Inverter low-efficiency diagnosis method, device, equipment and medium of photovoltaic system

Abstract

The invention discloses a low-efficiency diagnosis method, device and equipment for an inverter of a photovoltaic system and a storage medium. The method comprises the following steps: acquiring real-time equipment data and meteorological data of a to-be-diagnosed time period; estimating the working temperature and the dust attenuation factor of the photovoltaic module as key features; inputting the real-time data and the features into a trained generating capacity prediction model to obtain theoretical generating capacity; calculating the deviation ratio with the actual generating capacity, and automatically judging whether the inverter is in a low-efficiency operation state or not according to a pre-configured identification condition; and when low efficiency is identified, fault attribution is carried out by combining the internal state of the inverter and the direct current side parameters. According to the method, the accuracy and reliability of inverter low-efficiency diagnosis are remarkably improved, additional hardware is not needed, and the method is suitable for intelligent operation and maintenance of a photovoltaic power station.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation technology, and in particular to a method, apparatus, equipment and medium for diagnosing inefficiency in photovoltaic system inverters. Background Technology

[0002] In photovoltaic (PV) power generation systems, the inverter, as the core device that converts the direct current (DC) generated by PV modules into alternating current (AC), directly impacts the overall power generation revenue and operational reliability of the system. As PV power plants expand in scale and age, inverters may experience efficiency reductions due to factors such as aging internal components, decreased heat dissipation, and control algorithm drift, leading to hidden power generation losses. Therefore, timely and accurate diagnosis of inefficient inverter operation has become a key technical issue for the refined operation and maintenance of PV power plants and for improving economic efficiency.

[0003] Currently, common methods for diagnosing inverter inefficiency in the industry mainly fall into two categories: threshold-based alarms and theoretical model comparisons. Threshold-based alarms primarily trigger alarms by monitoring whether inverter operating parameters (such as DC voltage, AC current, and internal temperature) exceed preset fixed thresholds. While simple and direct, this method is insensitive to slow efficiency degradation in inverters, struggles to identify long-term power generation losses caused by slight efficiency declines, and its threshold setting relies on experience, making it prone to false alarms or missed alarms. Theoretical model comparisons, on the other hand, typically use simulation software to calculate the system's theoretical power generation and then compare it with the actual value. While this method has a certain theoretical basis, its accuracy heavily depends on precise and static system parameter inputs, making it difficult to adapt to the dynamic changes in the environment and equipment status during actual power plant operation. Therefore, the model is prone to inaccuracy, resulting in insufficient reliability of the diagnostic results.

[0004] It is evident that existing technologies still suffer from insufficient reliability in diagnosing inefficient inverters. Summary of the Invention

[0005] The technical problem to be solved by this invention is the low reliability of existing inefficient inverter diagnostic methods.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for diagnosing inefficiency of photovoltaic system inverters, comprising: Acquire real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed; Based on the real-time meteorological data and real-time equipment data, the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module are estimated. The real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor are input into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed. Based on the theoretical power generation and the actual power generation of the inverter during the period to be diagnosed, the power generation deviation rate is calculated, and according to the pre-configured inefficient operation identification conditions, the inverter is identified as being in an inefficient operation state based on the power generation deviation rate. When the inverter is detected to be in an inefficient operating state, a fault attribution analysis is performed based on the inverter's internal state parameters and DC-side operating parameters.

[0007] Furthermore, the estimation of the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data includes: The horizontal irradiance in the real-time meteorological data is converted into the inclined irradiance of the photovoltaic module through a radiation conversion model; Based on the slope irradiance, the ambient temperature and wind speed in the real-time meteorological data, and the photovoltaic module installation tilt angle in the real-time equipment data, the real-time operating temperature characteristics of the photovoltaic module are estimated through a thermal balance model. Based on the rainfall records and consecutive rainless days in the real-time meteorological data, and combined with the seasonal empirical coefficient, the real-time dust attenuation factor is calculated.

[0008] Furthermore, the inefficient operation identification condition is a continuous verification condition based on a preset static deviation threshold. The step of identifying whether the inverter is in an inefficient operation state based on the power generation deviation rate according to the pre-configured inefficient operation identification condition includes: If the daily power generation deviation rate exceeds the preset static deviation threshold, the day is marked as a suspected inefficient state. If the power generation deviation rate exceeds the preset static deviation threshold for N consecutive days, the inverter is confirmed to have entered an inefficient operating state, where N is an integer greater than 1.

[0009] Furthermore, the fault attribution analysis based on the inverter's internal state parameters and DC-side operating parameters includes: Query the internal fault information of the inverter; If no efficiency-related fault information is found, analyze the consistency of the DC-side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter. Based on the query results of the internal fault information or the consistency analysis results of the DC side electrical parameters, the corresponding fault attribution conclusion is output.

[0010] Furthermore, the analysis of the consistency of the DC-side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter includes: Calculate the dispersion rate of DC voltage or DC current for each MPPT branch; When the dispersion rate exceeds the second threshold, it is determined that the DC side electrical parameters are inconsistent.

[0011] Furthermore, the power generation prediction model is obtained through a training step, which includes: Acquire historical equipment data, historical power generation data, and corresponding historical meteorological data for the photovoltaic system; Based on the historical meteorological data and the historical equipment data, the historical operating temperature characteristics and historical dust degradation factor of the photovoltaic module are dynamically estimated. The historical equipment data, historical power generation data, historical meteorological data, historical operating temperature characteristics, and historical dust attenuation factor are fused to obtain a fused dataset. The power generation prediction model was trained based on the fused dataset.

[0012] Furthermore, after acquiring the historical equipment data, historical power generation data, and corresponding historical meteorological data of the photovoltaic system, the method further includes: The historical equipment data, historical power generation data, and historical meteorological data are cleaned to remove abnormal data items; The historical equipment data, historical power generation data, and historical meteorological data were aligned by time after cleaning.

[0013] The present invention also provides a device for diagnosing inverter inefficiency in a photovoltaic system, comprising: The real-time data acquisition module is used to acquire real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed. The physical characteristic estimation module is used to estimate the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data. The theoretical power generation prediction module is used to input the real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed. The inefficiency identification module is used to calculate the power generation deviation rate based on the theoretical power generation and the actual power generation of the inverter during the period to be diagnosed, and to identify whether the inverter is in an inefficient operating state based on the power generation deviation rate according to the pre-configured inefficient operation identification conditions. The fault attribution module is used to perform fault attribution analysis based on the inverter's internal state parameters and DC-side operating parameters when the inverter is identified as being in an inefficient operating state.

[0014] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the inverter inefficiency diagnosis method for a photovoltaic system as described above.

[0015] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the inverter inefficiency diagnosis method for a photovoltaic system as described above.

[0016] The beneficial effects of this invention are as follows: By acquiring real-time equipment data and real-time meteorological data of the photovoltaic system, and estimating the real-time operating temperature characteristics and real-time dust attenuation factor of the photovoltaic modules, and by inputting the real-time equipment data, real-time meteorological data, real-time operating temperature characteristics, and the real-time dust attenuation factor into a trained power generation prediction model, the theoretical power generation of the inverter during the period to be diagnosed is obtained. Then, the deviation rate between the theoretical power generation and the actual power generation is calculated. Based on the power generation deviation rate, according to the pre-configured inefficient operation identification conditions, it is identified whether the inverter is in an inefficient operation state. The addition of real-time operating temperature characteristics and real-time dust attenuation factor can effectively improve the accuracy and early identification capability of inverter inefficiency diagnosis, significantly reduce false alarms and missed alarms caused by a single threshold or fixed model, and improve the reliability of diagnosis. At the same time, no additional hardware deployment is required, and reliable diagnosis can be achieved solely based on existing data, which greatly improves the intelligence level and economy of photovoltaic power plant operation and maintenance. Attached Figure Description

[0017] The specific structure of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart of a photovoltaic system inverter inefficiency diagnosis method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the inverter inefficiency detection method according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating the alarm classification and work order generation process according to an embodiment of the present invention. Figure 4 This is a block diagram of an inverter inefficiency diagnostic device for a photovoltaic system according to an embodiment of the present invention. Figure 5 This is a schematic block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0021] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0023] like Figure 1 As shown, an embodiment of the present invention is: a method for diagnosing inefficiency in a photovoltaic system inverter, comprising steps S1-S5: S1. Obtain real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed.

[0024] In this step, real-time equipment data is the equipment data acquired during the diagnostic phase. Equipment data may include equipment configuration data and operating parameters. Equipment configuration data may include at least one of the following: inverter rated power, installation tilt angle, number of strings, and number of modules. Operating parameters may include at least one of the following: equipment age, grid connection voltage, grid frequency, and output power factor. Real-time historical meteorological data must be obtained from a meteorological bureau or commercial weather API and must match the power plant's geographical location. The data should include at least the following: total horizontal radiation intensity, ambient temperature, wind speed, wind direction angle, diffuse radiation intensity, relative humidity, precipitation, air pressure, and cloud cover.

[0025] S2. Based on the real-time meteorological data and real-time equipment data, estimate the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module.

[0026] In this embodiment, estimating the component operating temperature and dust attenuation factor during this period helps to improve the prediction accuracy of the power generation prediction model.

[0027] In a specific embodiment, estimating the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data includes steps S21-S23: S21. The horizontal irradiance in the real-time meteorological data is converted into the inclined irradiance of the photovoltaic module through a radiation conversion model.

[0028] In this step, meteorological data typically provides the total radiation intensity on a horizontal plane. However, since photovoltaic modules are installed at a certain tilt angle, the irradiance they receive (tilted irradiance) differs from the horizontal irradiance. This embodiment can employ a radiation conversion model, such as the Hay model, to accurately convert the horizontal irradiance into the tilted irradiance of the modules based on the power plant's geographical location, date and time, and the module's installation tilt angle, thereby more realistically reflecting the energy source driving power generation.

[0029] S22. Based on the slope irradiance, the ambient temperature and wind speed in the real-time meteorological data, and the photovoltaic module installation tilt angle in the real-time equipment data, the real-time operating temperature characteristics of the photovoltaic module are estimated using a thermal balance model.

[0030] In this step, since the operating temperature of photovoltaic (PV) modules significantly affects their power generation efficiency, a simplified thermal balance model is used to dynamically estimate the operating temperature of the PV modules. This model comprehensively considers the temperature rise effect after the modules receive irradiance, the baseline influence of ambient temperature, and the convective cooling effect caused by wind speed. Input parameters include ambient temperature, wind speed, calculated slope irradiance, and module characteristic parameters (such as installation tilt angle, which may also affect heat exchange). The temperature estimated using this thermal balance model more accurately reflects the actual operating state of the PV modules than simply using ambient temperature, thereby improving the accuracy of the power generation prediction model.

[0031] S23. Based on the rainfall records and consecutive rainless days in the real-time meteorological data, and combined with the seasonal empirical coefficient, calculate the real-time dust attenuation factor.

[0032] In this step, dust accumulation on the surface of photovoltaic modules reduces light transmission, leading to power generation loss. This impact is quantified using a model based on physical experience. Key inputs include historical rainfall records (rain can clean photovoltaic modules) and the number of consecutive rainless days (dust accumulation duration). Simultaneously, a dust attenuation factor (e.g., 0% to 10%) is dynamically calculated, representing the proportion of irradiance reduction. This factor is introduced as a feature, enabling the power generation prediction model to automatically learn and compensate for power generation changes caused by dust, improving the model's adaptability and prediction accuracy under different environmental conditions.

[0033] S3. Input the real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed.

[0034] In this step, when a diagnosis is needed for a recent period (e.g., the past 24 hours), real-time equipment data, real-time meteorological data, real-time operating temperature characteristics, and real-time dust attenuation factors for that period must be collected. These real-time features are then combined and input into a pre-trained power generation prediction model. The predicted value output by the model is the theoretical power generation that the inverter should produce under that specific environmental and equipment condition, denoted as […]. This value serves as a benchmark for assessing whether actual performance is normal.

[0035] In a specific embodiment, the power generation prediction model is obtained through a training step, which includes steps A1-A4: A1. Obtain historical equipment data, historical power generation data, and corresponding historical meteorological data of the photovoltaic system.

[0036] In this step, historical equipment data includes data obtained from the Supervisory Control and Data Acquisition (SCADA) database. Equipment data may include equipment configuration data and operating parameters. Equipment configuration data may include at least one of the following: inverter rated power, installation tilt angle, number of strings, and number of modules. Operating parameters may include at least one of the following: equipment operating years, grid connection voltage, grid frequency, and output power factor. Historical power generation data refers to the corresponding power generation recorded by the power plant monitoring system. Historical meteorological data must be obtained from a meteorological bureau or commercial weather API and must match the geographical location of the power plant. The data should include at least the total horizontal radiation intensity, ambient temperature, wind speed, wind direction angle, diffuse radiation intensity, relative humidity, precipitation, air pressure, and cloud cover. To build a reliable power generation prediction model, historical data from the past 1-2 years that has been confirmed as being in a healthy operating state is typically required, with a collection interval of, for example, 1 hour.

[0037] In a specific embodiment, after acquiring the historical equipment data, historical power generation data, and corresponding historical meteorological data of the photovoltaic system, the method further includes steps A11-A12: A11. Perform data cleaning on the historical equipment data, the historical power generation data, and the historical meteorological data to remove abnormal data items.

[0038] In this step, the purpose of data cleaning is to remove obviously unreasonable or erroneous data points to ensure the accuracy of subsequent analysis. Specific cleaning rules include, but are not limited to: removing data points that recorded power generation during nighttime periods (when irradiance is zero); removing abnormal data where irradiance exceeds theoretical limits (such as the solar constant at the top of the atmosphere); and removing data where parameters such as temperature, voltage, and current are significantly outside the reasonable operating range of the equipment. This step effectively reduces the interference of data noise on the training of the power generation prediction model.

[0039] A12. Time-align the cleaned historical equipment data, historical power generation data, and historical meteorological data.

[0040] In this step, since historical equipment data, historical power generation data, and historical meteorological data may originate from different systems, their timestamp accuracy and collection periods may not be entirely consistent. Time alignment is the process of matching and merging all this data according to a unified time base (e.g., the hour). Ensuring that meteorological conditions at the same point in time strictly correspond to the inverter's operating status and power generation is the foundation for constructing an accurate causal relationship dataset.

[0041] A2. Based on the historical meteorological data and the historical equipment data, dynamically estimate the historical operating temperature characteristics and historical dust attenuation factor of the photovoltaic module.

[0042] In this step, physical features are introduced to enhance the accuracy of the model. The physical model transforms the raw meteorological data into features that can more directly affect the power generation process.

[0043] A3. The historical equipment data, historical power generation data, historical meteorological data, historical operating temperature characteristics, and historical dust attenuation factor are fused to obtain a fused dataset.

[0044] In this step, all data sources obtained after cleaning, alignment, and feature engineering, including historical equipment data, historical power generation data, historical meteorological data, and newly estimated component operating temperature characteristics and dust attenuation factors, are integrated chronologically into a structured, multi-dimensional dataset. This fused dataset contains environmental drivers affecting power generation, equipment status, and key physical intermediate variables, forming the complete input for training a high-precision power generation prediction model.

[0045] A4. Train the power generation prediction model based on the fused dataset.

[0046] In this step, all features in the fused dataset except for historical power generation data are used as input variables (independent variables), and the inverter's historical power generation data is used as the target variable (dependent variable) to train a machine learning regression model. The model aims to learn the mapping relationship between various complex factors and power generation under the equipment's health condition. Algorithms such as XGBoost, gradient boosting trees, random forests, or LSTM can be used. To achieve optimal performance and automation, this embodiment can utilize the AutoML function of platforms such as H2O to automatically complete model training, hyperparameter tuning, and model selection.

[0047] In a specific embodiment, training the power generation prediction model based on the fused dataset further includes: Based on a pre-configured periodic triggering mechanism, the power generation prediction model is automatically retrained and updated using newly collected equipment data, historical power generation data, and corresponding historical meteorological data.

[0048] In this embodiment, a pre-configured periodic triggering mechanism (e.g., quarterly) automatically collects newly generated equipment data, power generation data, and corresponding meteorological data. This automatically triggers a new round of model training (e.g., re-invoking H2OAutoML), using recently accumulated historical equipment data, historical power generation data, historical meteorological data, and newly estimated component operating temperature characteristics and dust degradation factors to generate an updated power generation prediction model. Before deploying the new power generation prediction model to the production environment, its performance can be verified through A / B testing or other methods to determine if it outperforms the old model. Once verified, the old model is smoothly switched to the new one. This mechanism enables the diagnostic system to adapt to long-term performance drift caused by slow component aging and environmental changes, thus maintaining diagnostic accuracy and reliability throughout a multi-year maintenance cycle.

[0049] S4. Calculate the power generation deviation rate based on the theoretical power generation and the actual power generation of the inverter during the diagnostic period, and identify whether the inverter is in an inefficient operating state based on the power generation deviation rate according to the pre-configured inefficient operation identification conditions.

[0050] In this step, the actual power generation of the inverter during the same diagnostic period is obtained from the monitoring system and recorded as follows: The relative deviation rate is calculated using the following formula: This deviation rate quantifies the percentage by which actual power generation falls short of the theoretically expected value. A larger positive value indicates a more severe potential efficiency loss.

[0051] In this step, a relatively high static deviation threshold T is set (e.g., T = 30%). This static deviation threshold is designed to accommodate reasonable prediction errors in the power generation prediction model itself, local microclimate fluctuations, and normal noise in data acquisition, thereby avoiding false alarms for small or transient efficiency fluctuations. A continuous verification process is introduced: the inverter is only considered to be in a "confirmed inefficiency" state when the deviation rate exceeds the threshold T for N consecutive valid power generation days (e.g., N = 3). If only a single day exceeds the limit, it is marked as "suspected inefficiency," and only observed and recorded without immediately triggering a high-priority alarm. This dual-criteria mechanism greatly filters out accidental interference, ensuring extremely high reliability of the generated alarms and making maintenance actions more targeted.

[0052] In a specific embodiment, such as Figure 2 As shown, the inefficient operation identification condition is a continuous verification condition based on a preset static deviation threshold. The step of identifying whether the inverter is in an inefficient operation state based on the power generation deviation rate according to the pre-configured inefficient operation identification condition includes steps S41-S42: S41. If the daily power generation deviation rate exceeds the preset static deviation threshold, then the day is marked as a suspected inefficient state.

[0053] In this step, when the calculated power generation deviation rate for a certain day exceeds a preset high threshold (such as 30%), the inverter is not immediately diagnosed as faulty. Instead, the day is marked as a "suspected inefficiency" event. This is equivalent to a sensitive preliminary screening, recording all cases that significantly deviate from the theoretical value and placing them in the observation queue.

[0054] S42. If the power generation deviation rate exceeds the preset static deviation threshold for N consecutive days, then the inverter is confirmed to have entered an inefficient operating state, where N is an integer greater than 1.

[0055] In this step, it is checked whether the "suspected inefficiency" event is persistent. Only when multiple consecutive valid power generation days (e.g., N=3 days) meet the "suspected" condition does the system move from quantitative to qualitative change, determining that the inverter is indeed in a state of "confirmed inefficiency" health degradation. This mechanism effectively eliminates occasional deviations caused by single-day severe weather, temporary obstruction, or data anomalies, ensuring that alarm signals point to persistent, real efficiency problems, thereby reducing the false alarm rate to an extremely low level.

[0056] S5. When the inverter is found to be in an inefficient operating state, a fault attribution analysis is performed based on the inverter's internal state parameters and DC side operating parameters.

[0057] In this step, once the inverter is identified as "confirmed inefficiency," a fault attribution process is automatically initiated to guide maintenance. This process is entirely based on existing data and requires no additional hardware.

[0058] In a specific embodiment, such as Figure 3 As shown, the fault attribution analysis based on the inverter's internal state parameters and DC-side operating parameters includes steps S51-S53: S51. Query the internal fault information of the inverter.

[0059] In this step, the inverter typically has a built-in fault diagnosis system that records events such as IGBT overheating, heat dissipation failure, and communication anomalies in its internal fault register or alarm status table. First, this information is queried. If a fault code directly related to power generation efficiency is found, it can be quickly attributed to a hardware or control fault in the inverter itself, and a corresponding work order is generated.

[0060] S52. If no fault information related to efficiency is found, analyze the consistency of the DC side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter.

[0061] In this step, if the inverter itself does not report a clear fault, the problem may lie at the DC input. Analyze the DC voltage and current data for each MPPT branch. Under uniform illumination and consistent photovoltaic module conditions, the voltage / current of each branch should be relatively close. The degree of inconsistency can be quantified by calculating their dispersion rate (such as the ratio of standard deviation to mean).

[0062] In a specific embodiment, the analysis of the consistency of the DC-side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter includes the following steps: S521. Calculate the dispersion rate of DC voltage or DC current for each MPPT branch.

[0063] This step involves calculating the dispersion of the DC voltage (or current) values ​​for each branch, for example, by calculating its coefficient of variation (standard deviation divided by the mean). A higher dispersion indicates a greater difference in the operating points of each branch.

[0064] S522. When the dispersion rate exceeds the second threshold, it is determined that the DC side electrical parameters are inconsistent.

[0065] In this step, an empirical threshold is set for the dispersion rate. When the calculated dispersion rate exceeds this threshold, a significant inconsistency is determined to exist on the DC side. This usually indicates string mismatch, severe local shading, individual component failure, or connection problems, causing some MPPTs to fail to operate optimally, thus reducing the overall inverter efficiency. This can be attributed to an anomaly on the DC side.

[0066] S53. Based on the query results of the internal fault information or the consistency analysis results of the DC side electrical parameters, output the corresponding fault attribution conclusion.

[0067] In this step, if a fault code is found in step S51, "Inverter body fault" is output; if no fault is found in step S51 but step S52 determines that the DC side is inconsistent, "DC side abnormal" is output; if neither is abnormal, "May be affected by external environmental factors (such as complex clouds, uneven dust, etc., factors not fully captured in the features)" is output, and manual on-site verification is recommended. This three-level diagnostic classification mechanism can effectively distinguish the root cause of the problem and greatly improve the efficiency of operation and maintenance troubleshooting.

[0068] The aforementioned methods for diagnosing inverter inefficiency in photovoltaic systems offer significant advantages over traditional methods. Traditional methods, besides threshold-based alarms and theoretical model comparisons, have explored data-driven approaches to improve diagnostic accuracy. For example, some studies use operational data collected by Supervisory Control and Data Acquisition (SCADA) systems to train machine learning models for condition identification. However, these methods often rely on a single SCADA data source, resulting in limited data dimensions and significant noise interference. Their diagnostic capabilities are particularly vulnerable in the early stages of inverter efficiency degradation and when it is only slight. Other approaches attempt to comprehensively assess energy efficiency by introducing more physical features (such as module temperature and dust coverage) and combining them with infrared imaging and other detection methods. However, these methods typically target the entire power plant, making it difficult to pinpoint inverter-specific efficiency issues. Furthermore, they heavily rely on hardware deployments such as infrared cameras and additional sensors, leading to high implementation costs and hindering large-scale deployment in existing power plants. They also fail to effectively address the high false alarm rate prevalent in engineering practice.

[0069] The aforementioned inverter inefficiency diagnosis method for photovoltaic systems acquires real-time equipment data, real-time meteorological data, and estimates the real-time operating temperature characteristics and real-time dust attenuation factor of the photovoltaic modules. This data is then input into a trained power generation prediction model to obtain the theoretical power generation of the inverter during the diagnostic period. The deviation rate between the theoretical and actual power generation is calculated, and based on the pre-configured inefficiency operation identification conditions and the power generation deviation rate, the method identifies whether the inverter is in an inefficient operating state. The inclusion of real-time operating temperature characteristics and the real-time dust attenuation factor effectively improves the accuracy and early identification capability of inverter inefficiency diagnosis, significantly reducing false alarms and missed alarms caused by single thresholds or fixed models, thus improving diagnostic reliability. Furthermore, no additional hardware deployment is required; reliable diagnosis can be achieved solely based on existing data, significantly enhancing the intelligence and economy of photovoltaic power plant operation and maintenance.

[0070] like Figure 4As shown, this embodiment of the invention also provides a photovoltaic system inverter inefficiency diagnostic device, comprising: Real-time data acquisition module 10 is used to acquire real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed. The physical characteristic estimation module 20 is used to estimate the real-time operating temperature characteristics and real-time dust attenuation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data. The theoretical power generation prediction module 30 is used to input the real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed. The inefficiency identification module 40 is used to calculate the power generation deviation rate based on the theoretical power generation and the actual power generation of the inverter during the period to be diagnosed, and to identify whether the inverter is in an inefficient operating state based on the power generation deviation rate according to the pre-configured inefficient operation identification conditions. The fault attribution module 50 is used to perform fault attribution analysis based on the inverter's internal state parameters and DC side operating parameters when the inverter is identified as being in an inefficient operating state.

[0071] In a specific embodiment, the physical feature estimation module 20 includes: The irradiance conversion unit is used to convert the horizontal irradiance in the real-time meteorological data into the inclined irradiance of the photovoltaic module through a radiation conversion model. The operating temperature estimation unit is used to estimate the real-time operating temperature characteristics of the photovoltaic module based on the slope irradiance, the ambient temperature and wind speed in the real-time meteorological data, and the photovoltaic module installation tilt angle in the real-time equipment data, through a thermal balance model. The dust attenuation factor calculation unit is used to calculate the real-time dust attenuation factor based on the rainfall records and consecutive rainless days in the real-time meteorological data, combined with the seasonal empirical coefficient.

[0072] In a specific embodiment, the inefficient identification module 40 includes: A single-day identification and judgment unit is used to mark the day as a suspected inefficient state if the daily power generation deviation rate exceeds the preset static deviation threshold. The multi-day identification and judgment unit is used to confirm that the inverter has entered an inefficient operating state if the power generation deviation rate exceeds the preset static deviation threshold for N consecutive days, where N is an integer greater than 1.

[0073] In a specific embodiment, the fault attribution module 50 includes: The fault query unit is used to query the internal fault information of the inverter. An electrical parameter tracking unit is used to analyze the consistency of DC-side electrical parameters of each maximum power point tracking MPPT branch of the inverter if no efficiency-related fault information is found. The fault attribution unit is used to output the corresponding fault attribution conclusion based on the query results of the internal fault information or the consistency analysis results of the DC side electrical parameters.

[0074] In a specific embodiment, the electrical parameter tracking unit is specifically used for: Calculate the dispersion rate of DC voltage or DC current for each MPPT branch; When the dispersion rate exceeds the second threshold, it is determined that the DC side electrical parameters are inconsistent.

[0075] In a specific embodiment, it also includes a power generation prediction model training module, comprising: The historical data acquisition unit is used to acquire historical equipment data, historical power generation data, and corresponding historical meteorological data of the photovoltaic system. The historical physical characteristic estimation unit is used to dynamically estimate the historical operating temperature characteristics and historical dust attenuation factor of the photovoltaic module based on the historical meteorological data and the historical equipment data. The data fusion unit is used to fuse the historical equipment data, the historical power generation data, the historical meteorological data, the historical operating temperature characteristics, and the historical dust attenuation factor to obtain a fused dataset. The model training unit is used to train a power generation prediction model based on the fused dataset.

[0076] In a specific embodiment, the power generation prediction model training module further includes: The data cleaning unit is used to clean the historical equipment data, the historical power generation data, and the historical meteorological data to remove abnormal data items. The time alignment unit is used to align the cleaned historical equipment data, historical power generation data, and historical meteorological data in terms of time.

[0077] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the inverter inefficiency diagnostic device and each unit of the above-mentioned photovoltaic system can be referred to the corresponding description in the foregoing method embodiments. For the sake of convenience and brevity, it will not be repeated here.

[0078] The aforementioned inverter inefficiency diagnostic device for photovoltaic systems can be implemented as a computer program, which can be used in, for example... Figure 5 It runs on the computer device shown.

[0079] Please see Figure 5 , Figure 5This is a schematic block diagram of a computer device provided in an embodiment of this application. The computer device 500 can be a terminal or a server. The terminal can be an electronic device with communication functions, such as a smartphone, tablet, laptop, desktop computer, personal digital assistant, or wearable device. The server can be a standalone server or a server cluster composed of multiple servers.

[0080] See Figure 5 The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0081] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform a method for diagnosing inverter inefficiency in a photovoltaic system.

[0082] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0083] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can perform a method for diagnosing inverter inefficiency in a photovoltaic system.

[0084] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0085] The processor 502 is used to run a computer program 5032 stored in the memory to implement the inverter inefficiency diagnosis method of the photovoltaic system as described above.

[0086] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0087] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0088] Therefore, the present invention also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions. When executed by a processor, the program instructions cause the processor to perform the inverter inefficiency diagnosis method for a photovoltaic system as described above.

[0089] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0090] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0091] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of each unit is merely a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0092] The steps in the method of this invention can be adjusted, merged, or reduced in order according to actual needs. The units in the device of this invention can be merged, divided, or reduced according to actual needs. Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0093] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0094] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for diagnosing inverter inefficiency in a photovoltaic system, characterized in that, include: Acquire real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed; Based on the real-time meteorological data and real-time equipment data, the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module are estimated. The real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor are input into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed. Based on the theoretical power generation and the actual power generation of the inverter during the period to be diagnosed, the power generation deviation rate is calculated, and according to the pre-configured inefficient operation identification conditions, the inverter is identified as being in an inefficient operation state based on the power generation deviation rate. When the inverter is detected to be in an inefficient operating state, a fault attribution analysis is performed based on the inverter's internal state parameters and DC-side operating parameters.

2. The inverter inefficiency diagnosis method for photovoltaic systems according to claim 1, characterized in that, The estimation of the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data includes: The horizontal irradiance in the real-time meteorological data is converted into the inclined irradiance of the photovoltaic module through a radiation conversion model; Based on the slope irradiance, the ambient temperature and wind speed in the real-time meteorological data, and the photovoltaic module installation tilt angle in the real-time equipment data, the real-time operating temperature characteristics of the photovoltaic module are estimated through a thermal balance model. Based on the rainfall records and consecutive rainless days in the real-time meteorological data, and combined with the seasonal empirical coefficient, the real-time dust attenuation factor is calculated.

3. The inverter inefficiency diagnosis method for photovoltaic systems according to claim 1, characterized in that, The inefficient operation identification condition is a continuous verification condition based on a preset static deviation threshold. The step of identifying whether the inverter is in an inefficient operation state based on the power generation deviation rate according to the pre-configured inefficient operation identification condition includes: If the daily power generation deviation rate exceeds the preset static deviation threshold, the day is marked as a suspected inefficient state. If the power generation deviation rate exceeds the preset static deviation threshold for N consecutive days, the inverter is confirmed to have entered an inefficient operating state, where N is an integer greater than 1.

4. The inverter inefficiency diagnosis method for photovoltaic systems according to claim 1, characterized in that, The fault attribution analysis based on the inverter's internal state parameters and DC-side operating parameters includes: Query the internal fault information of the inverter; If no efficiency-related fault information is found, analyze the consistency of the DC-side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter. Based on the query results of the internal fault information or the consistency analysis results of the DC side electrical parameters, the corresponding fault attribution conclusion is output.

5. The inverter inefficiency diagnosis method for photovoltaic systems according to claim 4, characterized in that, The analysis of the consistency of DC-side electrical parameters of each maximum power point tracking (MPPT) branch of the inverter includes: Calculate the dispersion rate of DC voltage or DC current for each MPPT branch; When the dispersion rate exceeds the second threshold, it is determined that the DC side electrical parameters are inconsistent.

6. The method for diagnosing inverter inefficiency in a photovoltaic system according to any one of claims 1-5, characterized in that, The power generation prediction model is obtained through a training process, which includes: Acquire historical equipment data, historical power generation data, and corresponding historical meteorological data for the photovoltaic system; Based on the historical meteorological data and the historical equipment data, the historical operating temperature characteristics and historical dust degradation factor of the photovoltaic module are dynamically estimated. The historical equipment data, historical power generation data, historical meteorological data, historical operating temperature characteristics, and historical dust attenuation factor are fused to obtain a fused dataset. The power generation prediction model was trained based on the fused dataset.

7. The inverter inefficiency diagnosis method for photovoltaic systems according to claim 6, characterized in that, After acquiring the historical equipment data, historical power generation data, and corresponding historical meteorological data of the photovoltaic system, the method further includes: The historical equipment data, historical power generation data, and historical meteorological data are cleaned to remove abnormal data items; The historical equipment data, historical power generation data, and historical meteorological data were aligned by time after cleaning.

8. A diagnostic device for inverter inefficiency in a photovoltaic system, characterized in that, include: The real-time data acquisition module is used to acquire real-time equipment data and real-time meteorological data of the photovoltaic system during the period to be diagnosed. The physical characteristic estimation module is used to estimate the real-time operating temperature characteristics and real-time dust degradation factor of the photovoltaic module based on the real-time meteorological data and real-time equipment data. The theoretical power generation prediction module is used to input the real-time equipment data, the real-time meteorological data, the real-time operating temperature characteristics, and the real-time dust attenuation factor into the trained power generation prediction model to obtain the theoretical power generation of the inverter during the period to be diagnosed. The inefficiency identification module is used to calculate the power generation deviation rate based on the theoretical power generation and the actual power generation of the inverter during the period to be diagnosed, and to identify whether the inverter is in an inefficient operating state based on the power generation deviation rate according to the pre-configured inefficient operation identification conditions. The fault attribution module is used to perform fault attribution analysis based on the inverter's internal state parameters and DC-side operating parameters when the inverter is identified as being in an inefficient operating state.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the inverter inefficiency diagnosis method for a photovoltaic system as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the inverter inefficiency diagnosis method for the photovoltaic system as described in any one of claims 1 to 7.

Images