Abnormality early warning method and device of distributed photovoltaic power generation system and electronic equipment

Abstract

The application discloses an abnormal early warning method and device of a distributed photovoltaic power generation system and electronic equipment. The method comprises the following steps: obtaining current performance data, current environmental interference data and performance index data of a target distributed photovoltaic power generation system at a current time, and then determining a current initial performance score of the target distributed photovoltaic power generation system at the current time; based on the current initial performance score, detecting whether the performance of the target distributed photovoltaic power generation system has a target abnormal fluctuation at the current time; if the target abnormal fluctuation exists, determining an abnormal reason, and correcting the current initial performance score based on the abnormal reason to obtain a current target performance score of the target distributed photovoltaic power generation system at the current time; and based on the current target performance score, performing abnormal early warning on the target distributed photovoltaic power generation system. The application solves the technical problem of inaccurate abnormal early warning of the distributed photovoltaic power generation system in the related art.

Description

Technical Field

[0001] This application relates to the field of power systems, and more specifically, to an anomaly early warning method, device, and electronic equipment for a distributed photovoltaic power generation system. Background Technology

[0002] In the operation and maintenance management of distributed photovoltaic (PV) power generation systems, to ensure the power generation efficiency, it is necessary to detect and provide early warnings of abnormal performance degradation. Related technologies typically employ methods based on a single power generation threshold or a fixed performance ratio to provide early warnings of abnormal performance degradation. However, this method cannot effectively distinguish between performance degradation caused by environmental factors and performance degradation caused by the equipment itself within the PV system. This often leads to numerous false alarms in practical applications. For example, performance degradation caused by environmental factors may be misjudged as performance degradation caused by the equipment itself, or the system may fail to promptly identify performance degradation truly caused by the equipment itself, resulting in insufficient accuracy in early warnings and impacting operation and maintenance decisions. Therefore, related technologies suffer from the technical problem of inaccurate early warning for distributed PV power generation systems.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This application provides an anomaly early warning method, device, and electronic device for distributed photovoltaic power generation systems, so as to at least solve the technical problem of inaccurate anomaly early warning for distributed photovoltaic power generation systems in related technologies.

[0005] According to one aspect of the embodiments of this application, an anomaly early warning method for a distributed photovoltaic (PV) power generation system is provided, comprising: acquiring current performance data, current environmental interference data, and performance index data of a target distributed PV power generation system at the current moment, wherein the current environmental interference data refers to environmental data that affects the performance of the target distributed PV power generation system, and the performance index data is used to describe the change in the performance of the target distributed PV power generation system over time; determining the current initial performance score of the target distributed PV power generation system at the current moment based on the current performance data, current environmental interference data, and performance index data; detecting whether there is a target abnormal fluctuation in the performance of the target distributed PV power generation system at the current moment based on the current initial performance score; determining the cause of the abnormality when the performance of the target distributed PV power generation system is detected to be abnormal, and correcting the current initial performance score based on the cause of the abnormality to obtain the current target performance score of the target distributed PV power generation system at the current moment; and issuing an anomaly early warning for the target distributed PV power generation system based on the current target performance score.

[0006] According to another aspect of the embodiments of this application, an anomaly early warning device for a distributed photovoltaic power generation system is provided, comprising: a data acquisition module, configured to acquire current performance data, current environmental interference data, and performance index data of a target distributed photovoltaic power generation system at the current moment, wherein the current environmental interference data refers to environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change in the performance of the target distributed photovoltaic power generation system over time; a first determination module, configured to determine the current initial performance score of the target distributed photovoltaic power generation system at the current moment based on the current performance data, current environmental interference data, and performance index data; a detection module, configured to detect whether there is a target abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment based on the current initial performance score; a second determination module, configured to determine the cause of the abnormality when the abnormal fluctuation in the performance of the target distributed photovoltaic power generation system is detected, and to correct the current initial performance score based on the cause of the abnormality to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment; and an anomaly early warning module, configured to provide an anomaly early warning for the target distributed photovoltaic power generation system based on the current target performance score.

[0007] According to another aspect of the embodiments of this application, a non-volatile storage medium is provided, which stores multiple instructions, the instructions being adapted for a distributed photovoltaic power generation system anomaly early warning method that can be loaded by a processor and executed at any one of them.

[0008] According to another aspect of the embodiments of this application, an electronic device is provided, including: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any one of the following abnormal early warning methods for a distributed photovoltaic power generation system.

[0009] According to another aspect of the embodiments of this application, a computer program product is provided, which, when executed on a data processing device, is suitable for performing the steps of an anomaly early warning method for a distributed photovoltaic power generation system.

[0010] In this embodiment, the current performance data, current environmental interference data, and performance index data of the target distributed photovoltaic (PV) power generation system at the current moment are obtained. The current environmental interference data refers to environmental data that affects the performance of the target PV power generation system, and the performance index data describes the change in the performance of the target PV power generation system over time. Based on the current performance data, current environmental interference data, and performance index data, the initial performance score of the target PV power generation system at the current moment is determined. Based on the initial performance score, it is detected whether there are any abnormal fluctuations in the performance of the target PV power generation system at the current moment. If abnormal fluctuations are detected, the cause of the abnormality is determined, and the initial performance score is corrected based on the cause of the abnormality to obtain the current target performance score of the target PV power generation system at the current moment. Based on the current target performance score, an anomaly warning is issued for the target PV power generation system. The goal is to determine the current target performance score of the target distributed photovoltaic power generation system by analyzing current performance data, current environmental interference data, and performance index data, and to provide anomaly warnings for the target distributed photovoltaic power generation system based on the current target performance score. This achieves the technical effect of improving the accuracy of anomaly warnings for the target distributed photovoltaic power generation system, thereby solving the technical problem of inaccurate anomaly warnings for distributed photovoltaic power generation systems in related technologies. Attached Figure Description

[0011] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0012] Figure 1 This is a flowchart of an anomaly early warning method for a distributed photovoltaic power generation system provided according to an embodiment of this application;

[0013] Figure 2 This is a flowchart of an optional abnormal early warning method for a distributed photovoltaic power generation system provided according to an embodiment of this application;

[0014] Figure 3 This is a schematic diagram of an optional abnormal early warning device for a distributed photovoltaic power generation system provided according to an embodiment of this application. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0016] 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 apparatus 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 apparatus.

[0017] It should be noted that the information and data collected in this application (including but not limited to current performance data, current environmental interference data, performance index data, first historical initial performance score data, second historical initial performance score data, historical target performance score data, and forecast meteorological data, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding operation entry points for them to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0018] According to an embodiment of this application, an embodiment of an anomaly early warning method for a distributed photovoltaic power generation system is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0019] Figure 1 This is a flowchart of an anomaly early warning method for a distributed photovoltaic power generation system according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps:

[0020] Step S102: Obtain the current performance data, current environmental disturbance data, and performance index data of the target distributed photovoltaic power generation system at the current moment. The current environmental disturbance data refers to the environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change of the performance of the target distributed photovoltaic power generation system over time.

[0021] It is understood that current performance data may include, but is not limited to, the actual power generation of photovoltaic units in the target distributed photovoltaic power generation system, module temperature, string current, string voltage, inverter efficiency, and cumulative irradiance; current environmental interference data may include, but is not limited to, the surface area transmittance of modules, the temperature of local hot spots on modules, the grid supply voltage, and the ambient wind speed; performance index data may include, but is not limited to, the rate of performance change, the rate of performance degradation, and the year-on-year performance deviation. The target distributed photovoltaic power generation system includes multiple photovoltaic units, each of which includes several modules. These modules can be connected in series to form strings. A module is the smallest unit in the target distributed photovoltaic system capable of independently generating power, and a photovoltaic unit is the smallest unit in the target distributed photovoltaic system capable of independently performing performance monitoring. The inverter connects to the strings and converts the DC power generated by the strings into AC power. By acquiring the above-mentioned multi-dimensional performance perception data related to performance, a rich and comprehensive data foundation is provided for the early warning of anomalies in the target distributed photovoltaic power generation system.

[0022] Optionally, multi-dimensional performance sensing data of each photovoltaic unit in the distributed photovoltaic system can be acquired, including basic performance data (i.e., current performance data), current environmental interference data, and time-series dynamic data (i.e., performance index data). Basic performance data includes actual power generation, module temperature, string current, string voltage, inverter efficiency, and cumulative irradiance; environmental interference data includes module surface area transmittance, local hot spot temperature of the module, grid supply voltage, and ambient wind speed; time-series dynamic data includes the performance change rate within the first time period (i.e., the first historical time period), the performance degradation rate within the second time period (i.e., the second historical time period), and the year-on-year performance deviation within the third time period. A photovoltaic unit refers to the smallest unit in the target distributed photovoltaic power generation system capable of independently detecting power generation performance. It can be a string or a subarray composed of multiple strings, the specific division depending on the deployment granularity of the on-site detection equipment.

[0023] Basic performance data is a set of core parameters that directly reflect the conversion of solar energy into electrical energy by the target distributed photovoltaic power generation system. It is used to calculate the deviation between theoretical power generation and actual power generation. Specifically, it includes actual power generation, module temperature, string current, string voltage, inverter efficiency, and cumulative irradiance.

[0024] Actual power generation refers to the total amount of electrical energy actually delivered to the grid or load by the photovoltaic units in a target distributed photovoltaic power generation system within a specific time interval. The unit is usually kilowatt-hours (kWh), which can be obtained through the metering function of smart meters or inverters. Actual power generation is the final efficiency output of the photovoltaic units, directly reflecting the actual effect of converting input irradiance into electrical energy, and is the most intuitive indicator for judging whether efficiency has degraded. If irradiance is sufficient but power generation remains consistently low, it directly indicates abnormal degradation. Ignoring power generation may miss hidden anomalies such as normal parameters but low overall conversion efficiency.

[0025] Module temperature refers to the actual temperature of the photovoltaic panels in a target distributed photovoltaic power generation system during operation. It can be obtained through sensors installed on the module backsheet or using infrared temperature measurement technology. The photoelectric conversion efficiency of the module is negatively correlated with temperature. Excessively high module temperatures not only directly lead to a reduction in immediate efficiency, but also accelerate the aging of the module encapsulation layer over a long period, causing irreversible long-term degradation. Ignoring module temperature may lead to misjudging the performance degradation caused by high temperatures as module failure, or overlooking the potential aging risks caused by excessively high temperatures.

[0026] String current and string voltage refer to the current flowing through the entire string and the voltage across its terminals, respectively, and can be acquired using string-level current and voltage sensors. An abnormally low current compared to other parallel strings or historical data from the same period directly indicates a problem in that string, such as obstruction, wiring faults, or partial component failure. Abnormal voltage may indicate series problems, such as reverse bias caused by hot spots on components, diode conduction in the junction box, or abnormal maximum power point tracking. Combining both can be used to plot current-voltage curves and pinpoint the problematic string.

[0027] Inverter efficiency refers to the efficiency with which an inverter converts direct current (DC) to alternating current (AC), i.e., the ratio of AC output power to DC input power. This efficiency is provided by the inverter's own monitoring system. This data reflects whether the power generated by the modules can be effectively transmitted to the grid, helping to distinguish whether the problem originates from the DC or AC side. Ignoring inverter efficiency can lead to overlooking the cause of degradation in cases where the modules are functioning normally but the conversion process is malfunctioning.

[0028] Cumulative irradiance refers to the cumulative amount of solar radiation energy that falls on the plane containing a photovoltaic unit within a specific time interval. The unit is typically kilowatt-hours per square meter (kWh / m²), and it can be measured using an on-site irradiance meter. Cumulative irradiance serves as a benchmark for theoretical power generation. Ignoring low irradiance while focusing solely on low power generation can lead to a misjudgment of insufficient normal input as abnormal degradation. Conversely, if irradiance is sufficient but power generation is low, the anomaly is clearly caused by either the equipment itself within the target distributed photovoltaic power generation system or environmental interference.

[0029] Environmental disturbance data refers to external factors that affect the performance of photovoltaic units in a target distributed photovoltaic power generation system. It is used to identify and quantify the power generation loss caused by temporary and recoverable external factors, including the surface area transmittance of the module, the temperature of local hot spots on the module, the grid supply voltage, and the ambient wind speed.

[0030] Module surface transmittance refers to the degree to which the proportion of sunlight transmitted to the solar cells decreases due to contaminants such as dust, dirt, and bird droppings covering the module surface. It can be estimated indirectly through specialized optical sensors or by analyzing the differences in output characteristics between clean and contaminated states. Contaminant coverage blocks light, directly reducing the module's short-circuit current and thus lowering output power. This data allows for precise quantification of power generation loss caused by contamination. When a decrease in power generation is detected, if a significant decrease in transmittance is also observed, the primary cause can be identified as contaminant coverage.

[0031] Localized hotspot temperature refers to the abnormal temperature of a localized area on a module that is significantly higher than the module's average temperature due to factors such as cell defects, internal bypass diode failure, or severe shading. This temperature can be obtained through regular inspections using infrared thermal imagers mounted on drones or by detection with fixed thermal imagers. Hotspots not only cause power loss in the corresponding area but also accelerate module aging and can even lead to fires. This data can be used to directly locate defective modules.

[0032] The grid supply voltage refers to the grid voltage at the point of connection of the target distributed photovoltaic (PV) power generation system. This data comes from grid-side monitoring data or inverter AC-side monitoring data. When the grid voltage is too high or fluctuations exceed the inverter's allowable range, the inverter will proactively reduce its output power or disconnect from the grid in accordance with grid connection standards to protect grid safety. In this case, the decrease in power generation is a normal protective behavior of the inverter, not a fault. By monitoring this data, it is possible to avoid misreporting reduced efficiency caused by grid problems as a fault in the target distributed PV power generation system.

[0033] Ambient wind speed refers to the wind speed at the location of a photovoltaic array (composed of multiple strings), which can be measured by an anemometer. Wind speed affects the heat dissipation of the modules, thus affecting the module temperature. At the same time, strong winds may blow debris from the ground, cause trees to sway, create dynamic shadows on the modules, and cause rapid fluctuations in power generation. Recording wind speed helps to explain some short-term, drastic power fluctuations.

[0034] Time-series dynamic data reflects the trend of efficiency changes of a target distributed photovoltaic power generation system over time. It can reveal the changing trend and periodicity of efficiency from a time dimension, specifically including the efficiency change rate in the first time period, the efficiency decay rate in the second time period, and the year-on-year efficiency deviation in the third time period.

[0035] The efficiency change rate within the first time period corresponds to a short-term time scale, such as the most recent 24 hours or the most recent 7 days. The efficiency change rate can be the rate of change in average daily power generation, or the short-term volatility of the performance index. This indicator is used to capture sudden declines in efficiency.

[0036] The efficiency degradation rate within the second time period corresponds to a medium-term time scale, such as the most recent 3 months or the most recent 6 months. The efficiency degradation rate is used to quantify the linear or non-linear decline trend of efficiency. It can be obtained by performing a linear regression on the historical baseline efficiency data of the second historical time period and taking the normalized value of its slope as the efficiency degradation rate. This indicator is used to identify slow component aging or progressive contamination.

[0037] The third timeframe for performance year-over-year deviation corresponds to a long-term time scale, typically an annual comparison. Year-over-year deviation refers to the deviation rate between the current performance and the performance at the same historical time last year, used to eliminate the influence of seasonal factors. Solar irradiance, temperature, and sunshine duration all exhibit strong seasonal patterns; year-over-year comparisons, after eliminating seasonal factors, most directly reflect the long-term changes in the performance of the target distributed photovoltaic power generation system.

[0038] Step S104: Based on the current performance data, current environmental disturbance data, and performance index data, determine the current initial performance score of the target distributed photovoltaic power generation system at the current moment;

[0039] It is understandable that by integrating three different types of data—current performance data, current environmental disturbance data, and performance indicator data—the comprehensiveness and accuracy of the initial performance score can be ensured, and the errors that may arise from relying solely on one type of data can be reduced.

[0040] In one optional embodiment, the initial performance score of the target distributed photovoltaic (PV) power generation system at the current moment is determined based on current performance data, current environmental disturbance data, and performance index data. This includes: determining a basic performance score based on current performance data; determining an environmental disturbance coefficient based on current environmental disturbance data, wherein the environmental disturbance coefficient is used to quantify the impact of current environmental disturbance data on the performance of the target distributed PV power generation system; obtaining a performance index score by weighted calculation based on the performance change rate, performance decay rate, and performance year-on-year deviation in the performance index data, wherein the performance change rate is used to quantify the degree of change in the performance of the target distributed PV power generation system at the current moment compared to a first historical time period, the performance decay rate is used to quantify the degree of decay in the performance of the target distributed PV power generation system at the current moment compared to a second historical time period, and the performance year-on-year deviation is used to quantify the degree of difference in the performance of the target distributed PV power generation system at the current moment compared to the historical same period; and determining the initial performance score based on the basic performance score, environmental disturbance coefficient, and performance index score.

[0041] The process can be understood as follows: First, a basic performance score is determined based on current performance data. Second, an environmental interference coefficient is determined based on current environmental interference data. Then, based on the performance change rate, performance decay rate, and year-on-year performance deviation in the performance indicator data, combined with the dynamic weights of the performance change rate, performance decay rate, and year-on-year performance deviation, a weighted calculation method is used to obtain the performance indicator score. Finally, based on the aforementioned basic performance score, environmental interference coefficient, and performance indicator score, the current initial performance score of the target distributed photovoltaic power generation system at the current moment is calculated. Specifically, the performance change rate quantifies the degree of change in the performance of the target distributed photovoltaic power generation system at the current moment compared to the first historical time period (e.g., the most recent 24 hours or the most recent 7 days); the performance decay rate quantifies the degree of decay in the performance of the target distributed photovoltaic power generation system at the current moment compared to the second historical time period (e.g., the most recent 3 months or the most recent half year); and the year-on-year performance deviation quantifies the degree of difference in the performance of the target distributed photovoltaic power generation system at the current moment compared to the same historical time period (e.g., the same historical time in the same month and on the same date last year). By introducing the environmental disturbance coefficient, the impact of environmental factors on the performance of the target distributed photovoltaic power generation system can be quantified, distinguishing between performance fluctuations caused by environmental changes and anomalies in the target distributed photovoltaic power generation system itself. At the same time, by combining the performance change rate, performance decay rate, and performance year-on-year deviation, the performance of the target distributed photovoltaic power generation system can be analyzed from three dimensions: short-term, medium-term, and long-term, enhancing the sensitivity to gradual performance decay and sudden performance changes.

[0042] Optionally, since the acquired multi-dimensional performance perception data are independent and cannot be directly used to determine whether the performance is abnormal, it is necessary to transform the scattered data into a single, quantifiable evaluation indicator (i.e., the current initial performance score) through fusion calculation.

[0043] Optionally, the basic performance data can be normalized to obtain a basic performance score. Since the units and magnitudes of the basic performance data vary significantly, direct fusion could lead to large-scale data overly dominating the evaluation results. Therefore, normalization is necessary to convert all basic performance indicators into normalized scores within the range of 0-100. Linear normalization can be used, with reasonable optimal and worst values ​​set for the characteristics of different indicators. In the basic performance data, actual power generation, string current, string voltage, inverter efficiency, and cumulative irradiance are positive indicators; for positive indicators, a larger value indicates better performance. Module temperature is a negative indicator; for negative indicators, a larger value indicates worse performance. Based on the above analysis, the basic performance data can be normalized in the following way:

[0044]

[0045]

[0046] in, For the normalized score of the i-th indicator, the first formula is used to calculate the normalized score of the positive indicator, and the second formula is used to calculate the normalized score of the negative indicator. This is the actual measured value of the indicator; The optimal value of the i-th indicator is the historical maximum value or design rated value within a predetermined historical period when it is a positive indicator, and the historical minimum value or safety threshold within a predetermined historical period when it is a negative indicator. The worst value of the i-th indicator is the historical minimum or fault threshold in the predetermined historical time period when it is a positive indicator, and the historical maximum or danger threshold in the predetermined historical time period when it is a negative indicator.

[0047] After obtaining the normalized scores of each basic performance data point, a weighted average of these scores is taken to obtain the comprehensive basic performance score. The basic performance score can be determined as follows:

[0048]

[0049] in, Basic performance rating; The weight of the i-th indicator can be allocated according to the importance of each indicator to the overall effectiveness.

[0050] Optionally, environmental interference data can be categorized into positive and negative categories. Based on historical data, the direction and extent of the impact of each environmental interference factor on performance can be statistically analyzed, and the environmental interference coefficient can be calculated. Environmental factors have both enhancing and inhibiting effects on the performance of the target distributed photovoltaic power generation system. If all environmental data are directly superimposed, the direction of influence cannot be distinguished, leading to evaluation bias. Therefore, it is necessary to classify them before calculation. The specific classifications of positive and negative environmental factors are as follows:

[0051] Positive environmental factors: Ambient wind speed; when the ambient wind speed is moderate, heat dissipation is accelerated and efficiency is improved.

[0052] Negative environmental factors include: light transmittance of dust on the module surface, temperature of local hot spots on the module, and grid supply voltage. Among these, low light transmittance has the best suppression effect, high temperature has the best suppression effect, and voltage exceeding the limit has the best suppression effect.

[0053] Optionally, the environmental interference coefficient can be based on historical environmental interference data. A multiple linear regression model is used to establish the mapping relationship between each environmental factor and the performance deviation rate, obtaining the influence coefficient of each environmental factor. This coefficient is then calculated based on current environmental interference data. The performance deviation rate is the difference between the actual baseline performance score and the ideal performance score, representing the proportion of the ideal performance score. The ideal performance score refers to the baseline performance score under standard environmental conditions: wind speed of 2 m / s, light transmittance of 100%, no hot spots, and grid voltage of 220 V. It can be determined by the average of historical baseline performance scores that meet the standard environmental conditions.

[0054] The process of constructing a multiple linear regression model is as follows:

[0055] First, select the ambient wind speed over a past period of time. Component surface area gray transmittance Component local hot spot temperature Grid voltage Given n sets of data and their corresponding historical baseline performance scores, calculate the performance deviation rate for each set of data. :

[0056]

[0057] in, Rate the ideal performance. Assess the actual basic performance rating; A positive value indicates that the performance is lower than expected. A negative value indicates that the performance is higher than expected.

[0058] Then, based on ambient wind speed Component surface area gray transmittance Component local hot spot temperature Grid voltage As the independent variable, Using [variable name] as the dependent variable, construct a multiple linear regression model:

[0059]

[0060] Where a, b, c, and d are regression coefficients, which can be obtained by fitting the above n sets of data using the least squares method; e is a constant term, which can be set to 20, representing the baseline performance deviation rate when there is no environmental data.

[0061] Based on the above multiple linear regression model, the environmental disturbance coefficient The calculation formula is:

[0062]

[0063] The negative sign here serves to convert the performance deviation rate into an environmental interference coefficient, that is, when... When the value is positive, the efficiency is lower than expected, and environmental factors cause negative interference. When ΔE is negative, the efficiency is higher than expected, and environmental factors cause positive interference. It is positive.

[0064] Optionally, trend-weighted analysis is applied to the time-series dynamic data, assigning dynamic weights based on the time sensitivity of the efficiency change rate within the first time period, the efficiency degradation rate within the second time period, and the year-on-year efficiency deviation within the third time period. Time-series dynamic data reflects efficiency changes at different time scales, and these changes have varying early warning value for abnormal degradation of the target distributed photovoltaic power generation system. For example, short-term sharp fluctuations may indicate a sudden failure and require close monitoring, while long-term slow changes may indicate normal aging and warrant less attention. Therefore, different dynamic weights need to be assigned to different efficiency indicators based on their time sensitivity. For short-term efficiency change rates… It is highly sensitive to sudden failures, and the dynamic weight α should be set high to enable rapid response to drastic changes; for the medium-term performance decay rate... It reflects trend changes, with a moderately set dynamic weight β, used to capture gradual problems; for long-term performance year-on-year deviations... It is used for annual health comparisons. The dynamic weight γ is usually low, but it can provide validation from a long-term perspective. The specific value of the dynamic weight can be set by expert experience or determined by training a simple model using historical data, ensuring that α+β+γ=1 or meet other normalization criteria.

[0065] Optionally, the basic performance score, environmental disturbance coefficient, and time-series dynamic weights are calculated using a preset fusion formula to obtain an initial performance evaluation value. The preset fusion formula is:

[0066]

[0067] in, This is the initial performance evaluation value. The closer the value is to 1, the more ideal the performance; the lower the value, the worse the performance. Basic performance rating; Environmental interference coefficient; This constitutes the static-environmental correction term for effectiveness, meaning that if the current environment is unfavorable, the initial effectiveness assessment value will be reduced, and if the environment is favorable, it will be increased. , , These are the efficiency change rate in the first time period, the efficiency decay rate in the second time period, and the year-on-year performance deviation in the third time period, respectively. , , These are the dynamic weights for the rate of change in efficiency, the rate of decline in efficiency, and the year-on-year deviation in efficiency, respectively. It is a dynamic trend amplification term of efficiency, when , , When the value is negative, the value within the parentheses will be less than 1, thus further attenuating the static-environment correction term, amplifying the severity of the performance decline, and making the initial performance assessment value sensitive to negative trends.

[0068] Step S106: Based on the current initial performance score, detect whether there is any abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment;

[0069] It is understandable that fluctuation verification is performed based on the current initial performance score to detect whether there are any abnormal fluctuations in the performance of the target distributed photovoltaic power generation system at the current moment. By performing fluctuation verification, once abnormal fluctuations are detected, data source analysis can be performed in a timely manner to determine the cause of the anomaly, and the current initial performance score can be corrected based on the cause of the anomaly to obtain the current target performance score, thereby improving the accuracy of anomaly warning.

[0070] In one optional embodiment, based on the current initial performance score, detecting whether there is a target abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment includes: determining a first fluctuation detection result based on the first historical initial performance score data of the target distributed photovoltaic power generation system in a third historical time period and the current initial performance score, wherein the first fluctuation detection result is used to indicate whether there is a short-term or medium-term fluctuation in the performance of the target distributed photovoltaic power generation system; determining a second fluctuation detection result based on the second historical initial performance score data of the target distributed photovoltaic power generation system in a fourth historical time period and the current initial performance score, wherein the second fluctuation detection result is used to indicate whether there is a long-term fluctuation in the performance of the target distributed photovoltaic power generation system, and the fourth historical time period is longer than the third historical time period; and determining whether there is a target abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment based on the first fluctuation detection result and the second fluctuation detection result.

[0071] It is understood that, based on the initial historical performance score data of the target distributed photovoltaic (PV) power generation system in the third historical time period (e.g., the most recent 30 days) and the current initial performance score, a first fluctuation detection result is determined to indicate whether the performance of the target distributed PV power generation system experiences short-term or medium-term fluctuations. Based on the initial historical performance score data of the target distributed PV power generation system in the fourth historical time period (e.g., the most recent 180 days) and the current initial performance score, a second fluctuation detection result is determined to indicate whether the performance of the target distributed PV power generation system experiences long-term fluctuations. Based on the first and second fluctuation detection results, it is determined whether the performance of the target distributed PV power generation system exhibits any abnormal fluctuations at the current moment. By performing fluctuation detection on both short-term / medium-term time scales (the third historical time period) and long-term time scales (the fourth historical time period), comprehensive performance changes from sudden situations to long-term trends can be identified, improving the comprehensiveness and accuracy of anomaly warnings.

[0072] Optionally, the aforementioned abnormal fluctuations refer to abnormal fluctuations in the target distributed photovoltaic power generation system that exceed the normal fluctuation range. These abnormal fluctuations are not caused by external events (including meteorological and grid events), but rather by equipment malfunctions or abnormalities within the target distributed photovoltaic power generation system itself, component aging, etc. Fluctuations include short-term, medium-term, and long-term fluctuations. Whether short-term and medium-term fluctuations are abnormal is determined by the results of the first fluctuation detection, while long-term fluctuations are determined by the results of the second fluctuation detection.

[0073] Optionally, fluctuation verification can be performed on the initial performance assessment value to determine whether there are any abnormal fluctuations in the target that cannot be explained by environmental or external events. Temporary abnormal fluctuations are caused by environmental changes or external events, such as a decrease in power generation due to heavy rain; target abnormal fluctuations are caused by equipment failures or component aging in the target distributed photovoltaic power generation system, such as performance degradation due to hot spots. To avoid misclassifying temporary abnormal fluctuations as target abnormal fluctuations, and to prevent overlooking genuine target abnormal fluctuations, fluctuation verification needs to be implemented through multi-angle cross-validation, including statistical fluctuation detection, trend fitting, and correlation with external events.

[0074] In one optional embodiment, the first fluctuation detection result is determined based on the first historical initial efficiency score data of the target distributed photovoltaic power generation system in the third historical time period and the current initial efficiency score, including: performing statistical analysis on the first historical initial efficiency score data to obtain the first mean and first standard deviation of the first historical initial efficiency score data; determining the energy efficiency volatility rate and a first threshold based on the first mean and the first standard deviation; and determining the first fluctuation detection result based on the current initial efficiency score, the energy efficiency volatility rate, and the first threshold.

[0075] The first fluctuation detection result is determined as follows: First, statistical analysis is performed on the first historical initial efficiency score data to calculate the first mean and first standard deviation. Second, the energy efficiency volatility rate and first threshold are determined based on the first mean and first standard deviation. Finally, the current initial efficiency score is compared with the first threshold to obtain the first comparison result. Simultaneously, the energy efficiency volatility rate is compared with a preset volatility threshold to obtain the second comparison result. The first fluctuation detection result is determined based on the first and second comparison results. By comparing the current initial efficiency score with the first threshold, and the energy efficiency volatility rate with the preset volatility threshold, the existence of short-term fluctuations in the efficiency of the target distributed photovoltaic power generation system can be cross-verified, improving the accuracy and rationality of the first fluctuation detection result.

[0076] Optionally, if the initial performance score is greater than a first threshold, the first comparison result indicates that the performance of the target distributed photovoltaic (PV) power generation system exhibits abnormal fluctuations at the current time; otherwise, the first comparison result indicates that the performance of the target distributed PV power generation system does not exhibit abnormal fluctuations at the current time. If the energy efficiency volatility rate is greater than a preset volatility rate threshold, the second comparison result indicates that the performance of the target distributed PV power generation system exhibits abnormal fluctuations at the current time; otherwise, the second comparison result indicates that the performance of the target distributed PV power generation system does not exhibit abnormal fluctuations at the current time. If either the first comparison result or the second comparison result indicates that the performance of the target distributed PV power generation system exhibits abnormal fluctuations at the current time, the first fluctuation detection result is determined to indicate that the performance of the target distributed PV power generation system exhibits short-term or medium-term fluctuations; otherwise, the first fluctuation detection result is determined to indicate that the performance of the target distributed PV power generation system does not exhibit short-term or medium-term fluctuations.

[0077] Optionally, statistical fluctuation detection calculates the first standard deviation and efficiency volatility of the initial historical efficiency score data for the third historical time period (e.g., the most recent 30 days) to obtain the first fluctuation detection result, thereby identifying short-term or medium-term abnormal fluctuations. Under normal circumstances, the efficiency of photovoltaic units will fluctuate slightly around the mean, and the fluctuation range conforms to statistical laws. If the efficiency at a certain moment exceeds this range, it is determined to be an abnormal fluctuation. The specific process is as follows:

[0078] First, select the initial historical performance score data from the third historical time period (e.g., the most recent 30 days, a medium-term time scale) as the historical data sample, and calculate the first mean of the sample. and the first standard deviation The first standard deviation reflects the dispersion of the initial historical performance score data, i.e., the absolute range of normal fluctuations.

[0079] Then, calculate the efficiency volatility. That is, the ratio of the first standard deviation to the first mean. Efficiency volatility reflects the relative range of normal fluctuations, avoiding misjudgments of the fluctuation range due to differences in the first mean.

[0080] Based on the 3σ principle of normal distribution, a short-term anomaly threshold (i.e., the first threshold used to determine whether short-term fluctuations exist) can be set as follows: The mid-term anomaly threshold (i.e., the first threshold used to determine whether mid-term fluctuations exist) is: The initial performance evaluation value at the current moment is compared with the anomaly threshold. If it exceeds the anomaly threshold range, then there is abnormal fluctuation. If the first threshold is... This indicates the existence of short-term abnormal fluctuations; if the first threshold is... This indicates the presence of abnormal fluctuations in the medium term.

[0081] Setting the duration of the third historical time period to 30 days ensures sufficient historical data samples to guarantee reliable statistical results, while also reflecting recent performance fluctuations and preventing discrepancies between the initial historical performance score data and current patterns caused by seasonal variations. If the target distributed photovoltaic power generation system is located in an area with significant seasonal variations, such as large performance differences between summer and winter, the duration of the third historical time period can be adjusted to 15 days to ensure that the historical data samples match the current season.

[0082] In one optional embodiment, the second fluctuation detection result is determined based on the second historical initial efficiency score data of the target distributed photovoltaic power generation system in the fourth historical time period and the current initial efficiency score, including: fitting the second historical initial efficiency score data to obtain a fitting curve; determining the fitted initial efficiency score of the target distributed photovoltaic power generation system at the current time based on the fitted curve; determining the initial efficiency score difference based on the current initial efficiency score and the fitted initial efficiency score; and determining the second fluctuation detection result based on the initial efficiency score difference and a second threshold.

[0083] The second fluctuation detection result is determined as follows: First, the second historical initial efficiency score data is fitted to obtain a fitted curve. This fitted curve represents the expected path of the target photovoltaic system's efficiency change over time without long-term fluctuations. Second, based on this fitted curve, the fitted initial efficiency score of the target distributed photovoltaic power generation system at the current moment is obtained. This fitted initial efficiency score characterizes the theoretical initial efficiency score of the target distributed photovoltaic power generation system at the current moment without long-term fluctuations. Finally, based on the difference between the current initial efficiency score and the fitted initial efficiency score, and a second threshold, the second fluctuation detection result is obtained. By capturing the long-term trend of the target distributed photovoltaic power generation system's efficiency over time through the fitted curve, it is possible to better identify slowly progressing anomalies, such as component aging and long-term dust accumulation. These problems may not be easily detected in the short term, but they will seriously affect the efficiency of the target distributed photovoltaic power generation system in the long term.

[0084] Optionally, if the initial performance score difference is greater than the second threshold, it indicates that the performance of the target photovoltaic system at the current moment deviates from the long-term trend, and there may be long-term degradation or structural problems. In this case, the second fluctuation detection result is that the performance of the target distributed photovoltaic power generation system has long-term fluctuations; otherwise, the second fluctuation detection result is that the performance of the target distributed photovoltaic power generation system does not have long-term fluctuations.

[0085] Optionally, trend fitting uses a time series model to fit the second historical initial efficiency score data for the fourth historical time period to obtain the second fluctuation detection result, identifying abnormal fluctuations that deviate from the fitted curve. Under normal circumstances, the efficiency of photovoltaic units will exhibit a stable trend, such as a slowly decreasing linear trend or a periodic trend with seasonal fluctuations. If the actual efficiency deviates from this trend, there may be abnormal degradation, as detailed below:

[0086] First, the second historical initial performance score data of the fourth historical time period (e.g., the most recent 180 days, a long-term time scale) is selected as the fitting data. The stationarity of the second historical initial performance score data is tested. If the second historical initial performance score data is non-stationary, it is transformed into stationary second historical initial performance score data through differencing before proceeding to the next steps.

[0087] Then, a time series model was used to fit the second historical initial performance score data to obtain a fitted curve, which represents the expected path of performance change.

[0088] Calculate the error (i.e., the initial performance score difference) between the actual historical initial performance score and the corresponding fitted initial performance score at each time step. This yields the errors corresponding to multiple moments within the fourth historical time period. The standard deviation (i.e., the second threshold) of the errors corresponding to these multiple moments is then calculated. If the error at a certain moment Exceeding the second threshold If the value is 0, it is determined to be an abnormal fluctuation point that deviates from the fitted curve.

[0089] Statistical fluctuation detection focuses on short-term dispersion, identifying sudden and large fluctuations; trend fitting focuses on long-term patterns of change, identifying slow and cumulative deviations. Combining the two can cover abnormal fluctuations at different time scales.

[0090] Optionally, after obtaining the first fluctuation detection result and the second fluctuation detection result using the above process, and determining that there is an abnormal fluctuation at the current moment, it is further determined whether the abnormal fluctuation is a temporary abnormal fluctuation caused by external factors, or a target abnormal fluctuation caused by equipment failure or component aging in the target distributed photovoltaic power generation system. For example, there may be abnormal components caused by component failure or abnormal components caused by environmental factors.

[0091] Optionally, external event correlation can be used to determine whether abnormal fluctuations are temporary abnormal fluctuations caused by external factors. External event correlation determines whether abnormal fluctuations are caused by external factors by correlating historical meteorological events and power grid events of the same period. If the performance fluctuations of the target distributed photovoltaic power generation system are caused by external events such as rainstorms, sandstorms, or power grid outages, they are considered temporary abnormal fluctuations; if no external events occur, they are considered target abnormal fluctuations. The specific process is as follows:

[0092] First, establish an event database covering both meteorological and power grid events, recording the time of occurrence, duration, and scope of impact of each event. Meteorological events should include at least heavy rain, heavy snow, sandstorms, strong radiation, and low temperatures; power grid events should include at least power outages, voltage dips, and frequency anomalies.

[0093] Then, when abnormal fluctuations are detected at the current moment through statistical fluctuation detection or trend fitting, the target external events with the same impact range that occurred at the same historical moment during the same period in the event database are retrieved.

[0094] If external events related to the target are retrieved, the causal relationship between the external events and the abnormal fluctuations is further determined. The specific process is as follows:

[0095] Calculate the rate of change in irradiance and the rate of fluctuation in effectiveness before and after the occurrence of the external event. If the absolute value of the correlation coefficient between the two is greater than the preset value, it is a strong correlation, and the abnormal fluctuation is determined to be caused by the external event and is a temporary abnormal fluctuation. If no external event of the same period is retrieved, or the absolute value of the correlation coefficient is less than the preset value, it is a weak correlation, and the abnormal fluctuation is determined to be a target abnormal fluctuation that cannot be explained by the external event.

[0096] By employing three dimensions—statistical fluctuation detection, trend fitting, and external event correlation—a complementary approach can accurately identify abnormal performance fluctuations. Statistical fluctuation detection, based on the statistical characteristics of the initial historical performance score data, defines short-term or medium-term fluctuation thresholds, enabling the identification of sudden and significant abnormal deviations. Trend fitting, through time series models, uncovers long-term performance change patterns, capturing slowly accumulating abnormal degradation and compensating for the limitations of short-term detection in identifying gradual issues. External event correlation, by matching meteorological and power grid events, clarifies the cause of abnormal fluctuations, distinguishing between temporary abnormal fluctuations caused by external environments or events and target abnormal fluctuations caused by equipment malfunctions or component aging. These three methods corroborate each other, both defining whether fluctuations exceed normal fluctuation ranges through quantitative standards and eliminating external interference through causal tracing, avoiding misjudging temporary abnormal fluctuations as target abnormal fluctuations. This ensures that the judgment of fluctuation nature is both comprehensive and accurate, guaranteeing the reliability of early warnings for abnormal performance degradation of target distributed photovoltaic power generation systems.

[0097] Step S108: If abnormal fluctuations in the efficiency of the target distributed photovoltaic power generation system are detected, the cause of the abnormality is determined, and the current initial efficiency score is corrected based on the cause of the abnormality to obtain the current target efficiency score of the target distributed photovoltaic power generation system at the current moment.

[0098] It is understandable that when abnormal fluctuations in the performance of a target distributed photovoltaic (PV) power generation system are detected, data source tracing analysis is used to determine the abnormal causes of these fluctuations. Based on these causes, the initial performance score is then corrected to obtain the current target performance score of the PV power generation system at the current moment. Data source tracing analysis accurately identifies the direct causes of these abnormal fluctuations, avoiding the blindness of anomaly warnings and ensuring their targeted nature. Furthermore, determining the current target performance score based on the causes makes it more closely reflect the current state of the PV power generation system, improving the adaptability and accuracy of anomaly warnings.

[0099] Optionally, when fluctuation verification determines that there are abnormal fluctuations in the target that cannot be explained by the environment or external events, it is necessary to perform preliminary fault or anomaly localization, determine the cause of the anomaly, and optimize the evaluation parameters based on the cause. This specifically includes two parts: data source analysis and evaluation parameter optimization. The purpose of data source analysis is to find the abnormal cause of the target's abnormal fluctuations and avoid blind early warnings; parameter optimization is to adjust the evaluation parameters based on the cause of the anomaly, eliminate the interference of abnormal factors on the current initial performance score, obtain a more realistic current target performance score, and provide an accurate basis for determining the anomaly warning level.

[0100] Data tracing analysis includes retrieving thermal imaging data of the components to locate hot spot areas, checking the dust accumulation data of the components to confirm the light transmittance, and analyzing the inverter efficiency curve to identify sudden drops in conversion efficiency, corresponding to three dimensions: components, external environment, and core equipment.

[0101] First, thermal imaging data of the module is retrieved to locate the hot spot area. Finding a hot spot on the thermal image (i.e., thermal imaging data) allows for preliminary identification of the problem. Further, the identified hot spot area is inspected on-site. If the hot spot area is located at the edge of the module and is obstructed by tree branches, billboards, or other objects, the cause of the anomaly is external obstruction. If the hot spot area is not significantly obstructed, but there are scratches or microcracks on the module surface, the cause of the anomaly is microcracks in the solar cells. If the hot spot area is concentrated near the module junction box, the cause of the anomaly is poor contact in the junction box leading to localized overheating.

[0102] Secondly, the transmittance is confirmed by checking the dust accumulation data of the components. Historical data on the transmittance of dust accumulation on the component surface is analyzed. If a continuous downward trend in transmittance is found recently, and it is highly correlated with the downward trend of the initial performance score, then component dust accumulation is determined to be a potential cause of abnormal fluctuations in the target. Furthermore, a portable transmittance meter is taken to the site to perform repeated measurements on different areas of the component surface. If the measurement results are consistent with historical data, then component dust accumulation is confirmed as the cause of the abnormality. If the transmittance measured on-site does not match the historical data, then it is necessary to investigate whether the transmittance meter is faulty or the image recognition algorithm is biased.

[0103] Finally, the inverter efficiency curve is analyzed to identify sudden drops in conversion efficiency. The inverter's efficiency-load curve is plotted within the time period corresponding to the target abnormal fluctuation and compared with the historical normal curve. If the efficiency curve is found to shift downward as a whole or to drop sharply in a specific load range, the cause of the abnormality may be an abnormality in the inverter itself or its DC-side input.

[0104] In one optional embodiment, the initial performance score is corrected based on the cause of the anomaly to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment. This includes: if the cause of the anomaly is an abnormal component within the target distributed photovoltaic power generation system due to its own component failure, determining the health index of the abnormal component based on its hotspot area and output power at the current moment, wherein the hotspot area is obtained based on the thermal imaging data of the abnormal component at the current moment; correcting the basic performance score based on the health index to obtain a corrected performance score; and correcting the initial performance score based on the corrected performance score to obtain the current target performance score.

[0105] Understandably, if the cause of the anomaly indicates the presence of an abnormal component within the target distributed photovoltaic power generation system due to its own fault, the initial performance score is corrected as follows: First, the health index of the abnormal component is determined based on its current hotspot area and output power. Second, the base performance score is corrected based on the abnormal component's health index to obtain a corrected performance score. Finally, the initial performance score is corrected based on the corrected performance score to obtain the current target performance score. By accurately measuring the hotspot area and adjusting the health index of the abnormal component accordingly, the impact of component anomalies on the performance of the target distributed photovoltaic power generation system can be quantified more precisely, making the correction of the initial performance score more reasonable and improving the accuracy of the current target performance score.

[0106] Optionally, the hot spot region can be located and the hot spot area can be determined based on the thermal imaging data of the abnormal component at the current moment.

[0107] Optionally, adjusting the component health index applies when the anomaly is due to a problem with the component itself (i.e., there are abnormal components within the target distributed photovoltaic power generation system due to their own failure). This parameter reflects the performance status of a component, with a value between 0 and 1, where 0 indicates the component is unusable and 1 indicates it is in good condition. In the initial evaluation model (the fusion formula used to calculate the initial performance evaluation value), the component health index is set to 1 by default. A grading standard can be pre-set based on the impact of component anomalies on power. When an abnormal component is present, the grading standard is adjusted according to the severity of the problem. For example, the grading criteria can be set as follows:

[0108] Minor faults (hot spot area <5cm², microcrack in a single cell): =0.9-0.95;

[0109] Moderate fault (hot spot area) (5-15cm², multiple battery cells with microcracks) =0.8-0.89;

[0110] Severe faults (hot spot area > 15cm², module encapsulation layer aging): =0.7-0.79;

[0111] Extremely severe fault (component output power < 50% of rated power): <0.7.

[0112] Component issues primarily affect indicators such as actual power generation and string current in the basic performance data. Therefore, it is necessary to correct the basic performance score using the component's health index to obtain the corrected performance score. The environmental disturbance coefficient and time-series dynamic data are not affected by the component health index, so they remain unchanged. Based on the corrected performance score, the corrected performance evaluation value (i.e., the current target performance score) can be calculated by substituting it back into the fusion formula. .

[0113] In one optional embodiment, the initial performance score is corrected based on the cause of the anomaly to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment. This includes: if the cause of the anomaly is the presence of an abnormal component within the target distributed photovoltaic power generation system due to environmental factors, determining an environmental attenuation coefficient based on the transmittance and target temperature of the abnormal component at the current moment, wherein the transmittance is obtained based on the dust accumulation data of the abnormal component at the current moment, and the target temperature is obtained based on the component temperature data of the abnormal component at the current moment; correcting the environmental interference coefficient based on the environmental attenuation coefficient to obtain a corrected interference coefficient; and correcting the initial performance score based on the corrected interference coefficient to obtain the current target performance score.

[0114] It is understandable that if the anomaly indicates the presence of abnormal components within the target distributed photovoltaic (PV) power generation system due to environmental factors, the initial performance score is corrected as follows: First, based on the dust accumulation data and component temperature data of the abnormal component at the current moment, the transmittance and target temperature of the abnormal component at the current moment are determined, thereby determining the environmental degradation coefficient. This environmental degradation coefficient is used to quantify the performance degradation caused by environmental factors. Second, based on the environmental degradation coefficient, the environmental interference coefficient is corrected to obtain the corrected interference coefficient. Finally, the initial performance score is corrected based on the corrected interference coefficient to obtain the current target performance score. By correlating transmittance with the environmental degradation coefficient, the negative impact of environmental factors on the performance of the target distributed PV power generation system is successfully quantified. This avoids misinterpreting performance degradation caused by environmental factors as a fault in the equipment (such as components, inverters, and component cells) within the target distributed PV power generation system, reducing the risk of underreporting performance degradation caused by environmental factors.

[0115] Optionally, light transmittance measures the ability of a module to receive direct sunlight through dust accumulation, which directly affects the module's power generation efficiency. The light transmittance of the dust-accumulated module can be determined by acquiring dust accumulation data from optical sensors installed on the module or by using image recognition technology.

[0116] Optionally, adjusting the environmental degradation coefficient is applicable when the anomaly is due to abnormal components within the target distributed photovoltaic power generation system caused by environmental factors. The environmental degradation coefficient is a parameter that corrects for environmental interference; it is set to 1 by default when calculating the initial performance evaluation value. A grading standard can be pre-set based on the degree of environmental anomaly; when environmental factors exceed the normal range, the environmental degradation coefficient is adjusted. This enhances the environmental disturbance coefficient's response to abnormal environments. For example, the grading standard could be set as follows:

[0117] Minor abnormalities (transmittance) 70%-80%, local temperature 45-50℃): =1.1-1.2;

[0118] Moderate abnormality (transmittance) 60%-69%, local temperature 51-55℃): =1.3-1.4;

[0119] Severe abnormalities (transmittance <60%, local temperature >55℃): =1.5-1.6.

[0120] The local temperature is the target temperature of the abnormal component, which is obtained based on the component temperature data of the abnormal component at the current moment.

[0121] Corrected environmental interference coefficient (i.e., corrected interference coefficient) At this point, the basic performance score and time-series dynamic data remain unchanged. Based on the corrected interference coefficient, the corrected performance evaluation value can be calculated by substituting it back into the fusion formula. .

[0122] Optionally, the corrected efficacy assessment value can also be determined by switching the assessment model. Switching the assessment model is suitable when the causes of the anomaly are complex or the initial assessment model cannot accurately quantify the impact of the anomaly. The initial assessment model uses a linear fusion formula, which is suitable for single, linear efficacy influencing factors. When nonlinear effects exist, it is necessary to switch to a nonlinear assessment model, such as a model trained based on machine learning, and use the new model to calculate the corrected efficacy assessment value. .

[0123] Step S110: Based on the current target performance score, issue an anomaly warning for the target distributed photovoltaic power generation system.

[0124] It is understandable that the process of issuing anomaly warnings for the target distributed photovoltaic power generation system based on the current target performance score fully combines the correction of the current initial performance score with in-depth analysis of abnormal fluctuations, ensuring the accuracy and effectiveness of the anomaly warnings and achieving the reliability of the anomaly attenuation warning for the target distributed photovoltaic power generation system.

[0125] In one optional embodiment, an anomaly warning is issued for the target distributed photovoltaic power generation system based on the current target performance score, including: acquiring historical target performance score data of the target distributed photovoltaic power generation system and predicted meteorological data of the area where the target distributed photovoltaic power generation system is located; performing statistical analysis on the historical target performance score data to obtain a second mean and a second standard deviation of the historical target performance score data; determining the expected performance benchmark value of the target distributed photovoltaic power generation system at the current moment based on the second mean, the annual decay rate and service life of the target distributed photovoltaic power generation system; determining the performance correction coefficient based on the predicted meteorological data; determining a third threshold, a fourth threshold and a fifth threshold based on the second standard deviation, the expected performance benchmark value and the performance correction coefficient; and determining the warning level of the target distributed photovoltaic power generation system at the current moment based on the current target performance score, the third threshold, the fourth threshold and the fifth threshold, and issuing an anomaly warning.

[0126] The process involves acquiring historical target performance scores for the target distributed photovoltaic (PV) power generation system and forecast meteorological data for the region where the system is located. Statistical analysis is performed on the historical performance scores to calculate the second mean and second standard deviation. Based on these second mean and second standard deviation, combined with the annual degradation rate and service life of the target PV power generation system, the expected performance benchmark value for the system at the current moment is determined. Based on the forecast meteorological data, performance correction coefficients are determined, and combined with the second standard deviation and the expected performance benchmark value, third, fourth, and fifth thresholds are determined. Based on the current target performance score, the third threshold, the fourth threshold, and the fifth threshold, the warning level for the target PV power generation system at the current moment is determined, and an abnormality warning is issued. The expected performance benchmark value reflects the performance level that the target PV power generation system should have at the current moment under normal degradation rates; the performance correction coefficient reflects the possible impact of forecast meteorological data on the performance of the target PV power generation system. By setting warning thresholds (including the third, fourth, and fifth thresholds) based on historical target performance scores, annual degradation rate, service life, and future meteorological forecasts, the actual performance of the target distributed photovoltaic power generation system can be reflected more accurately and in more detail over time and in the environment, reducing the possibility of false alarms and missed alarms and improving the accuracy of abnormal warnings.

[0127] Optionally, the corrected performance evaluation value can be compared with a number of preset dynamic thresholds (including a third threshold, a fourth threshold, and a fifth threshold), which are generated based on historical target performance score data for the same period in history and future forecast meteorological data for the area where the target distributed photovoltaic power generation system is located.

[0128] Optionally, the dynamic threshold generation logic integrates historical patterns and future expectations, enabling the anomaly warning benchmark to adapt to seasonal changes, normal equipment aging patterns, and future weather conditions, thereby avoiding false alarms during rainy weather or missed alarms during high-radiation weather. The specific dynamic threshold generation method is as follows:

[0129] First, obtain historical target performance score data for the target distributed photovoltaic power generation system from historical dates corresponding to the current time over the past few years. Then, perform statistical analysis on these historical target performance score data and calculate its second mean. Second standard deviation Simultaneously, based on the annual degradation rate f provided by the equipment manufacturer or obtained through long-term data analysis, and combined with the service life of the target distributed photovoltaic power generation system, the second mean is adjusted for degradation to obtain the theoretical expected performance benchmark value at the current moment. , where n is the number of years of use.

[0130] Then, based on the predicted irradiance and ambient temperature over a future period (i.e., predicted meteorological data), the expected performance correction coefficient under future weather conditions is calculated. ;

[0131] Combining baselines and meteorological corrections, a set of dynamic thresholds is generated, typically including a attention threshold (i.e., the third threshold), a warning threshold (i.e., the fourth threshold), and an emergency threshold (i.e., the fifth threshold). The attention threshold... This threshold is slightly lower than the historical benchmark level for the same period after weather correction. Exceeding this threshold only indicates that the performance of the target distributed photovoltaic power generation system has begun to deviate from the normal expected range, suggesting that attention is needed; Warning threshold This threshold is significantly lower than the benchmark, indicating that the performance degradation of the target distributed photovoltaic power generation system is unlikely to be caused solely by normal fluctuations or weather factors, and there is a high probability of equipment malfunction or performance degradation; Emergency threshold This threshold is far below the benchmark, indicating that the performance of the target distributed photovoltaic power generation system has deteriorated significantly and may be accompanied by major failures or safety hazards.

[0132] Optionally, based on the comparison results, a gradient anomaly warning signal associated with the efficiency degradation rate and potential power generation loss of the target distributed photovoltaic power generation system can be triggered. Gradient warnings can be divided into Level 1, Level 2, and Level 3 warnings (i.e., warning levels), which are determined when the efficiency assessment value is corrected. Below the attention threshold However, it is above the warning threshold. When a Level 1 warning is triggered, the warning signal may include the abnormal photovoltaic unit number, corrected assessment value, degradation rate, potential monthly power generation loss, identified cause of the anomaly, and recommended measures. The warning signal is pushed to the regional operation and maintenance manager via SMS and the operation and maintenance platform APP (application). When the corrected performance assessment value... Below the warning threshold But above the emergency threshold When a Level 2 warning is triggered, the warning signal may include the abnormal photovoltaic unit number, corrected performance assessment value, degradation rate, potential monthly power generation loss, and recommended measures. In addition to being sent to the operation and maintenance manager, the warning signal must also be copied to the regional supervisor, who must track the progress of the handling. When the corrected performance assessment value... When the value falls below the emergency threshold, a Level 3 warning is triggered. The warning signal may include the abnormal photovoltaic unit number, the corrected assessment value, the degradation rate, the potential monthly power generation loss, the specific cause of the abnormality, and emergency measures. In addition to being pushed to the operation and maintenance manager and the regional supervisor, the warning signal must also be reported to the operation and maintenance headquarters. The headquarters will arrange technical experts to provide remote guidance to ensure the rapid restoration of the efficiency of the target distributed photovoltaic power generation system.

[0133] Optionally, after the warning signal is triggered, a closed-loop management mechanism can be established. After the operation and maintenance personnel have finished handling the issue, multi-dimensional performance perception data is re-collected, and a new corrected performance evaluation value is calculated. If the corrected performance evaluation value rises back to the attention threshold, the mechanism will be activated. If the above conditions are met, the warning will be marked as lifted in the operation and maintenance platform. If the corrected performance evaluation value does not recover, the data source analysis needs to be carried out again, and the handling measures need to be adjusted until the performance of the target distributed photovoltaic power generation system returns to normal, ensuring that the abnormal degradation is completely resolved and avoiding repeated warnings.

[0134] Through the above steps S102 to S110, the current target performance score of the target distributed photovoltaic power generation system can be determined by analyzing the current performance data, current environmental interference data, and performance index data. Based on the current target performance score, anomaly warnings can be issued for the target distributed photovoltaic power generation system. This achieves the technical effect of improving the accuracy of anomaly warnings for the target distributed photovoltaic power generation system, thereby solving the technical problem of inaccurate anomaly warnings for distributed photovoltaic power generation systems in related technologies.

[0135] Based on the above embodiments and optional embodiments, this application proposes an implementation method for an optional early warning method of an anomaly in a distributed photovoltaic power generation system. This implementation method can be understood as an early warning method for abnormal degradation of distributed photovoltaic power generation efficiency. Figure 2 This is a flowchart of an optional anomaly early warning method for a distributed photovoltaic power generation system provided according to an embodiment of this application, such as... Figure 2 As shown, the steps of the early warning method for abnormal performance degradation of distributed photovoltaic power generation include:

[0136] Step S1: Obtain multi-dimensional performance perception data of each photovoltaic unit in the distributed photovoltaic system, including basic performance data (i.e., current performance data), current environmental interference data, and time series dynamic data (i.e., performance index data).

[0137] Basic performance data includes actual power generation, module temperature, string current, string voltage, inverter efficiency, and cumulative irradiance; environmental interference data includes module surface area transmittance, module local hot spot temperature, grid supply voltage, and ambient wind speed; time-series dynamic data includes the performance change rate in the first time period (i.e., the first historical time period), the performance degradation rate in the second time period (i.e., the second historical time period), and the year-on-year performance deviation in the third time period.

[0138] A photovoltaic unit refers to the smallest unit in a target distributed photovoltaic power generation system that can independently perform power generation efficiency testing. It can be a string or a subarray composed of multiple strings, depending on the deployment granularity of the on-site testing equipment.

[0139] By integrating multi-dimensional performance perception data, fluctuation verification, data source analysis, and optimized evaluation parameters (including health index, environmental degradation coefficient, etc.), a closed-loop mechanism can incorporate multi-source data, such as basic performance data, current environmental interference data, and time-series dynamic data, into the entire performance evaluation process. For example, in the initial evaluation stage, the power generation is normalized by comprehensively considering data such as cumulative irradiance and local hot spot temperature of the modules, reducing the interference of environmental fluctuations on the evaluation results. In the fluctuation verification stage, abnormal fluctuations that do not match known environmental factors or operational events are identified through time-series dynamic data, improving the sensitivity of anomaly identification. In the data source analysis stage, the changing trends of multi-dimensional data are traced in reverse for abnormal fluctuations, distinguishing between the equipment's own degradation and external incidental influences in the target distributed photovoltaic power generation system, and dynamically correcting the evaluation parameters. This makes the early warning judgment no longer dependent on a single static threshold, but deeply correlated with the actual operating status of the target distributed photovoltaic power generation system, which can effectively reduce the false alarm rate compared with the traditional fixed threshold method.

[0140] The initial assessment phase aims to eliminate environmental interference and establish a dynamic performance benchmark. The fluctuation verification phase aims to identify latent anomalies by combining historical data (including first and second historical initial performance score data) with external event logs for difference analysis, avoiding misclassifying normal fluctuations as abnormal fluctuations and temporary abnormal fluctuations as target anomalies. The data source analysis phase aims to pinpoint the root cause by using multi-dimensional data correlation analysis to distinguish the abnormal causes of performance degradation and adaptively optimize assessment parameters to achieve a balance between false alarm suppression and false negative prevention. Through this multi-stage verification and source tracing mechanism, accurate identification of abnormal degradation in target distributed photovoltaic power generation systems can be achieved, thereby improving the accuracy of anomaly early warning.

[0141] In summary, by integrating multi-dimensional performance perception data, constructing a fluctuation verification and data traceability analysis mechanism, and dynamically optimizing evaluation parameters based on the causes of anomalies, it is possible to effectively distinguish between the actual performance degradation of the target distributed photovoltaic power generation system and the performance degradation caused by environmental fluctuations, reduce the risk of false alarms and missed alarms, improve the accuracy and timeliness of anomaly warnings, and provide a reliable basis for the refined operation and maintenance of the target distributed photovoltaic power generation system.

[0142] Basic performance data is a set of core parameters that directly reflect the conversion of solar energy into electrical energy by the target distributed photovoltaic power generation system. It is used to calculate the deviation between theoretical power generation and actual power generation. Specifically, it includes actual power generation, module temperature, string current, string voltage, inverter efficiency, and cumulative irradiance.

[0143] Actual power generation refers to the total amount of electrical energy actually delivered to the grid or load by the photovoltaic units in a target distributed photovoltaic power generation system within a specific time interval. The unit is usually kilowatt-hours (kWh), which can be obtained through the metering function of smart meters or inverters. Actual power generation is the final efficiency output of the photovoltaic units, directly reflecting the actual effect of converting input irradiance into electrical energy, and is the most intuitive indicator for judging whether efficiency has degraded. If irradiance is sufficient but power generation remains consistently low, it directly indicates abnormal degradation. Ignoring power generation may miss hidden anomalies such as normal parameters but low overall conversion efficiency.

[0144] Module temperature refers to the actual temperature of the photovoltaic panels in a target distributed photovoltaic power generation system during operation. It can be obtained through sensors installed on the module backsheet or using infrared temperature measurement technology. The photoelectric conversion efficiency of the module is negatively correlated with temperature. Excessively high module temperatures not only directly lead to a reduction in immediate efficiency, but also accelerate the aging of the module encapsulation layer over a long period, causing irreversible long-term degradation. Ignoring module temperature may lead to misjudging the performance degradation caused by high temperatures as module failure, or overlooking the potential aging risks caused by excessively high temperatures.

[0145] String current and string voltage refer to the current flowing through the entire string and the voltage across its terminals, respectively, and can be acquired using string-level current and voltage sensors. An abnormally low current compared to other parallel strings or historical data from the same period directly indicates a problem in that string, such as obstruction, wiring faults, or partial component failure. Abnormal voltage may indicate series problems, such as reverse bias caused by hot spots on components, diode conduction in the junction box, or abnormal maximum power point tracking. Combining both can be used to plot current-voltage curves and pinpoint the problematic string.

[0146] Inverter efficiency refers to the efficiency with which an inverter converts direct current (DC) to alternating current (AC), i.e., the ratio of AC output power to DC input power. This efficiency is provided by the inverter's own monitoring system. This data reflects whether the power generated by the modules can be effectively transmitted to the grid, helping to distinguish whether the problem originates from the DC or AC side. Ignoring inverter efficiency can lead to overlooking the cause of degradation in cases where the modules are functioning normally but the conversion process is malfunctioning.

[0147] Cumulative irradiance refers to the cumulative amount of solar radiation energy that falls on the plane containing a photovoltaic unit within a specific time interval. The unit is typically kilowatt-hours per square meter (kWh / m²), and it can be measured using an on-site irradiance meter. Cumulative irradiance serves as a benchmark for theoretical power generation. Ignoring low irradiance while focusing solely on low power generation can lead to a misjudgment of insufficient normal input as abnormal degradation. Conversely, if irradiance is sufficient but power generation is low, the anomaly is clearly caused by either the equipment itself within the target distributed photovoltaic power generation system or environmental interference.

[0148] Environmental disturbance data refers to external factors that affect the performance of photovoltaic units in a target distributed photovoltaic power generation system. It is used to identify and quantify the power generation loss caused by temporary and recoverable external factors, including the surface area transmittance of the module, the temperature of local hot spots on the module, the grid supply voltage, and the ambient wind speed.

[0149] Module surface transmittance refers to the degree to which the proportion of sunlight transmitted to the solar cells decreases due to contaminants such as dust, dirt, and bird droppings covering the module surface. It can be estimated indirectly through specialized optical sensors or by analyzing the differences in output characteristics between clean and contaminated states. Contaminant coverage blocks light, directly reducing the module's short-circuit current and thus lowering output power. This data allows for precise quantification of power generation loss caused by contamination. When a decrease in power generation is detected, if a significant decrease in transmittance is also observed, the primary cause can be identified as contaminant coverage.

[0150] Localized hotspot temperature refers to the abnormal temperature of a localized area on a module that is significantly higher than the module's average temperature due to factors such as cell defects, internal bypass diode failure, or severe shading. This temperature can be obtained through regular inspections using infrared thermal imagers mounted on drones or by detection with fixed thermal imagers. Hotspots not only cause power loss in the corresponding area but also accelerate module aging and can even lead to fires. This data can be used to directly locate defective modules.

[0151] The grid supply voltage refers to the grid voltage at the point of connection of the target distributed photovoltaic (PV) power generation system. This data comes from grid-side monitoring data or inverter AC-side monitoring data. When the grid voltage is too high or fluctuations exceed the inverter's allowable range, the inverter will proactively reduce its output power or disconnect from the grid in accordance with grid connection standards to protect grid safety. In this case, the decrease in power generation is a normal protective behavior of the inverter, not a fault. By monitoring this data, it is possible to avoid misreporting reduced efficiency caused by grid problems as a fault in the target distributed PV power generation system.

[0152] Ambient wind speed refers to the wind speed at the location of a photovoltaic array (composed of multiple strings), which can be measured by an anemometer. Wind speed affects the heat dissipation of the modules, thus affecting the module temperature. At the same time, strong winds may blow debris from the ground, cause trees to sway, create dynamic shadows on the modules, and cause rapid fluctuations in power generation. Recording wind speed helps to explain some short-term, drastic power fluctuations.

[0153] Time-series dynamic data reflects the trend of efficiency changes of a target distributed photovoltaic power generation system over time. It can reveal the changing trend and periodicity of efficiency from a time dimension, specifically including the efficiency change rate in the first time period, the efficiency decay rate in the second time period, and the year-on-year efficiency deviation in the third time period.

[0154] The efficiency change rate within the first time period corresponds to a short-term time scale, such as the most recent 24 hours or the most recent 7 days. The efficiency change rate can be the rate of change in average daily power generation, or the short-term volatility of the performance index. This indicator is used to capture sudden declines in efficiency.

[0155] The efficiency degradation rate within the second time period corresponds to a medium-term time scale, such as the most recent 3 months or the most recent 6 months. The efficiency degradation rate is used to quantify the linear or non-linear decline trend of efficiency. It can be obtained by performing a linear regression on the historical baseline efficiency data of the second historical time period and taking the normalized value of its slope as the efficiency degradation rate. This indicator is used to identify slow component aging or progressive contamination.

[0156] The third timeframe for performance year-over-year deviation corresponds to a long-term time scale, typically an annual comparison. Year-over-year deviation refers to the deviation rate between the current performance and the performance at the same historical time last year, used to eliminate the influence of seasonal factors. Solar irradiance, temperature, and sunshine duration all exhibit strong seasonal patterns; year-over-year comparisons, after eliminating seasonal factors, most directly reflect the long-term changes in the performance of the target distributed photovoltaic power generation system.

[0157] The aforementioned basic performance data focuses on the equipment status of core conversion links and target distributed photovoltaic power generation systems, directly related to power output and key equipment operating parameters, serving as the foundation for identifying abnormal degradation. Environmental interference data focuses on external, temporary, and recoverable influencing factors, quantifying performance fluctuations not caused by the equipment itself, and providing support for distinguishing between normal interference and abnormal failures. Time-series dynamic data spans short-term, medium-term, and long-term time scales, capturing sudden changes, gradual degradation, and long-term trends in performance, eliminating interference from time-related factors such as seasonality. These three types of data comprehensively cover the dimensions of performance impact, fully collecting key performance-related information, ensuring that the identification of abnormal performance degradation in target distributed photovoltaic power generation systems neither misses real problems nor misjudges normal fluctuations as abnormal fluctuations.

[0158] Step S2: Calculate the initial performance evaluation value (i.e., the current initial performance score) based on multi-dimensional performance perception data.

[0159] Since the acquired multi-dimensional performance perception data are independent and cannot be directly used to determine whether performance is abnormal, it is necessary to transform the scattered data into a single, quantifiable evaluation indicator, the initial performance evaluation value, through fusion calculation. The specific process for determining the initial performance evaluation value is as follows:

[0160] Step S21: Normalize the basic performance data to obtain the basic performance score.

[0161] Because the units and magnitudes of the basic performance data vary significantly, direct fusion could lead to large-scale data overly dominating the evaluation results. Therefore, normalization is necessary to convert all basic performance indicators into normalized scores within the range of 0-100. Linear normalization can be used, with reasonable optimal and worst values ​​set for the characteristics of different indicators. In the basic performance data, actual power generation, string current, string voltage, inverter efficiency, and cumulative irradiance are positive indicators; for positive indicators, a larger value indicates better performance. Module temperature is a negative indicator; for negative indicators, a larger value indicates worse performance. Based on the above analysis, the basic performance data is normalized using the following method:

[0162]

[0163]

[0164] in, For the normalized score of the i-th indicator, the first formula is used to calculate the normalized score of the positive indicator, and the second formula is used to calculate the normalized score of the negative indicator. This is the actual measured value of the indicator; The optimal value of the i-th indicator is the historical maximum value or design rated value within a predetermined historical period when it is a positive indicator, and the historical minimum value or safety threshold within a predetermined historical period when it is a negative indicator. The worst value of the i-th indicator is the historical minimum or fault threshold in the predetermined historical time period when it is a positive indicator, and the historical maximum or danger threshold in the predetermined historical time period when it is a negative indicator.

[0165] After obtaining the normalized scores of each basic performance data point, a weighted average of these scores is calculated to obtain the comprehensive basic performance score. The basic performance score is determined as follows:

[0166]

[0167] in, Basic performance rating; The weight of the i-th indicator can be allocated according to the importance of each indicator to the overall effectiveness.

[0168] Step S22: Classify the environmental interference data into positive and negative categories, and calculate the environmental interference coefficient based on the historical data to determine the direction and degree of influence of each environmental interference factor on the performance.

[0169] Environmental factors have both enhancing and inhibiting effects on the performance of a target distributed photovoltaic power generation system. Directly overlaying all environmental data would obscure the direction of influence, leading to assessment bias. Therefore, it is necessary to classify and then calculate these factors. The specific classifications of positive and negative environmental factors are as follows:

[0170] Positive environmental factors: Ambient wind speed; when the ambient wind speed is moderate, heat dissipation is accelerated and efficiency is improved.

[0171] Negative environmental factors include: light transmittance of dust on the module surface, temperature of local hot spots on the module, and grid supply voltage. Among these, low light transmittance has the best suppression effect, high temperature has the best suppression effect, and voltage exceeding the limit has the best suppression effect.

[0172] Specifically, the environmental interference coefficient can be calculated based on historical environmental interference data. A multiple linear regression model is used to establish the mapping relationship between each environmental factor and the performance deviation rate, obtaining the influence coefficient of each environmental factor. This coefficient is then calculated based on current environmental interference data. The performance deviation rate is the difference between the actual baseline performance score and the ideal performance score, representing a percentage of the ideal performance score. The ideal performance score refers to the baseline performance score under standard environmental conditions: wind speed of 2 m / s, light transmittance of 100%, no hot spots, and grid voltage of 220 V. It can be determined by the average of historical baseline performance scores that meet the standard environmental conditions.

[0173] The process of constructing a multiple linear regression model is as follows:

[0174] First, select the ambient wind speed over a past period of time. Component surface area gray transmittance Component local hot spot temperature Grid voltage Given n sets of data and their corresponding historical baseline performance scores, calculate the performance deviation rate for each set of data. :

[0175]

[0176] in, Rate the ideal performance. Assess the actual basic performance rating; A positive value indicates that the performance is lower than expected. A negative value indicates that the performance is higher than expected.

[0177] Then, based on ambient wind speed Component surface area gray transmittance Component local hot spot temperature Grid voltage As the independent variable, Using [variable name] as the dependent variable, construct a multiple linear regression model:

[0178]

[0179] Where a, b, c, and d are regression coefficients, which can be obtained by fitting the above n sets of data using the least squares method; e is a constant term, which can be set to 20, representing the baseline performance deviation rate when there is no environmental data.

[0180] Based on the above multiple linear regression model, the environmental disturbance coefficient The calculation formula is:

[0181]

[0182] The negative sign here serves to convert the performance deviation rate into an environmental interference coefficient, that is, when... When the value is positive, the efficiency is lower than expected, and environmental factors cause negative interference. When ΔE is negative, the efficiency is higher than expected, and environmental factors cause positive interference. It is positive.

[0183] Step S23: Apply trend weighting to the time-series dynamic data, assigning dynamic weights based on the efficiency change rate within the first time period, the efficiency decay rate within the second time period, and the time sensitivity of the year-on-year efficiency deviation within the third time period.

[0184] Time-series dynamic data reflects performance changes at different time scales. However, changes at different time scales have varying early warning value for abnormal degradation of the target distributed photovoltaic power generation system. For example, short-term sharp fluctuations may indicate a sudden failure and require close monitoring, while long-term slow changes may indicate normal aging and can be given less attention. Therefore, different dynamic weights need to be assigned to different performance indicators based on their time sensitivity. For short-term performance change rates… It is highly sensitive to sudden failures, and the dynamic weight α should be set high to enable rapid response to drastic changes; for the medium-term performance decay rate... It reflects trend changes, with a moderately set dynamic weight β, used to capture gradual problems; for long-term performance year-on-year deviations... It is used for annual health comparisons. The dynamic weight γ is usually low, but it can provide validation from a long-term perspective. The specific value of the dynamic weight can be set by expert experience or determined by training a simple model using historical data, ensuring that α+β+γ=1 or meet other normalization criteria.

[0185] Step S24: The basic performance score, environmental interference coefficient, and time series dynamic weight are calculated using a preset fusion formula to obtain the initial performance evaluation value.

[0186] The preset fusion formula is:

[0187]

[0188] in, This is the initial performance evaluation value. The closer the value is to 1, the more ideal the performance; the lower the value, the worse the performance. Basic performance rating; Environmental interference coefficient; This constitutes the static-environmental correction term for effectiveness, meaning that if the current environment is unfavorable, the initial effectiveness assessment value will be reduced, and if the environment is favorable, it will be increased. , , These are the efficiency change rate in the first time period, the efficiency decay rate in the second time period, and the year-on-year performance deviation in the third time period, respectively. , , These are the dynamic weights for the rate of change in efficiency, the rate of decline in efficiency, and the year-on-year deviation in efficiency, respectively. It is a dynamic trend amplification term of efficiency, when , , When the value is negative, the value within the parentheses will be less than 1, thus further attenuating the static-environment correction term, amplifying the severity of the performance decline, and making the initial performance assessment value sensitive to negative trends.

[0189] Step S3: Perform fluctuation verification on the initial performance evaluation value to determine whether there are any abnormal fluctuations in the target that cannot be explained by the environment or external events.

[0190] Temporary abnormal fluctuations are caused by environmental changes or external events, such as heavy rain leading to a decrease in power generation; target abnormal fluctuations are caused by equipment failures or component aging in the target distributed photovoltaic power generation system, such as hot spots causing performance degradation. To avoid misclassifying temporary abnormal fluctuations as target abnormal fluctuations, and to prevent overlooking genuine target abnormal fluctuations, fluctuation verification needs to be achieved through multi-angle cross-validation, including statistical fluctuation detection, trend fitting, and correlation with external events.

[0191] Step S31: Statistical fluctuation detection obtains the first fluctuation detection result by calculating the first standard deviation and performance volatility of the first historical initial performance score data in the third historical time period (e.g., the most recent 30 days), thereby identifying short-term or medium-term abnormal fluctuations.

[0192] Fluctuation detection identifies short-term or medium-term abnormal fluctuations by calculating the first standard deviation and efficiency volatility of the initial historical efficiency score data for the third historical time period. Under normal circumstances, the efficiency of photovoltaic units fluctuates slightly around the mean, and the fluctuation range conforms to statistical laws. If the efficiency at a certain moment exceeds this range, it is determined to be an abnormal fluctuation. The specific process is as follows:

[0193] First, select the initial historical performance score data from the third historical time period (e.g., the most recent 30 days, a medium-term time scale) as the historical data sample, and calculate the first mean of the sample. and the first standard deviation The first standard deviation reflects the dispersion of the initial historical performance score data, i.e., the absolute range of normal fluctuations.

[0194] Then, calculate the efficiency volatility. That is, the ratio of the first standard deviation to the first mean. Efficiency volatility reflects the relative range of normal fluctuations, avoiding misjudgments of the fluctuation range due to differences in the first mean.

[0195] Based on the 3σ principle of normal distribution, a short-term anomaly threshold (i.e., the first threshold used to determine whether short-term fluctuations exist) can be set as follows: The mid-term anomaly threshold (i.e., the first threshold used to determine whether mid-term fluctuations exist) is: The initial performance evaluation value at the current moment is compared with the anomaly threshold. If it exceeds the anomaly threshold range, then there is abnormal fluctuation. If the first threshold is... This indicates the existence of short-term abnormal fluctuations; if the first threshold is... This indicates the presence of abnormal fluctuations in the medium term.

[0196] Setting the duration of the third historical time period to 30 days ensures sufficient historical data samples to guarantee reliable statistical results, while also reflecting recent performance fluctuations and preventing discrepancies between the initial historical performance score data and current patterns caused by seasonal variations. If the target distributed photovoltaic power generation system is located in an area with significant seasonal variations, such as large performance differences between summer and winter, the duration of the third historical time period can be adjusted to 15 days to ensure that the historical data samples match the current season.

[0197] Step S32: Trend fitting uses a time series model to fit the second historical initial performance score data of the fourth historical time period to obtain the second fluctuation detection result and identify abnormal fluctuations that deviate from the fitted curve.

[0198] Trend fitting identifies anomalous fluctuations that deviate from the fitted curve by using a time series model to fit the initial historical performance score data for the fourth historical period. Under normal circumstances, the performance of photovoltaic units exhibits a stable trend, such as a slow, linear decline or a cyclical trend with seasonal fluctuations. If the actual performance deviates from this trend, abnormal degradation may exist. The specific process is as follows:

[0199] First, the second historical initial performance score data of the fourth historical time period (e.g., the most recent 180 days, a long-term time scale) is selected as the fitting data. The stationarity of the second historical initial performance score data is tested. If the second historical initial performance score data is non-stationary, it is transformed into stationary second historical initial performance score data through differencing before proceeding to the next steps.

[0200] Then, a time series model was used to fit the second historical initial performance score data to obtain a fitted curve, which represents the expected path of performance change.

[0201] Calculate the error (i.e., the initial performance score difference) between the actual historical initial performance score and the corresponding fitted initial performance score at each time step. This yields the errors corresponding to multiple moments within the fourth historical time period. The standard deviation (i.e., the second threshold) of the errors corresponding to these multiple moments is then calculated. If the error at a certain moment Exceeding the second threshold If the value is 0, it is determined to be an abnormal fluctuation point that deviates from the fitted curve.

[0202] Statistical fluctuation detection focuses on short-term dispersion, identifying sudden and large fluctuations; trend fitting focuses on long-term patterns of change, identifying slow and cumulative deviations. Combining the two can cover abnormal fluctuations at different time scales.

[0203] Step S33: External event correlation. By correlating concurrent meteorological data and power grid events, determine whether the abnormal fluctuations are temporary abnormal fluctuations caused by external factors.

[0204] External event correlation involves correlating historical meteorological and power grid events to determine whether abnormal fluctuations are caused by external factors. If the performance fluctuations of the target distributed photovoltaic power generation system are caused by external events such as heavy rain, sandstorms, or power outages, they are considered temporary abnormal fluctuations; if no external event occurs, they are considered target-specific abnormal fluctuations. The specific process is as follows:

[0205] First, establish an event database covering both meteorological and power grid events, recording the time of occurrence, duration, and scope of impact of each event. Meteorological events should include at least heavy rain, heavy snow, sandstorms, strong radiation, and low temperatures; power grid events should include at least power outages, voltage dips, and frequency anomalies.

[0206] Then, when abnormal fluctuations are detected at the current moment through statistical fluctuation detection or trend fitting, the target external events with the same impact range that occurred at the same historical moment during the same period in the event database are retrieved.

[0207] If external events related to the target are retrieved, the causal relationship between the external events and the abnormal fluctuations is further determined. The specific process is as follows:

[0208] Calculate the rate of change in irradiance and the rate of fluctuation in effectiveness before and after the occurrence of the external event. If the absolute value of the correlation coefficient between the two is greater than the preset value, it is a strong correlation, and the abnormal fluctuation is determined to be caused by the external event and is a temporary abnormal fluctuation. If no external event of the same period is retrieved, or the absolute value of the correlation coefficient is less than the preset value, it is a weak correlation, and the abnormal fluctuation is determined to be a target abnormal fluctuation that cannot be explained by the external event.

[0209] By employing three dimensions—statistical fluctuation detection, trend fitting, and external event correlation—a complementary approach can accurately identify abnormal performance fluctuations. Statistical fluctuation detection, based on the statistical characteristics of the initial historical performance score data, defines short-term or medium-term fluctuation thresholds, enabling the identification of sudden and significant abnormal deviations. Trend fitting, through time series models, uncovers long-term performance change patterns, capturing slowly accumulating abnormal degradation and compensating for the limitations of short-term detection in identifying gradual issues. External event correlation, by matching meteorological and power grid events, clarifies the cause of abnormal fluctuations, distinguishing between temporary abnormal fluctuations caused by external environments or events and target abnormal fluctuations caused by equipment malfunctions or component aging. These three methods corroborate each other, both defining whether fluctuations exceed normal fluctuation ranges through quantitative standards and eliminating external interference through causal tracing, avoiding misjudging temporary abnormal fluctuations as target abnormal fluctuations. This ensures that the judgment of fluctuation nature is both comprehensive and accurate, guaranteeing the reliability of early warnings for abnormal performance degradation of target distributed photovoltaic power generation systems.

[0210] Step S4: In response to the existing abnormal fluctuations in the target, perform data source analysis on the multi-dimensional performance perception data to locate the cause of the abnormal fluctuations (i.e., the cause of the abnormality), optimize the evaluation parameters, and obtain the corrected performance evaluation value (i.e., the current target performance score).

[0211] Once fluctuation verification identifies abnormal fluctuations in the target that cannot be explained by environmental or external events, it is necessary to initially locate the fault or anomaly, determine the cause of the anomaly, and optimize the evaluation parameters based on the cause. This includes two parts: data source analysis and evaluation parameter optimization. The purpose of data source analysis is to find the cause of the abnormal fluctuations in the target, avoiding blind early warnings. Parameter optimization adjusts the evaluation parameters based on the cause of the anomaly, eliminating the interference of abnormal factors on the current initial performance score, obtaining a more realistic current target performance score, and providing an accurate basis for determining the anomaly warning level.

[0212] Data tracing analysis includes retrieving thermal imaging data of the components to locate hot spot areas, checking the dust accumulation data of the components to confirm the light transmittance, and analyzing the inverter efficiency curve to identify sudden drops in conversion efficiency, corresponding to three dimensions: components, external environment, and core equipment.

[0213] First, thermal imaging data of the module is retrieved to locate the hot spot area. Finding a hot spot on the thermal image (i.e., thermal imaging data) allows for preliminary identification of the problem. Further, the identified hot spot area is inspected on-site. If the hot spot area is located at the edge of the module and is obstructed by tree branches, billboards, or other objects, the cause of the anomaly is external obstruction. If the hot spot area is not significantly obstructed, but there are scratches or microcracks on the module surface, the cause of the anomaly is microcracks in the solar cells. If the hot spot area is concentrated near the module junction box, the cause of the anomaly is poor contact in the junction box leading to localized overheating.

[0214] Secondly, the transmittance is confirmed by checking the dust accumulation data of the components. Historical data on the transmittance of dust accumulation on the component surface is analyzed. If a continuous downward trend in transmittance is found recently, and it is highly correlated with the downward trend of the initial performance score, then component dust accumulation is determined to be a potential cause of abnormal fluctuations in the target. Furthermore, a portable transmittance meter is taken to the site to perform repeated measurements on different areas of the component surface. If the measurement results are consistent with historical data, then component dust accumulation is confirmed as the cause of the abnormality. If the transmittance measured on-site does not match the historical data, then it is necessary to investigate whether the transmittance meter is faulty or the image recognition algorithm is biased.

[0215] Finally, the inverter efficiency curve is analyzed to identify sudden drops in conversion efficiency. The inverter's efficiency-load curve is plotted within the time period corresponding to the target abnormal fluctuation and compared with the historical normal curve. If the efficiency curve is found to shift downward as a whole or to drop sharply in a specific load range, the cause of the abnormality may be an abnormality in the inverter itself or its DC-side input.

[0216] Optimization of evaluation parameters includes adjusting the component health index, environmental degradation coefficient, or switching evaluation model (i.e., a model based on a preset fusion formula used to determine the current initial performance score of the target distributed photovoltaic power generation system). Adjusting the component health index is applicable when the anomaly is due to a problem with the component itself (i.e., there are abnormal components within the target distributed photovoltaic power generation system due to their own failures). This parameter reflects the performance status of a component, with a value ranging from 0 to 1, where 0 indicates the component is unusable and 1 indicates it is in good condition. In the initial evaluation model, the component health index is set to 1 by default. A pre-defined grading standard can be set based on the impact of component anomalies on power. When an abnormal component is present, the grading standard is adjusted according to the severity of the problem. For example, the grading criteria can be set as follows:

[0217] Minor faults (hot spot area <5cm², microcrack in a single cell): =0.9-0.95;

[0218] Moderate fault (hot spot area) (5-15cm², multiple battery cells with microcracks) =0.8-0.89;

[0219] Severe faults (hot spot area > 15cm², module encapsulation layer aging): =0.7-0.79;

[0220] Extremely severe fault (component output power < 50% of rated power): <0.7.

[0221] Component issues primarily affect indicators such as actual power generation and string current in the basic performance data. Therefore, it is necessary to correct the basic performance score using the component's health index to obtain the corrected performance score. The environmental disturbance coefficient and time-series dynamic data are unaffected by the component health index and therefore remain unchanged. Based on the corrected performance score, the corrected performance evaluation value can be calculated by substituting it back into the fusion formula. .

[0222] Adjusting the environmental degradation coefficient is applicable when the anomaly is due to abnormal components within the target distributed photovoltaic power generation system caused by environmental factors. The environmental degradation coefficient is a parameter that corrects for environmental interference; it is set to 1 by default when calculating the initial performance evaluation value. A grading standard can be pre-set based on the degree of environmental anomaly; when environmental factors exceed the normal range, the environmental degradation coefficient is adjusted. This enhances the environmental disturbance coefficient's response to abnormal environments. For example, the grading standard could be set as follows:

[0223] Minor abnormalities (transmittance) 70%-80%, local temperature 45-50℃): =1.1-1.2;

[0224] Moderate abnormality (transmittance) 60%-69%, local temperature 51-55℃): =1.3-1.4;

[0225] Severe abnormalities (transmittance <60%, local temperature >55℃): =1.5-1.6.

[0226] The local temperature is the target temperature of the abnormal component, which is obtained based on the component temperature data of the abnormal component at the current moment.

[0227] Corrected environmental interference coefficient (i.e., corrected interference coefficient) At this point, the basic performance score and time-series dynamic data remain unchanged. Based on the corrected interference coefficient, the corrected performance evaluation value can be calculated by substituting it back into the fusion formula. .

[0228] Switching to a different evaluation model is suitable when the causes of anomalies are complex or when the initial evaluation model cannot accurately quantify the impact of anomalies. The initial evaluation model uses a linear fusion formula, suitable for single, linear performance-influencing factors. When nonlinear effects exist, it is necessary to switch to a nonlinear evaluation model, such as a machine learning-based model, and use the new model to calculate the corrected performance evaluation value. .

[0229] The corrected performance evaluation value obtained by optimizing the above evaluation parameters has removed the influence of the identified anomalies and can better reflect the remaining, unexplained degradation risks or the true health status of the target distributed photovoltaic power generation system.

[0230] Step S5: Based on the corrected performance evaluation value, trigger the corresponding level of abnormality warning signal according to the degree of abnormality.

[0231] Step S51: The corrected performance evaluation value is compared with a number of preset dynamic thresholds (including a third threshold, a fourth threshold, and a fifth threshold). These multiple dynamic thresholds are generated based on historical target performance score data for the same period in history and future predicted meteorological data for the area where the target distributed photovoltaic power generation system is located.

[0232] The dynamic threshold generation logic integrates historical patterns and future expectations, enabling the anomaly early warning benchmark to adapt to seasonal changes, normal equipment aging patterns, and future weather conditions. This avoids false alarms during rainy weather or missed alarms during periods of high radiation. The specific dynamic threshold generation method is as follows:

[0233] First, obtain historical target performance score data for the target distributed photovoltaic power generation system from historical dates corresponding to the current time over the past few years. Then, perform statistical analysis on these historical target performance score data and calculate its second mean. Second standard deviation Simultaneously, based on the annual degradation rate f provided by the equipment manufacturer or obtained through long-term data analysis, and combined with the service life of the target distributed photovoltaic power generation system, the second mean is adjusted for degradation to obtain the theoretical expected performance benchmark value at the current moment. , where n is the number of years of use.

[0234] Then, based on the predicted irradiance and ambient temperature over a future period (i.e., predicted meteorological data), the expected performance correction coefficient under future weather conditions is calculated. ;

[0235] Combining baselines and meteorological corrections, a set of dynamic thresholds is generated, typically including a attention threshold (i.e., the third threshold), a warning threshold (i.e., the fourth threshold), and an emergency threshold (i.e., the fifth threshold). The attention threshold... This threshold is slightly lower than the historical benchmark level for the same period after weather correction. Exceeding this threshold only indicates that the performance of the target distributed photovoltaic power generation system has begun to deviate from the normal expected range, suggesting that attention is needed; Warning threshold This threshold is significantly lower than the benchmark, indicating that the performance degradation of the target distributed photovoltaic power generation system is unlikely to be caused solely by normal fluctuations or weather factors, and there is a high probability of equipment malfunction or performance degradation; Emergency threshold This threshold is far below the benchmark, indicating that the performance of the target distributed photovoltaic power generation system has deteriorated significantly and may be accompanied by major failures or safety hazards.

[0236] Step S52: Based on the comparison results, trigger a gradient anomaly warning signal associated with the efficiency degradation rate and potential power generation loss of the target distributed photovoltaic power generation system.

[0237] The tiered early warning system can be divided into Level 1, Level 2, and Level 3 (i.e., warning levels). When the performance evaluation value is corrected... Below the attention threshold However, it is above the warning threshold. When a Level 1 warning is triggered, the warning signal may include the abnormal photovoltaic unit number, corrected assessment value, degradation rate, potential monthly power generation loss, identified cause of the anomaly, and recommended measures. The warning signal is pushed to the regional operation and maintenance manager via SMS and the operation and maintenance platform APP (application). When the corrected performance assessment value... Below the warning threshold But above the emergency threshold When a Level 2 warning is triggered, the warning signal may include the abnormal photovoltaic unit number, corrected performance assessment value, degradation rate, potential monthly power generation loss, and recommended measures. In addition to being sent to the operation and maintenance manager, the warning signal must also be copied to the regional supervisor, who must track the progress of the handling. When the corrected performance assessment value... When the value falls below the emergency threshold, a Level 3 warning is triggered. The warning signal may include the abnormal photovoltaic unit number, the corrected assessment value, the degradation rate, the potential monthly power generation loss, the specific cause of the abnormality, and emergency measures. In addition to being pushed to the operation and maintenance manager and the regional supervisor, the warning signal must also be reported to the operation and maintenance headquarters. The headquarters will arrange technical experts to provide remote guidance to ensure the rapid restoration of the efficiency of the target distributed photovoltaic power generation system.

[0238] In addition, a closed-loop management mechanism needs to be established after the warning signal is triggered. After the operation and maintenance personnel have finished handling the issue, multi-dimensional performance perception data should be collected again, and a new corrected performance evaluation value should be calculated. If the corrected performance evaluation value rises back to the attention threshold, the system will take action. If the above conditions are met, the warning will be marked as lifted in the operation and maintenance platform. If the corrected performance evaluation value does not recover, the data source analysis needs to be carried out again, and the handling measures need to be adjusted until the performance of the target distributed photovoltaic power generation system returns to normal, ensuring that the abnormal degradation is completely resolved and avoiding repeated warnings.

[0239] The above-mentioned optional implementation methods achieve at least the following effects: by acquiring multi-dimensional performance perception data including basic performance data, environmental interference data, and time-series dynamic data, and combining this with fluctuation verification to accurately identify anomalies that cannot be explained by the environment or external events, false alarms that misjudge normal environmental impacts as equipment anomalies in the target distributed photovoltaic power generation system can be reduced; at the same time, by using data source tracing analysis, the causes of abnormal fluctuations can be accurately located, and evaluation parameters can be optimized to generate more accurate corrected performance evaluation values, avoiding the omission of gradual performance degradation and improving the accuracy of anomaly early warning; finally, by triggering corresponding level early warning signals based on the degree of anomaly, accurate and reliable basis can be provided for operation and maintenance decisions, thereby ensuring the power generation efficiency and stable operation of the target distributed photovoltaic power generation system.

[0240] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0241] This embodiment also provides an anomaly early warning device for a distributed photovoltaic power generation system. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the terms "module" and "device" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0242] According to an embodiment of this application, an apparatus embodiment for implementing an anomaly early warning method for a distributed photovoltaic power generation system is also provided. Figure 3 This is a schematic diagram of an abnormality early warning device for a distributed photovoltaic power generation system according to an embodiment of this application, such as... Figure 3 As shown, the above-mentioned abnormal early warning device for distributed photovoltaic power generation system includes a data acquisition module 302, a first determination module 304, a detection module 306, a second determination module 308, and an abnormal early warning module 310. The device will be described below.

[0243] The data acquisition module 302 is used to acquire the current performance data, current environmental disturbance data and performance index data of the target distributed photovoltaic power generation system at the current moment. The current environmental disturbance data refers to the environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change of the performance of the target distributed photovoltaic power generation system over time.

[0244] The first determining module 304, connected to the data acquisition module 302, is used to determine the current initial performance score of the target distributed photovoltaic power generation system at the current moment based on the current performance data, current environmental interference data, and performance index data.

[0245] The detection module 306, connected to the first determination module 304, is used to detect whether there is any abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment, based on the current initial performance score.

[0246] The second determining module 308, connected to the detection module 306, is used to determine the cause of the abnormality when the efficiency of the target distributed photovoltaic power generation system is detected to be abnormally fluctuating, and to correct the current initial efficiency score based on the cause of the abnormality, so as to obtain the current target efficiency score of the target distributed photovoltaic power generation system at the current moment.

[0247] The anomaly warning module 310, connected to the second determination module 308, is used to provide anomaly warnings for the target distributed photovoltaic power generation system based on the current target performance score.

[0248] This application provides an anomaly warning device for a distributed photovoltaic power generation system. By setting up a data acquisition module 302, a first determination module 304, a detection module 306, a second determination module 308, and an anomaly warning module 310, the device aims to determine the current target performance score of the target distributed photovoltaic power generation system by analyzing current performance data, current environmental interference data, and performance index data. Based on the current target performance score, it provides an anomaly warning for the target distributed photovoltaic power generation system, thereby improving the accuracy of anomaly warnings and solving the technical problem of inaccurate anomaly warnings in distributed photovoltaic power generation systems in related technologies.

[0249] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0250] It should be noted that the data acquisition module 302, the first determination module 304, the detection module 306, the second determination module 308, and the anomaly warning module 310 mentioned above correspond to steps S102 to S110 in the embodiments. The instances and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in the above embodiments. It should be noted that the above modules, as part of the device, can run in a computer terminal.

[0251] It should be noted that the optional or preferred implementation methods of this embodiment can be found in the relevant descriptions in the embodiments, and will not be repeated here.

[0252] The aforementioned abnormal early warning device for distributed photovoltaic power generation system may also include a processor and a memory. The data acquisition module 302, the first determination module 304, the detection module 306, the second determination module 308, the abnormal early warning module 310, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0253] The processor contains a core that retrieves the corresponding program unit from memory. One or more cores may be configured. Memory may include non-persistent memory in computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory includes at least one memory chip.

[0254] This application provides a non-volatile storage medium storing a program that, when executed by a processor, implements an anomaly warning method for a distributed photovoltaic power generation system.

[0255] This application provides an electronic device including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring current performance data, current environmental interference data, and performance index data of a target distributed photovoltaic (PV) power generation system at the current moment. The current environmental interference data refers to environmental data affecting the performance of the target PV power generation system, and the performance index data describes the change in the performance of the target PV power generation system over time. Based on the current performance data, current environmental interference data, and performance index data, it determines the current initial performance score of the target PV power generation system at the current moment. Based on the current initial performance score, it detects whether there is any abnormal fluctuation in the performance of the target PV power generation system at the current moment. If abnormal fluctuations in the performance of the target PV power generation system are detected, it determines the cause of the abnormality and corrects the current initial performance score based on the cause of the abnormality to obtain the current target performance score of the target PV power generation system at the current moment. Based on the current target performance score, it provides an anomaly warning for the target PV power generation system. The device in this document can be a server, PC, etc.

[0256] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: acquiring current performance data, current environmental interference data, and performance index data of a target distributed photovoltaic power generation system at the current moment, wherein the current environmental interference data refers to environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change in the performance of the target distributed photovoltaic power generation system over time; determining the current initial performance score of the target distributed photovoltaic power generation system at the current moment based on the current performance data, current environmental interference data, and performance index data; detecting whether there is a target abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment based on the current initial performance score; determining the cause of the abnormality if the performance of the target distributed photovoltaic power generation system is detected to be abnormal, and correcting the current initial performance score based on the cause of the abnormality to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment; and issuing an abnormality warning for the target distributed photovoltaic power generation system based on the current target performance score.

[0257] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0258] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0259] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0260] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0261] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0262] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0263] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0264] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0265] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0266] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An anomaly early warning method for a distributed photovoltaic power generation system, characterized in that, include: Acquire the current performance data, current environmental disturbance data, and performance index data of the target distributed photovoltaic power generation system at the current moment. The current environmental disturbance data refers to environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change of the performance of the target distributed photovoltaic power generation system over time. Based on the current performance data, the current environmental disturbance data, and the performance index data, determine the current initial performance score of the target distributed photovoltaic power generation system at the current moment; Based on the current initial performance score, detect whether there is any abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment; If an abnormal fluctuation in the efficiency of the target distributed photovoltaic power generation system is detected, the cause of the abnormality is determined, and the current initial efficiency score is corrected based on the cause of the abnormality to obtain the current target efficiency score of the target distributed photovoltaic power generation system at the current time. Based on the current target performance score, an anomaly warning is issued for the target distributed photovoltaic power generation system.

2. The method according to claim 1, characterized in that, The process of determining the initial performance score of the target distributed photovoltaic power generation system at the current moment based on the current performance data, the current environmental disturbance data, and the performance index data includes: Based on the current performance data, a basic performance score is determined; Based on the current environmental interference data, an environmental interference coefficient is determined, wherein the environmental interference coefficient is used to quantify the degree of impact of the current environmental interference data on the efficiency of the target distributed photovoltaic power generation system; Based on the efficiency change rate, efficiency decay rate, and year-on-year efficiency deviation in the efficiency index data, a weighted calculation method is used to obtain the efficiency index score. The efficiency change rate is used to quantify the degree of change in the efficiency of the target distributed photovoltaic power generation system at the current time compared with the first historical time period. The efficiency decay rate is used to quantify the degree of decay in the efficiency of the target distributed photovoltaic power generation system at the current time compared with the second historical time period. The year-on-year efficiency deviation is used to quantify the degree of difference in the efficiency of the target distributed photovoltaic power generation system at the current time compared with the historical time period at the current time. The current initial performance score is determined based on the basic performance score, the environmental interference coefficient, and the performance index score.

3. The method according to claim 1, characterized in that, The step of detecting whether there is an abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current time, based on the current initial performance score, includes: Based on the first historical initial performance score data of the target distributed photovoltaic power generation system in the third historical time period, and the current initial performance score, a first fluctuation detection result is determined, wherein the first fluctuation detection result is used to indicate whether there is short-term or medium-term fluctuation in the performance of the target distributed photovoltaic power generation system. Based on the second historical initial performance score data of the target distributed photovoltaic power generation system in the fourth historical time period, and the current initial performance score, a second fluctuation detection result is determined, wherein the second fluctuation detection result is used to indicate whether there is long-term fluctuation in the performance of the target distributed photovoltaic power generation system, and the fourth historical time period is longer than the third historical time period; Based on the first fluctuation detection result and the second fluctuation detection result, it is determined whether there is any abnormal fluctuation in the efficiency of the target distributed photovoltaic power generation system at the current moment.

4. The method according to claim 3, characterized in that, The determination of the first fluctuation detection result based on the first historical initial performance score data of the target distributed photovoltaic power generation system in the third historical time period and the current initial performance score includes: Statistical analysis is performed on the first historical initial performance score data to obtain the first mean and first standard deviation of the first historical initial performance score data; Based on the first mean and the first standard deviation, determine the energy efficiency volatility and the first threshold; Based on the current initial performance score, the energy efficiency volatility, and the first threshold, a first volatility detection result is determined.

5. The method according to claim 3, characterized in that, The determination of the second fluctuation detection result based on the second historical initial performance score data of the target distributed photovoltaic power generation system in the fourth historical time period and the current initial performance score includes: The second historical initial performance score data is fitted to obtain a fitted curve; Based on the fitted curve, the initial performance score of the target distributed photovoltaic power generation system at the current time is determined; Based on the current initial performance score and the fitted initial performance score, the difference in initial performance scores is determined; The second fluctuation detection result is determined based on the initial performance score difference and the second threshold.

6. The method according to claim 1, characterized in that, The process of correcting the initial performance score based on the cause of the anomaly to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment includes: When the cause of the anomaly is that there is an abnormal component within the target distributed photovoltaic power generation system due to its own component failure, the health index of the abnormal component is determined based on the hot spot area and output power of the abnormal component at the current moment, wherein the hot spot area is obtained based on the thermal imaging data of the abnormal component at the current moment; Based on the aforementioned health index, the basic performance score is corrected to obtain the corrected performance score. Based on the revised performance score, the current initial performance score is revised to obtain the current target performance score.

7. The method according to claim 1, characterized in that, The process of correcting the initial performance score based on the cause of the anomaly to obtain the current target performance score of the target distributed photovoltaic power generation system at the current moment includes: When the cause of the anomaly is that there is an abnormal component inside the target distributed photovoltaic power generation system due to environmental factors, the environmental attenuation coefficient is determined based on the light transmittance and target temperature of the abnormal component at the current time. The light transmittance is obtained based on the dust accumulation data of the abnormal component at the current time, and the target temperature is obtained based on the component temperature data of the abnormal component at the current time. Based on the environmental attenuation coefficient, the environmental interference coefficient is corrected to obtain the corrected interference coefficient; Based on the corrected interference coefficient, the current initial performance score is corrected to obtain the current target performance score.

8. The method according to any one of claims 1 to 7, characterized in that, The step of providing anomaly warnings for the target distributed photovoltaic power generation system based on the current target performance score includes: Obtain historical target performance score data of the target distributed photovoltaic power generation system, as well as predicted meteorological data of the area where the target distributed photovoltaic power generation system is located; Statistical analysis was performed on the historical target effectiveness score data to obtain the second mean and second standard deviation of the historical target effectiveness score data; Based on the second mean, and the annual decay rate and service life of the target distributed photovoltaic power generation system, the expected performance benchmark value of the target distributed photovoltaic power generation system at the current moment is determined. Based on the predicted meteorological data, determine the effectiveness correction coefficient; Based on the second standard deviation, the expected performance benchmark value, and the performance correction coefficient, the third threshold, the fourth threshold, and the fifth threshold are determined. Based on the current target performance score, the third threshold, the fourth threshold, and the fifth threshold, the warning level of the target distributed photovoltaic power generation system at the current moment is determined, and an abnormal warning is issued.

9. An anomaly early warning device for a distributed photovoltaic power generation system, characterized in that, include: The data acquisition module is used to acquire the current performance data, current environmental disturbance data, and performance index data of the target distributed photovoltaic power generation system at the current moment. The current environmental disturbance data refers to environmental data that affects the performance of the target distributed photovoltaic power generation system, and the performance index data is used to describe the change of the performance of the target distributed photovoltaic power generation system over time. The first determining module is used to determine the current initial performance score of the target distributed photovoltaic power generation system at the current moment based on the current performance data, the current environmental interference data, and the performance index data. The detection module is used to detect, based on the current initial performance score, whether there is any abnormal fluctuation in the performance of the target distributed photovoltaic power generation system at the current moment; The second determining module is used to determine the cause of the abnormality when the efficiency of the target distributed photovoltaic power generation system is detected to be abnormally fluctuating, and to correct the current initial efficiency score based on the cause of the abnormality to obtain the current target efficiency score of the target distributed photovoltaic power generation system at the current time. An anomaly warning module is used to provide anomaly warnings for the target distributed photovoltaic power generation system based on the current target performance score.

10. An electronic device, characterized in that, include: One or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the abnormal early warning method for a distributed photovoltaic power generation system according to any one of claims 1 to 8.

Images