a window longer than 15 minutes will obscure fault characteristics. Current characteristics (DA) are extracted from a list of current data collected within the target time window at a preset frequency (e.g., 1Hz). Current characteristics (DA) include, but are not limited to, the average current (e.g., 5.2A) and current fluctuation rate (variance or standard deviation of the current data list). Voltage characteristics (DU) are extracted from a list of voltage data collected within the target time window at a preset frequency. Voltage characteristics (DU) include, but are not limited to, voltage peak values ​​(e.g., 24.5V) and valley values ​​(e.g., 23.8V). Feature values ​​can be obtained through a sliding window algorithm.

[0024] S200, based on DA, DU and the anomaly detection model, obtain the initial anomaly level value corresponding to the solar panel under test; wherein, the anomaly detection model has a first threshold and a second threshold; the second threshold is greater than the first threshold.

[0025] Specifically, the anomaly detection model can be any model known to those skilled in the art capable of anomaly detection. As an example, the anomaly detection model can be an improved YOLOv8 model trained on the PVEL-AD dataset (containing 36,543 photovoltaic defect images and corresponding current / voltage features), with the input being a 4-dimensional feature vector composed of DA and DU (mean current, volatility, peak voltage, and valley voltage), and the output being a normalized outlier value of 0-1 (the higher the value, the greater the probability of an anomaly).

[0026] The first threshold is the critical value for distinguishing between normal and potential abnormal (e.g., 0.7), which is based on the distribution of abnormal values ​​in 1000 normal samples and is taken as the upper limit of the 95% confidence interval; the second threshold is the critical value for clearly determining an abnormality (e.g., 0.85), which is taken as the upper limit of the 95% confidence interval plus 3 standard deviations.

[0027] S300, if the initial anomaly level value is greater than the first threshold and less than the second threshold, then based on DA and DU, the list of critical solar panels G = (G1, G2, ..., G...) is obtained. i , ..., G n ); i = 1, 2, ..., n; where n is the number of critical solar panels; G iLet be the i-th key solar panel; the key solar panel is located in the area where the solar panel under test is located, and the matching degree between the corresponding key power generation characteristics and the power generation characteristics of the solar panel under test is greater than the preset matching degree threshold; the key power generation characteristics include the current characteristics and voltage characteristics of the corresponding key solar panel within the target time window; the power generation characteristics of the solar panel under test include DA and DU.

[0028] Specifically, the area where the solar panel under test is located can be an artificially defined area or a pre-defined administrative region (such as a district of a city). Key power generation characteristics include the current and voltage characteristics of the corresponding key solar panel within the target time window. The data acquisition and feature extraction methods for these key power generation characteristics are the same as in step S100 above. The matching degree between the key power generation characteristics of the key solar panel and the power generation characteristics of the solar panel under test can be calculated as follows: calculate the matching degree between the current characteristics and the matching degree between the voltage characteristics separately, and then perform a weighted sum to obtain the final matching degree; or combine the current and voltage characteristics into a comprehensive feature and then obtain the matching degree. Specific calculation methods can include matching degree calculation formulas such as cosine distance or Euclidean distance.

[0029] Here, when the initial value is between the first threshold and the second threshold, several key solar panels in the same area with a matching degree of higher than the preset threshold for the power generation characteristics (current and voltage) of the solar panel under test are selected. Since solar panels in the same area are likely to be affected by the same external environment, a high matching degree means that their power generation characteristics are similar under similar environmental interference. The characteristics of these key solar panels become a reference for judging whether the characteristics of the solar panel under test are affected by the environment, thus introducing the consideration of the consistency of the environment in the same area.

[0030] S400, if n / a is greater than the preset threshold for the proportion of the same area, then obtain the third threshold; wherein, the third threshold is greater than the first threshold, and the third abnormality value is less than or equal to the second threshold; the third threshold is determined according to the first threshold and n / a; a is the number of solar panels contained in the area where the solar panel to be tested is located.

[0031] Specifically, n / a is the ratio of the number of key solar panels n in the area where the solar panel under test is located to the total number a of solar panels in the area (e.g., if the area contains 100 solar panels, n = 60, then n / a = 60%).

[0032] Furthermore, the third threshold SZ conforms to the following characteristics: SZ = YZ + (EZ - YZ) × n / a; where YZ is the first threshold and EZ is the second threshold.

[0033] As an example: the preset threshold for the proportion of the same area is 50%. For example, when the first threshold = 0.7, the second threshold = 0.9, and n / a = 60%, the third threshold = 0.7 + 0.2 × 0.6 = 0.82.

[0034] Here, the larger the n / a ratio, the stronger the consistency of the influence of the same external environment (such as sunlight and temperature) on most solar panels in the same area—that is, the fluctuations in the current characteristic DA and voltage characteristic DU of the tested panel are more likely caused by common environmental factors in the area, rather than its own fault. At this time, it is necessary to raise the threshold for judging anomalies to avoid misjudging environmental interference as faults. In the formula, the (EZ-YZ)×n / a term increases with the increase of n / a, making the third threshold SZ approach the second threshold EZ (but never exceeding EZ). It is precisely by dynamically raising the threshold that the scenario with significant environmental influence is adapted: when n / a is close to 100% (almost all solar panels in the area have matching characteristics), SZ is close to EZ, which means that only when the initial anomaly degree value of the tested panel is close to the "clearly abnormal" EZ will it be judged as an anomaly, effectively filtering out characteristic fluctuations caused by environmental interference.

[0035] The setting of SZ not exceeding EZ is to preserve the "bottom-line" function of the second threshold EZ: EZ, as a preset "clearly abnormal" threshold, corresponds to severe abnormal characteristics that can be directly judged as faults regardless of environmental changes (such as a sudden drop in current of more than 50%, far exceeding the fluctuation range that the environment may cause). Even if n / a is extremely high (with a great influence from the environment), SZ will at most approach EZ and will not exceed it, ensuring that when the initial abnormality value of the board under test reaches EZ or above, it will inevitably be judged as abnormal, avoiding the omission of true severe faults due to over-considering environmental factors.

[0036] S500: If the initial anomaly level value is greater than the third threshold, the solar panel under test is determined to be abnormal.

[0037] Specifically, if n / a exceeds the ratio threshold, it indicates that most panels in the area are affected by the same environment (high feature matching degree). If the initial abnormal value of the test panel is higher than the third threshold, it means that its deviation is likely not caused by the environment, but by its own fault (such as component aging or partial shading).

[0038] When the initial anomaly level exceeds the third threshold (SZ), it indicates that even in scenarios with "significant environmental impact" (n / a > preset proportional threshold), the abnormal characteristics of the test panel still exceed the dynamically adjusted judgment threshold. In this case, the dominant role of "environmental interference" can be ruled out—because if the fluctuation is caused by the environment, its anomaly level should be consistent with that of key solar panels in the same area (i.e., "intercepted" within the normal range by SZ), while the portion exceeding SZ is more likely to originate from the test panel's own faults (such as component microcracks, loose wiring, etc.). For example, if the initial anomaly level of the test panel is 0.85, and the third threshold SZ = 0.82 (calculated based on n / a = 60%), then the result of 0.85 > 0.82 indicates that its anomaly level has exceeded the "range that can be explained by environmental interference," and therefore it is judged as abnormal.

[0039] In one exemplary embodiment of this application, after step S200, the method further includes:

[0040] S201, if the initial anomaly level value is greater than the second threshold, then the solar panel under test is determined to be abnormal.

[0041] Specifically, if the initial anomaly level value is greater than the second threshold, which is the critical value for clearly determining an anomaly, the solar panel under test is directly determined to be abnormal.

[0042] In one exemplary embodiment of this application, after step S300, the method further includes:

[0043] S301, if n / a is less than or equal to the preset threshold for the proportion of the same area, then the solar panel under test is determined to be abnormal.

[0044] Specifically, the smaller the n / a ratio, the stronger the consistency of the influence of the same external environment (such as sunlight and temperature) on a small number of solar panels in the same area—that is, the fluctuations in the current characteristic DA and voltage characteristic DU of the tested panel are more likely to be due to a fault in the tested solar panel itself. In this case, the tested solar panel is directly determined to be abnormal.

[0045] In one exemplary embodiment of this application, after step S300, the method further includes:

[0046] S600, if the initial anomaly level value is equal to or less than the third threshold, then obtain the first intermediate solar panels included within the preset range where the area of ​​the solar panel to be tested is located, so as to obtain the first intermediate solar panel list YB = (YB1, YB2, ..., YB...). j , ..., YB m ); j = 1, 2, ..., m; where m is the number of the first intermediate solar panels; YB jThe j-th first intermediate solar panel; the matching degree between the external environmental characteristics of the first intermediate solar panel on the target time window day and the external environmental characteristics of the solar panel under test on the target time window day is greater than the preset external environmental matching degree threshold.

[0047] Specifically, the preset range can be an area with a larger geographical radius centered on the location of the solar panel under test. Within this range, there may be similar macro-climate influences (such as different photovoltaic power stations in the same city), but the micro-environment (such as local shading) may differ. External environmental characteristics include weather type characteristics, temperature and humidity characteristics, and light intensity characteristics. The feature extraction and encoding methods for the external environmental characteristics of the first intermediate solar panel on the target time window and the external environmental characteristics of the solar panel under test on the target time window are as follows:

[0048] Furthermore, the weather type feature extraction method is as follows: Weather type data for the target time window is obtained from a photovoltaic system-integrated weather station (e.g., Vaisala WXT530) or a third-party meteorological API (e.g., China Weather Network open interface). The dominant weather type (e.g., "sunny" or "cloudy") is extracted on a daily basis. If there is a weather transition on the same day (e.g., sunny in the morning and rainy in the afternoon), the dominant weather type (≥60%) within the target time window (e.g., 9:50-10:00) is selected. One core feature is extracted—the dominant weather type within the target time window (categorical variable). Temperature and humidity features are extracted as follows: Real-time data is collected using temperature and humidity sensors (e.g., SHT30) installed on a solar panel array, with a sampling frequency of 1 minute / time. For the target time window (e.g., 9:40-10:00), the following statistics are calculated: Temperature features: average temperature (°C) within the window, temperature fluctuation range (maximum - minimum, °C); Humidity features: average humidity (%) within the window, humidity fluctuation range (maximum - minimum, %). Feature dimensions: A total of 4 features (mean temperature, temperature fluctuation, mean humidity, and humidity fluctuation). Light intensity feature extraction method: Data is collected using an irradiance sensor of the photovoltaic system (e.g., Kipp & Zonn CMP11), with a sampling frequency of 1 minute / sample. For the target time window, the following statistics are calculated: average light intensity (W / m²). 2 Peak illumination (W / m) 2The duration of stable illumination (in seconds, referring to the continuous duration of illumination intensity fluctuation ≤ 5%) is defined as follows. There are three feature dimensions: average illumination, peak illumination, and stable duration. To transform non-numerical features into a computable vector form while preserving their physical meaning, the following encoding method is used: Weather type feature encoding (categorical variable → numerical vector) employs one-hot encoding. The specific steps are: Define a standard weather type set: {sunny, cloudy, overcast, rainy, snowy, dusty} (covering over 95% of common scenarios); assign a binary vector to each type, specifically: Sunny → [1,0,0,0,0,0], Cloudy → [0,1,0,0,0,0], Overcast → [0,0,1,0,0,0], Rainy → [0,0,0,1,0,0], Snowy → [0,0,0,0,1,0], Dusty → [0,0,0,0,0,1]. This encoding method avoids the "implicit priority of numerical values" problem caused by label encoding (e.g., sunny = 1, cloudy = 2), ensuring that the differences between different weather types are only reflected through vector matching degree. The temperature and humidity feature encoding (continuous variable → normalized value) adopts min-max normalization, mapping the feature values ​​to the [0,1] interval. Temperature features: average temperature: value range [-20℃, 40℃] (covering the working environment of photovoltaic systems in most areas), such as 25℃ after normalization is (25+20) / (40+20) = 0.75; temperature fluctuation: value range [0℃, 15℃] (the maximum daily fluctuation is usually ≤15℃), such as 5℃ after normalization is 5 / 15≈0.33. Humidity characteristics: Average humidity: range [10%, 90%], e.g., 60% normalized is (60-10) / (90-10) = 0.625; Humidity fluctuation: range [0%, 30%], e.g., 10% normalized is 10 / 30 ≈ 0.33. Light intensity characteristic encoding (continuous variable → normalized value) also uses min-max normalization: Average light intensity: range [0W / m²]. 2 1000W / m 2 (Standard test conditions: STC light intensity is 1000 W / m²) 2 ), such as 800W / m 2 Normalized value is 0.8; Peak illumination: value range [0W / m²]. 2 1200W / m 2 (Extremely strong light scenarios), such as 1000W / m 2 After normalization, it is 1000 / 1200≈0.83; Stable duration: value range [0 seconds, 900 seconds] (maximum stable duration within a 15-minute window), such as 600 seconds after normalization, it is 600 / 900≈0.67.

[0049] After encoding, a comprehensive large vector is obtained, and matching is performed based on the comprehensive large vector. The specific matching method can be calculated based on cosine distance or Euclidean distance, etc.

[0050] In this embodiment, when the initial anomaly level value is ≤ SZ, it indicates that the fluctuation of the test panel may be caused by environmental interference in the same area. However, relying solely on data from the same area may have limitations (e.g., all panels in the area are affected by local anomalies). By expanding the scope to a "preset range" and introducing solar panels affected by similar macroscopic environments as a reference, it is possible to verify whether the fluctuation is caused by common environmental factors over a larger area, thus avoiding misjudgments caused by local environmental anomalies in the region.

[0051] S700, based on DA, DU, and YB, obtain the second intermediate solar panel list EB = (EB1, EB2, ..., EB...). x , ..., EB y ); x = 1, 2, ..., y; where y is the number of the second intermediate solar panels; EB x The xth intermediate solar panel is a second intermediate solar panel; the second intermediate solar panel has at least one corresponding key date; the second intermediate solar panel is a first intermediate solar panel that meets the first intermediate condition; the first intermediate condition is: the matching degree between the external environmental characteristics corresponding to the key date and the external environmental characteristics of the solar panel under test corresponding to the key date is greater than a preset external environmental matching degree threshold; and the matching degree between the power generation characteristics corresponding to the key date and the power generation characteristics of the solar panel under test corresponding to the key date is greater than a preset power matching degree threshold; the key date is any day within a preset number of days before that day.

[0052] Specifically, as an example: if the current date is July 31, 2025, and the preset number of days is 5, then the key date can be any one of July 30, 2025, July 29, 2025, July 28, 2025, July 27, 2025, and July 26, 2025.

[0053] The key date is set to "any day within a preset number of days prior to the current date" (e.g., if the current date is July 31, 2025, and the preset number of days is 5, then the key dates are July 26 to 30). This setting is based on the short-term stability of photovoltaic systems: the hardware parameters (such as open-circuit voltage, short-circuit current, and fill factor) and aging status of solar panels typically do not show significant abrupt changes within 5-30 days (attenuation rate ≤ 0.1% / day). Therefore, recent historical data can effectively reflect its inherent power generation characteristics. Choosing this time range avoids interference from long-term (e.g., more than 30 days) environmental seasonal changes (such as irradiance angle and average temperature) on feature comparisons, while ensuring the representativeness of historical references by covering environmental fluctuations at different times with a sufficient sample size (e.g., 5 days).

[0054] By calculating the environmental feature vector (including quantified values ​​of weather type, temperature, humidity, and light intensity) of key dates and the vector similarity (e.g., cosine similarity ≥ 0.8) of the corresponding date of the panel under test, it is ensured that the two sets of samples are in the same environmental baseline on a certain historical day. This is a prerequisite for eliminating environmental interference, because only under the same environment can the difference in power generation characteristics be attributed to the characteristics of the solar panel itself (rather than the difference in environmental input).

[0055] Based on consistent environmental benchmarks, the similarity of power generation characteristic vectors (statistics of current characteristic DA and voltage characteristic DU, such as mean and volatility) is used to verify the consistency of the power generation response patterns of the two boards on critical dates. For example, if on July 28 (critical date) both the board under test and a board in YB are under "sunny weather, 800W / m" sunlight conditions... 2 If the two are in an environment with a temperature of 25℃, and the average current is 5.2±0.1A and the peak voltage is 24.5±0.2V, then the output characteristics of the two are highly consistent under the same input environment.

[0056] The YB (intermediate batch) selected by S600 only guarantees a match in external environmental characteristics for that day. Further screening is conducted within YB to select a subset of solar panels whose external environmental characteristics and power generation characteristics are similar to those of the solar panel under test on a specific day within a recent time range. This is called EB (external batch). Each solar panel in EB has power generation characteristics similar to those of the solar panel under test (since the external environment affects power generation characteristics, similar power generation characteristics under the same external environment indicate that such a second intermediate solar panel is similar to the power generation characteristics of the solar panel under test after excluding environmental factors). Furthermore, the external environmental characteristics of each solar panel in EB on that day are also similar to those of the solar panel under test. This means that the samples in EB and the panel under test not only have a high degree of similarity in the current environment, but also have consistent power generation characteristics under similar historical environments. The similarity of their power generation characteristics is not affected by environmental fluctuations, but rather stems from the consistency of their intrinsic characteristics (such as materials, processes, and aging states). If most of the solar panels in the EB exhibit different power generation characteristics from the solar panel under test under the same external environment, it indicates that the abnormality of the solar panel under test is most likely due to its own inherent causes. Conversely, if most of the solar panels in the EB exhibit the same power generation characteristics as the solar panel under test under the same external environment, it indicates that the abnormality of the solar panel under test is most likely due to environmental causes.

[0057] This embodiment uses the setting of short-term historical key dates and the dual matching of environmental and power generation characteristics to extract reference samples (EB) from YB that have consistent with the inherent power generation characteristics of the solar panel under test.

[0058] S800, based on EB, obtain the third intermediate solar panel list SB = (SB1, SB2, ..., SB...). p , ..., SB q); p = 1, 2, ..., q; where q is the number of third solar panels; SB p Let p be the third solar panel; the third solar panel is a second intermediate solar panel that meets the second intermediate condition; the second intermediate condition is: the matching degree between the power generation characteristics corresponding to the target time window and the power generation characteristics of the solar panel under test in the target time window is greater than the preset matching degree threshold.

[0059] S900, if q / y is greater than the preset threshold for the proportion of different regions, then the solar panel under test is determined to be non-abnormal.

[0060] Specifically, this embodiment, based on the obtained third threshold, further examines several solar panels (EB) within a preset range that have similar power generation performance to the solar panel under test in different regions (similar power generation characteristics under the same environmental characteristics, excluding environmental influences) and similar external environmental characteristics on the day of testing. If most of these solar panels have similar power generation characteristics to the solar panel under test within the target time window, it indicates that the current current and voltage anomalies of the solar panel under test are likely due to environmental factors. That is, if q / y is greater than the preset threshold for the proportion of different regions, it indicates that the anomaly of the solar panel under test is likely due to environmental reasons. At this point, it is necessary to raise the threshold for judging anomalies to avoid misjudging environmental interference as a fault. That is, continue to use the obtained third threshold as the judgment standard. That is, determine that the solar panel under test is not abnormal. That is, further determine whether to use the third threshold as the judgment basis, improving the accuracy of the final judgment of the anomaly of the solar panel under test.

[0061] In one exemplary embodiment of this application, after S800, the method further includes:

[0062] S1000, if q / y is less than or equal to the preset non-same area ratio threshold, then the solar panel under test is determined to be abnormal.

[0063] Specifically, if q / y is less than or equal to the preset threshold for the proportion of non-same areas, it indicates that the anomaly of the solar panel under test is most likely due to its own inherent causes. Since it already exceeds the first threshold, it is directly identified as an anomaly to avoid missed detections.

[0064] In an exemplary embodiment of this application, an electronic device capable of implementing the above-described method is also provided.

[0065] Those skilled in the art will understand that various aspects of this application can be implemented as a system, method, or program product. Therefore, various aspects of this application can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, collectively referred to herein as a "circuit," "module," or "system."

[0066] An electronic device according to this embodiment of the present application. The electronic device is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of this application.

[0067] Electronic devices are manifested in the form of general-purpose computing devices. Components of an electronic device may include, but are not limited to: at least one processor, at least one memory, and buses connecting different system components (including memory and processor).

[0068] The memory stores program code that can be executed by a processor, causing the processor to perform the steps described in the "Exemplary Methods" section above, according to various exemplary embodiments of this application.

[0069] The storage may include readable media in the form of volatile storage, such as random access memory (RAM) and / or cache memory, and may further include read-only memory (ROM).

[0070] The storage may also include programs / utilities having a set (at least one) of program modules, including but not limited to: an operating system, one or more applications, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0071] A bus can represent one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus that uses any of the various bus architectures.

[0072] The electronic device can also communicate with one or more external devices (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device, and / or any device that enables the electronic device to communicate with one or more other computing devices (e.g., routers, modems, etc.). This communication can be achieved through input / output (I / O) interfaces. Furthermore, the electronic device can communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter. As shown in the figure, the network adapter communicates with other modules of the electronic device via a bus. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0073] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the method according to the embodiments of this application.

[0074] In exemplary embodiments of this application, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible implementations, various aspects of this application may also be implemented as a program product including program code, which, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of this application described in the "Exemplary Methods" section above.

[0075] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0076] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0077] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0078] Program code for performing the operations of this application can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0079] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this application, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0080] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0081] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting anomalies in solar panels, characterized in that, The method includes: Obtain the current characteristic DA and voltage characteristic DU of the solar panel under test within the target time window; where the end time of the target time window is the current time. Based on DA, DU and the anomaly detection model, the initial anomaly level value corresponding to the solar panel under test is obtained; wherein, the anomaly detection model has a first threshold and a second threshold; the second threshold is greater than the first threshold; the initial anomaly level value is the normalized anomaly value of 0-1 output by the anomaly detection model after inputting the 4-dimensional feature vector composed of DA and DU into the anomaly detection model; the higher the normalized anomaly value, the greater the probability of an anomaly. If the initial anomaly level value is greater than the first threshold and less than the second threshold, then based on DA and DU, the list of critical solar panels G=(G1, G2, ..., G...) is obtained. i , ..., G n ); i = 1, 2, ..., n; where n is the number of critical solar panels; G i Let be the i-th key solar panel; the key solar panel is located within the area where the solar panel under test is located, and the matching degree between the corresponding key power generation characteristics and the power generation characteristics of the solar panel under test is greater than the preset matching degree threshold; the key power generation characteristics include the current characteristics and voltage characteristics of the corresponding key solar panel within the target time window; the power generation characteristics of the solar panel under test include DA and DU; If n / a is greater than the preset threshold for the proportion of the same area, then a third threshold is obtained; wherein, the third threshold is greater than the first threshold, and the third threshold is less than or equal to the second threshold; the third threshold is determined based on the first threshold and n / a; a is the number of solar panels contained in the area where the solar panel to be tested is located; the third threshold SZ meets the following characteristics: SZ = YZ + (EZ - YZ) × n / a; where YZ is the first threshold and EZ is the second threshold. If the initial anomaly level value is greater than the third threshold, the solar panel under test is determined to be abnormal. If the initial anomaly level value is equal to or less than the third threshold, then the first intermediate solar panels included within the preset range of the area where the solar panel under test is located are obtained, so as to obtain the first intermediate solar panel list YB=(YB1, YB2, ..., YB j , ..., YB m ); j=1,2,…,m; where m is the number of the first intermediate solar panels; YB j For the j-th first intermediate solar panel; the matching degree between the external environmental characteristics of the first intermediate solar panel on the target time window day and the external environmental characteristics of the solar panel under test on the target time window day is greater than the preset external environmental matching degree threshold. Based on DA, DU, and YB, the second intermediate solar panel list EB = (EB1, EB2, ..., EB) is obtained. x , ..., EB y ); x = 1, 2, ..., y; where y is the number of the second intermediate solar panels; EB x The x-th second intermediate solar panel; the second intermediate solar panel has at least one corresponding key date; the second intermediate solar panel is a first intermediate solar panel that meets the first intermediate condition; the first intermediate condition is: the matching degree between the external environmental characteristics corresponding to the key date and the external environmental characteristics of the solar panel under test corresponding to the key date is greater than a preset external environmental matching degree threshold; and the matching degree between the power generation characteristics corresponding to the key date and the power generation characteristics of the solar panel under test corresponding to the key date is greater than a preset power matching degree threshold; the key date is any day within a preset number of days before that day; Based on EB, the third intermediate solar panel list SB = (SB1, SB2, ..., SB) is obtained. p , ..., SB q ); p = 1, 2, ..., q; where q is the number of third solar panels; SB p For the p-th third solar panel; the third solar panel is the second intermediate solar panel that meets the second intermediate condition; the second intermediate condition is: the matching degree between the power generation characteristics corresponding to the target time window and the power generation characteristics of the solar panel under test in the target time window is greater than the preset matching degree threshold. If q / y is greater than the preset threshold for the proportion of different regions, then the solar panel under test is determined to be non-abnormal.

2. The method for detecting anomalies in solar panels according to claim 1, characterized in that, Based on EB, the third intermediate solar panel list SB = (SB1, SB2, ..., SB) is obtained. p , ..., SB q Following this, the method further includes: If q / y is less than or equal to the preset threshold for the proportion of different regions, then the solar panel under test is determined to be abnormal.

3. The method for detecting anomalies in solar panels according to claim 1, characterized in that, After obtaining the initial anomaly level value corresponding to the solar panel under test based on DA, DU, and the anomaly detection model, the method further includes: If the initial anomaly level value is greater than the second threshold, the solar panel under test is determined to be abnormal.

4. The method for detecting anomalies in solar panels according to claim 1, characterized in that, If the initial anomaly level is greater than the first threshold and less than the second threshold, then based on DA and DU, the list of critical solar panels G = (G1, G2, ..., G...) is obtained. i , ..., G n Following this, the method further includes: If n / a is less than or equal to the preset threshold for the proportion of the same area, then the solar panel under test is determined to be abnormal.

5. The method for detecting anomalies in solar panels according to claim 1, characterized in that, External environmental characteristics include weather type characteristics, temperature and humidity characteristics, and light intensity characteristics.

6. A non-transitory computer-readable storage medium, characterized in that, The storage medium stores at least one instruction or at least one program segment, which is loaded and executed by a processor to implement the solar panel anomaly detection method as described in any one of claims 1-5.

7. An electronic device, characterized in that, Includes a processor and the non-transitory computer-readable storage medium as described in claim 6.

Images