Method and system for managing photovoltaic four-function gateway

Abstract

The invention relates to the technical field of photovoltaic gateways, and provides a management method and system of a photovoltaic four-function gateway, and the management method comprises the steps: constructing a photovoltaic power generation power prediction model based on electrical parameters, equipment state parameters and environmental parameters, and outputting a photovoltaic power generation power prediction value of a future time period; calculating the fluctuation degree of the photovoltaic power generation power prediction value according to the photovoltaic power generation power prediction value in the future time period, and adaptively adjusting the disturbance step length of the MPPT algorithm based on the fluctuation degree to track the maximum power point of the photovoltaic module; constructing a photovoltaic power generation power fitting model based on the environmental parameters and the electrical parameters, judging the dust accumulation state of the photovoltaic panel by comparing the deviation between the real-time power generation power and the fitting power, and generating a cleaning suggestion based on future weather forecast; and based on the electrical parameters, the equipment state parameters and the environmental parameters, constructing a fault diagnosis model, and performing real-time evaluation and fault early warning on the state of the photovoltaic panel.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic gateway technology, and in particular to a management method and system for a photovoltaic four-way gateway. Background Technology

[0002] With the continued growth of global energy demand and the increasing severity of environmental problems, photovoltaic power generation, as a clean and renewable energy solution, has ushered in unprecedented development opportunities. However, the following problems still exist in the application of photovoltaics: MPPT (Maximum Power Point Tracking) technology is the core of photovoltaic power generation systems, used to track the maximum power point of solar panels in real time to ensure the system operates at optimal efficiency. However, in reality, due to unreasonable perturbation step size settings or rapid changes in the external environment (such as sudden changes in irradiance), the MPPT algorithm often cannot adjust in time, leading to overshoot and a significant reduction in efficiency. Dust accumulation on the surface of photovoltaic modules is one of the main factors contributing to energy efficiency losses in power plants. Dust not only directly reduces light transmittance and power generation, but also hinders module heat dissipation, leading to increased operating temperature and further deteriorating conversion efficiency. Long-term dust accumulation can even cause surface corrosion, accelerating module aging and significantly shortening its lifespan. Statistics show that uncollected dust can cause a 2%-10% loss of electricity per month, which has a significant impact on operational revenue. Currently, the industry urgently needs a simple and effective method for monitoring dust accumulation to rationally schedule cleaning times and find the optimal balance between maintenance costs and power generation revenue. Common photovoltaic (PV) module failures include cell aging, dust accumulation, microcracks, and diode failure. Current main detection methods include light transmittance monitoring, electroluminescence (EL) testing, and infrared thermal imaging. However, these methods have many limitations, such as stringent environmental requirements, complex operation, and high costs, making them unsuitable for the efficient operation and maintenance needs of large-scale PV power plants. Summary of the Invention

[0003] To address the aforementioned shortcomings, the present invention aims to propose a management method and system for a photovoltaic four-dimensional gateway, which seeks to solve the key problems of insufficient adaptability of the MPPT algorithm, energy efficiency loss caused by dust accumulation, and limitations of fault detection methods in existing photovoltaic technologies, thereby improving the operating efficiency of photovoltaic power plants and reducing operation and maintenance costs.

[0004] To achieve this objective, the present invention adopts the following technical solution: A management method for a photovoltaic four-way gateway, the management method comprising the following steps: S1: Collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; S2: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a photovoltaic power generation prediction model and output the predicted photovoltaic power generation value for future periods; S3: Based on the predicted photovoltaic power generation value for the future period, calculate the fluctuation level of the predicted photovoltaic power generation value, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation level to track the maximum power point of the photovoltaic module. S4: Based on the environmental and electrical parameters, a photovoltaic power generation fitting model is constructed. By comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, the dust accumulation status of the photovoltaic panels is determined, and cleaning suggestions are generated based on future weather forecasts. S5: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a fault diagnosis model to perform real-time assessment and fault warning of the photovoltaic panel status.

[0005] Preferably, step S1 includes: The electrical parameters include the input and output voltage, input and output current, input and output power, active power, reactive power, and power generation of the photovoltaic modules and inverters. The equipment status parameters include the inverter's operating mode and temperature, the photovoltaic module status, and the photovoltaic switch status. The environmental parameters include wind speed, wind direction, light intensity, air pressure, humidity, outdoor temperature, direct radiation, diffuse radiation, solar altitude angle, solar azimuth angle, and weather conditions.

[0006] Preferably, step S2 includes: S21: Perform data preprocessing on the electrical parameters, equipment status parameters and environmental parameters. The data preprocessing includes data cleaning, handling of missing and outlier values ​​and data normalization. S22: Based on the preprocessed data, feature extraction and construction are performed. The feature extraction and construction includes extracting meteorological features from the environmental parameters and constructing statistical and signal analysis features from the electrical parameters and meteorological features, including mean, root mean square value, peak value, maximum difference, peak index, waveform index, impulse index, margin index, skewness, and kurtosis. S23: Based on the results of feature extraction and construction, a long short-term memory network model is constructed and used to predict photovoltaic power generation, outputting the predicted photovoltaic power generation values ​​at several future time granularities.

[0007] Preferably, step S3 includes: S31: Obtain the predicted photovoltaic power generation for future periods, calculate the fluctuation of the predicted photovoltaic power generation, and obtain the rolling standard deviation of the predicted photovoltaic power. Satisfying the relation ,in Indicates the size of the time window. Indicates the current time point, This represents the predicted photovoltaic power generation at a specific time point i. This represents the average value of the predicted photovoltaic power generation from t+1 to t+w+1. S32: Set a threshold between the calculated volatility and the standard deviation. The comparison is performed, and the perturbation voltage step size of the MPPT algorithm is adaptively adjusted based on the comparison result, where: If the volatility is less than or equal to the standard deviation threshold Then the first disturbance voltage is used. Adjust the output voltage of the photovoltaic panel; If the degree of volatility is greater than the standard deviation threshold Then the second disturbance voltage is used. Adjusting the output voltage of the photovoltaic panel, the second disturbance voltage Less than the first disturbance voltage ; S33: After applying a disturbance voltage, measure the change in output power of the photovoltaic system, and adjust the subsequent disturbance direction based on the change in output power: if the output power increases, maintain the current disturbance direction; if the output power decreases, reverse the disturbance direction.

[0008] Preferably, step S4 includes: S41: Select a historical period when the photovoltaic panels are in a clean state, and obtain the irradiance, outdoor temperature and photovoltaic power generation of the selected period as modeling data; S42: Perform data preprocessing on the modeling data, including outlier identification and removal, missing value filling and data normalization. Based on the preprocessed data, fit the outdoor temperature, irradiance and photovoltaic power generation to construct a photovoltaic power generation fitting model. S43: Obtain real-time irradiance and outdoor temperature, and calculate the expected photovoltaic power generation based on the photovoltaic power generation fitting model; S44: Compare the actual photovoltaic power generation with the expected photovoltaic power generation. If the actual photovoltaic power generation is consistently less than the expected photovoltaic power generation and the relative deviation exceeds the preset deviation threshold, then the photovoltaic panel is determined to be dusty, and the dust accumulation degree is calculated. S45: Based on dust accumulation, predicted power generation, predicted power generation attenuation, electricity price, and cleaning cost, quantify net revenue and generate cleaning recommendations: If the expected increase in revenue after cleaning is greater than the cleaning cost, generate a cleaning recommendation that suggests cleaning; otherwise, generate a cleaning recommendation that does not suggest cleaning.

[0009] Preferably, the photovoltaic power generation fitting model satisfies the following relationship: ; in, , and Represents the fitting coefficient. Indicates the expected photovoltaic power generation capacity. Indicates irradiance, Indicates outdoor temperature. This represents the irradiance correction factor.

[0010] Preferably, the dust accumulation degree satisfies the following relationship: ; in, Indicates the expected photovoltaic power generation capacity. Indicates the actual power generation of photovoltaic power; Power generation attenuation .

[0011] Preferably, the calculation of revenue satisfies the following relationship: ; in, Indicates the power generation attenuation. Indicates the value of electricity. It is represented as time.

[0012] Preferably, step S5 includes: S51: Collect historical fault tag data corresponding to photovoltaics, and perform data preprocessing on the operating status data, including outlier identification and screening, missing value filling, time series synchronization and data normalization. S52: Based on the preprocessed data, construct derived features reflecting the operating status of the photovoltaic panel; based on the derived features and the historical fault label data, construct and train a machine learning fault diagnosis model. S53: Utilize the trained machine learning fault diagnosis model to perform real-time assessment and fault warning of photovoltaic panel status.

[0013] A management system for a photovoltaic (PV) four-functional gateway, wherein the management system is applied to a management method for a PV four-functional gateway as described above, and the management system includes: The data acquisition module is used to collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; The photovoltaic short-term forecasting module is used to construct a photovoltaic power generation forecasting model based on the electrical parameters, equipment status parameters, and environmental parameters, and output the photovoltaic power generation forecast value for future periods. The MPPT adaptive adjustment module is used to calculate the fluctuation of the photovoltaic power generation forecast value based on the photovoltaic power generation forecast value for the future period, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation to track the maximum power point of the photovoltaic module. The photovoltaic panel operation and maintenance optimization module is used to construct a photovoltaic power generation fitting model based on the environmental and electrical parameters. By comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, the module determines the dust accumulation status of the photovoltaic panels and generates cleaning suggestions based on future weather forecasts. The fault diagnosis module is used to construct a fault diagnosis model based on the electrical parameters, equipment status parameters, and environmental parameters, and to perform real-time assessment and fault warning of the photovoltaic panel status.

[0014] One of the above technical solutions has the following advantages or beneficial effects: This invention effectively addresses the problems of poor adaptability of the MPPT algorithm, insufficient dust accumulation monitoring, and high fault detection costs in existing photovoltaic systems. First, by collecting electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station, a comprehensive and accurate data foundation is provided for subsequent power prediction, MPPT adjustment, dust accumulation monitoring, and fault diagnosis. Next, a photovoltaic power generation prediction model is constructed based on these parameters, outputting predicted values ​​for future periods. This allows the system to anticipate power generation trends in advance, providing a basis for subsequent adaptive MPPT adjustment. Then, the perturbation step size of the MPPT algorithm is adaptively adjusted according to the predicted power generation fluctuations, ensuring that the photovoltaic modules always operate near their maximum power point in complex environments, thus improving power generation efficiency. Simultaneously, by constructing a photovoltaic power generation fitting model and comparing the deviation between the actual photovoltaic power generation and the fitted power, the dust accumulation status of the photovoltaic panels is determined, and cleaning suggestions are generated based on future weather forecasts, achieving accurate monitoring and timely handling of dust accumulation problems and reducing power generation losses. Finally, a fault diagnosis model is constructed based on the collected parameters to perform real-time assessment of the photovoltaic panel status and provide fault warnings, reducing fault detection costs and improving system reliability and operation and maintenance efficiency. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0016] Figure 1 This is a flowchart of the management method for a photovoltaic four-way gateway provided in an embodiment of the present invention; Figure 2This is a schematic diagram of the management system of the photovoltaic four-way gateway provided in an embodiment of the present invention. Detailed Implementation

[0017] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0018] In this invention, the terms "comprising," "including," or any other variations thereof are intended to cover a 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 a process, method, article, or apparatus. Without further limitation, 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 said element.

[0019] A management method for a photovoltaic four-way gateway, such as Figure 1 As shown, in a preferred embodiment of the present invention, the method includes the following steps: S1: Collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; It should be noted that electrical parameters refer to the input and output voltage, current, and power of photovoltaic modules and inverters, which are used to reflect the real-time operating status of the photovoltaic system; equipment status parameters include the inverter's operating mode and temperature, which are used to monitor the health status and operating mode of the equipment; environmental parameters such as temperature and irradiance are used to provide information on the external conditions for the operation of the photovoltaic system.

[0020] Understandably, collecting electrical, equipment, and environmental parameters of photovoltaic power plants can provide a reliable data foundation for subsequent power prediction, MPPT adjustment, and fault diagnosis. The collection of these parameters is achieved through existing sensors and monitoring modules. Electrical parameters reflect the system's operating status, equipment parameters reflect its health status, and environmental parameters provide external conditions. Together, these three constitute the input data source for the intelligent diagnostic system.

[0021] Specifically, data can be collected in the following ways: 1) By directly communicating with the inverter through a photovoltaic four-way gateway, system-level electrical parameters such as the inverter's input and output current, voltage, and power can be obtained; 2) Environmental parameters such as ambient temperature, irradiance, and wind speed can be measured using a weather station, and the weather station data can be directly connected to the gateway. This collected data is then aggregated into the photovoltaic four-way gateway via a data cable or wireless transmission module, forming a complete dataset. The combination of multiple sensors can improve the comprehensiveness and accuracy of data collection, providing high-quality data support for subsequent intelligent diagnostics.

[0022] S2: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a photovoltaic power generation prediction model and output the predicted photovoltaic power generation value for future periods; It should be noted that the photovoltaic power generation prediction model is an algorithmic model based on historical and real-time data, used to predict photovoltaic power generation over a future period. This model typically includes a data preprocessing module (for cleaning and normalizing data), a feature extraction module (for extracting features such as mean and variance), and a model training module (such as a Long Short-Term Memory (LSTM) network). Its purpose is to provide reference values ​​for future power generation for MPPT adjustment and dust accumulation determination.

[0023] Understandably, the purpose of building a photovoltaic power generation prediction model is to know the future power generation in advance so that the smart gateway can dynamically adjust system parameters and optimize power generation efficiency. The model predicts future power changes by learning patterns in historical data, thereby providing a basis for the intelligent adjustment of the system.

[0024] Specifically, an LSTM neural network model can be used for short-term power prediction, with inputs including historical power, irradiance, temperature, etc., and outputting power prediction values ​​for the next 30 minutes (5-minute granularity); it can also be combined with weather clustering analysis to improve the model's adaptability to different weather conditions.

[0025] S3: Based on the predicted photovoltaic power generation value for the future period, calculate the fluctuation level of the predicted photovoltaic power generation value, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation level to track the maximum power point of the photovoltaic module. It should be noted that the fluctuation of photovoltaic power generation forecast refers to the magnitude of change in power generation over a future period, typically quantified by calculating the rolling standard deviation. The MPPT algorithm is a control algorithm used to track the maximum power point of photovoltaic modules; its perturbation step size determines the tracking accuracy and speed. The purpose of adaptively adjusting the perturbation step size is to dynamically adjust the tracking strategy based on power fluctuations, ensuring efficient system operation even in complex environments.

[0026] Understandably, by calculating the fluctuations in the predicted photovoltaic power generation, we can assess the power changes the system will face in the future. When the fluctuations are small, a larger perturbation step size can quickly track the maximum power point; when the fluctuations are large, a smaller perturbation step size can avoid overshoot and oscillations, ensuring stable system operation. This adaptive adjustment mechanism enables the MPPT algorithm to maintain efficient power point tracking performance under different environments.

[0027] S4: Based on the environmental and electrical parameters, a photovoltaic power generation fitting model is constructed. By comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, the dust accumulation status of the photovoltaic panels is determined, and cleaning suggestions are generated based on future weather forecasts. It should be noted that the photovoltaic power generation fitting model refers to a power-environment parameter relationship model established based on historical data under clean conditions; real-time power generation refers to the power generation of the photovoltaic panels in actual operation, while the fitted power refers to the expected photovoltaic power generation calculated according to the photovoltaic power generation fitting model, i.e., the ideal power generation of the module. The deviation between the two can reflect the degree of dust accumulation on the photovoltaic panels. The cleaning recommendation is generated based on the degree of dust accumulation and economic analysis, and is used to guide whether the photovoltaic panels need to be cleaned.

[0028] Understandably, by comparing the model's predictions with the actual outputs, the decrease in power generation efficiency caused by dust accumulation can be identified, thereby avoiding blind cleaning or delaying cleaning and achieving economic and scientific cleaning decisions.

[0029] S5: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a fault diagnosis model to perform real-time assessment and fault warning of the photovoltaic panel status.

[0030] It should be noted that the fault diagnosis model refers to a classification or regression model trained based on historical fault data, used to identify whether photovoltaic panels have faults and their types; real-time assessment and fault early warning refers to the system judging whether a fault may occur under the current operating state and issuing a warning in advance.

[0031] Understandably, by integrating data from multiple sources such as electrical, environmental, and equipment sources, the system's operating status can be comprehensively reflected, improving the accuracy and timeliness of fault identification and avoiding the lag and high cost of relying on manual inspections in traditional methods.

[0032] Specifically, an XGBoost model can be constructed, with inputs including features such as voltage, current, temperature, and power / irradiance ratio, and outputting fault type labels; time series features can also be introduced to improve the model's sensitivity to early faults.

[0033] Preferably, step S1 includes: The electrical parameters include the input and output voltage, input and output current, input and output power, active power, reactive power, and power generation of the photovoltaic modules and inverters. The equipment status parameters include the inverter's operating mode and temperature, the photovoltaic module status, and the photovoltaic switch status. The environmental parameters include wind speed, wind direction, light intensity, air pressure, humidity, outdoor temperature, direct radiation, diffuse radiation, solar altitude angle, solar azimuth angle, and weather conditions.

[0034] Electrical parameters are core data reflecting the power output characteristics of a photovoltaic power station. For example, input and output voltage are used to determine whether the system is operating within a reasonable voltage range, and power data is used to evaluate power generation efficiency. Equipment status parameters are used to monitor the system's operating status, such as inverter temperature, which can be used to determine whether it is overheating. Environmental parameters are external conditions that affect photovoltaic power generation. For example, sunlight intensity directly determines the total amount of energy received by the photovoltaic panels, and weather conditions such as sunny, rainy, and cloudy days.

[0035] Understandably, by refining the three types of parameter data, subsequent prediction, fitting, and diagnostic models can have a unified, complete, and comparable data foundation, avoiding model drift caused by missing dimensions or unknown states.

[0036] Specifically, electrical parameters can be collected by the inverter's built-in sensors, environmental parameters can be obtained through weather stations or sensors, and equipment status parameters can be read through the communication interface; the collection frequency can be set to once per minute to ensure data continuity; edge computing devices can also be used for local preprocessing to reduce transmission delay.

[0037] Preferably, step S2 includes: S21: Perform data preprocessing on the electrical parameters, equipment status parameters and environmental parameters. The data preprocessing includes data cleaning, handling of missing and outlier values ​​and data normalization. It should be noted that data preprocessing is a preliminary step in data mining, data cleaning is the process of removing noise and erroneous data, missing value handling is a method to fill in data gaps, and data normalization is a technique to unify data to a specific range.

[0038] Understandably, raw data often contains noise, missing values, and scaling differences. Directly using this data for analysis and modeling would severely reduce the accuracy and reliability of the model. Data cleaning removes noise and errors, restoring the pure essence of the data; missing value processing scientifically fills in data gaps, ensuring information integrity; and normalization unifies the data scale, eliminating dimensional differences and making the data more suitable for the operational characteristics of the model algorithm.

[0039] S22: Based on the preprocessed data, feature extraction and construction are performed. The feature extraction and construction includes extracting meteorological features from the environmental parameters and constructing statistical and signal analysis features from the electrical parameters and meteorological features, including mean, root mean square value, peak value, maximum difference, peak index, waveform index, impulse index, margin index, skewness, and kurtosis. It should be noted that feature extraction is the process of filtering out key features that have a significant impact on the prediction target from massive amounts of data, while feature construction is the operation of combining and transforming the original features to generate new features to improve the model's expressive power. Meteorological features in environmental parameters include wind level, wind speed, wind direction, and light intensity, reflecting the impact of the environment on power generation; statistical and signal analysis features, such as mean, root mean square value, and peak value, are quantitative indicators extracted from electrical parameters and meteorological features, used to capture the statistical characteristics and temporal fluctuation patterns of the data.

[0040] S23: Based on the results of feature extraction and construction, a long short-term memory network model is constructed and used to predict photovoltaic power generation, outputting the predicted photovoltaic power generation values ​​at several future time granularities.

[0041] It should be noted that Long Short-Term Memory (LSTM) networks are a special type of recurrent neural network (RNN) architecture specifically designed to handle long-distance dependencies in time-series data, possessing the ability to remember long-term information. The photovoltaic power generation prediction model is a machine learning model trained on historical data, used to predict future power generation. The results of feature extraction and construction provide input features for the model, encompassing time-series and statistical features extracted from electrical and environmental parameters, providing the model with a key source of knowledge.

[0042] Preferably, step S3 includes: S31: Obtain the predicted photovoltaic power generation for future periods, calculate the fluctuation of the predicted photovoltaic power generation, and obtain the rolling standard deviation of the predicted photovoltaic power. Satisfying the relation ,in Indicates the size of the time window. Indicates the current time point, This represents the predicted photovoltaic power generation at a specific time point i. This represents the average value of the predicted photovoltaic power generation from t+1 to t+w+1. S32: Set a threshold between the calculated volatility and the standard deviation. The comparison is performed, and the perturbation voltage step size of the MPPT algorithm is adaptively adjusted based on the comparison result, where: If the volatility is less than or equal to the standard deviation threshold Then the first disturbance voltage is used. Adjust the output voltage of the photovoltaic panel; If the degree of volatility is greater than the standard deviation threshold Then the second disturbance voltage is used. Adjusting the output voltage of the photovoltaic panel, the second disturbance voltage Less than the first disturbance voltage ; S33: After applying a disturbance voltage, measure the change in output power of the photovoltaic system, and adjust the subsequent disturbance direction based on the change in output power: if the output power increases, maintain the current disturbance direction; if the output power decreases, reverse the disturbance direction.

[0043] It should be noted that the rolling standard deviation of photovoltaic power forecasts It is a statistical measure that quantifies the severity of fluctuations in photovoltaic power generation over a future period. The time window size w determines the number of historical data points considered in the calculation, the current time point t indicates the fluctuation state at this moment in the calculation, and the predicted value... This is the predicted power generation value at time point i, and it is predicted by a Long Short-Term Memory (LSTM) network. It is the average of the predicted values ​​within the time window; these elements define it collaboratively. This accurately characterizes the power fluctuation features, providing a quantitative basis for subsequent disturbance step size adjustment.

[0044] Understandably, the MPPT algorithm's perturbation step size is adjusted based on fluctuations in predicted power output to ensure photovoltaic modules always operate near their maximum power point and improve photovoltaic power generation efficiency. When predicted power fluctuations are large, a smaller perturbation step size is used, enabling the system to track the maximum power point more accurately and avoiding missing the optimal operating point due to an excessively large step size. Conversely, when power fluctuations are small, the perturbation step size can be appropriately increased to improve tracking speed. This method achieves adaptive adjustment of the MPPT algorithm, enhancing the adaptability and stability of the photovoltaic system in complex environments.

[0045] In one embodiment, the input features for constructing the photovoltaic power generation prediction model PVModel include the actual photovoltaic power generation, irradiance, outdoor temperature, direct / scattered radiation at the same time as the prediction point yesterday and the day before, the average / maximum / minimum photovoltaic power generation, average / maximum / minimum irradiance, average / maximum / minimum outdoor temperature, average / maximum / minimum direct / scattered radiation, and the actual photovoltaic power generation, actual irradiance, outdoor temperature, and direct / scattered radiation every 5 minutes for the current time and 1 hour ago. These features are normalized and used as input. The output is the photovoltaic power generation value at a 5-minute granularity for the next 6 points from the prediction point, i.e., r. i+1 ,r i+2 ,...,r i+6The prediction model is a Long Short-Term Memory (LSTM) network model, with an overall architecture consisting of three layers: an input layer, an LSTM hidden layer, and an output layer. The input layer receives normalized input features. LSTM hidden layers (2 layers in total): Layer 1 contains 64 hidden units, with dropout rate = 0.2 and return_sequences = True; Layer 2 contains 128 hidden units, with dropout rate = 0.2 and return_sequences = False. Output layer: 6 nodes, outputting normalized 5-minute granular photovoltaic power generation values.

[0046] The model training parameters are: optimizer Adam (learning rate 0.001), loss function mean squared error (MSE), batch size = 64, and number of iterations = 100.

[0047] During forecasting, input features are extracted from historical data based on future forecast times. After normalization, these features are input into the PVModel. The forecast results are then inversely normalized to obtain the photovoltaic power generation forecast values ​​at a granularity of six 5-minute intervals, i.e., r. i+1 ,r i+2 ,...,r i+6 .

[0048] Set the time window size for the MPPT algorithm to w=6, and set a threshold for the standard deviation. =3. When the system obtains the sequence of 5-minute granular photovoltaic power generation prediction values ​​for 6 points over the next 30 minutes as 42, 45, 39, 55, 41, 50, it first calculates the average value of the prediction values ​​within the time window. Then, the rolling standard deviation is calculated based on the relationship. ≈6.09. Since 6.09 > 3 ( The system determines that the current fluctuation level is large, and therefore adopts a smaller second disturbance voltage. =0.2V to adjust the output voltage of the photovoltaic panel; after applying this disturbance voltage, the output power of the photovoltaic system was measured to increase from 45kW to 46kW. Based on this, the system determined that the current disturbance direction was correct and maintained this disturbance direction in subsequent control to continue tracking the maximum power point.

[0049] Preferably, step S4 includes: S41: Select a historical period when the photovoltaic panels are in a clean state, and obtain the irradiance, outdoor temperature and photovoltaic power generation of the selected period as modeling data; S42: Perform data preprocessing on the modeling data, including outlier identification and removal, missing value filling and data normalization. Based on the preprocessed data, fit the outdoor temperature, irradiance and photovoltaic power generation to construct a photovoltaic power generation fitting model. S43: Obtain real-time irradiance and outdoor temperature, and calculate the expected photovoltaic power generation based on the photovoltaic power generation fitting model; S44: Compare the actual photovoltaic power generation with the expected photovoltaic power generation. If the actual photovoltaic power generation is consistently less than the expected photovoltaic power generation and the relative deviation exceeds the preset deviation threshold, then the photovoltaic panel is determined to be dusty, and the dust accumulation degree is calculated. S45: Based on dust accumulation, predicted power generation, predicted power generation attenuation, electricity price, and cleaning cost, quantify net revenue and generate cleaning recommendations: If the expected increase in revenue after cleaning is greater than the cleaning cost, generate a cleaning recommendation that suggests cleaning; otherwise, generate a cleaning recommendation that does not suggest cleaning.

[0050] It should be noted that outdoor temperature affects power generation by influencing the properties of semiconductor materials; dust accumulation is used to quantify the impact of dust accumulation on power generation efficiency by comparing the deviation between real-time power generation and the fitted power (expected photovoltaic power generation); net benefit quantification is a process of economic benefit assessment by comparing the increase in power generation revenue after cleaning with the cleaning cost.

[0051] Understandably, the purpose of step S4 is to accurately monitor the dust accumulation status of photovoltaic panels and make scientific decisions regarding the timing of cleaning. The principle is to construct a fitted power model to simulate the power generation of the photovoltaic panels under dust-free conditions, and then compare it with the actual power generation. If the actual power is consistently lower than the fitted power and the deviation exceeds a preset threshold, it is determined that the photovoltaic panels have a dust accumulation problem. Further, a net benefit quantification calculation is performed to determine the difference between the increase in power generation benefit after cleaning and the cleaning cost. Higher dust levels indicate a decrease in predicted power generation and an increase in predicted power attenuation, meaning a greater potential power generation loss and a higher benefit after cleaning. If the increase in benefit exceeds the cost, cleaning is recommended; otherwise, it is not. This process not only accurately identifies power generation losses caused by dust accumulation but also optimizes cleaning decisions from an economic perspective, avoiding unnecessary cleaning expenses and ensuring a balance between the operation and maintenance costs and power generation benefits of the photovoltaic power plant.

[0052] For example, by combining weather forecasts and obtaining hourly weather forecasts for the next 15 days, if the irradiance continues to be high and the expected increase in revenue after cleaning is large, then cleaning is recommended. If there are many rainy days and fewer hours with high irradiance, the increase in revenue is less than the cleaning cost, or if there is heavy rain or storms that can effectively clean the photovoltaic panels, then cleaning is not recommended.

[0053] Preferably, the photovoltaic power generation fitting model satisfies the following relationship: ; in, , and Represents the fitting coefficient. Indicates the expected photovoltaic power generation capacity. Indicates irradiance, Indicates outdoor temperature. This represents the irradiance correction factor.

[0054] Specifically, the fitting coefficients are used to quantify the impact of irradiance and outdoor temperature on power generation. This reflects the expected power generation capacity of a photovoltaic (PV) panel under given irradiance and outdoor temperature. Irradiance is an environmental parameter affecting the power generation efficiency of PV panels. The power generation of PV panels is affected by irradiance and outdoor temperature. By constructing a fitting model, the influence relationship between these two factors on power generation can be quantified. Once dust accumulates on the surface of the PV panel, the actual power generation will be lower than the expected PV power generation, and the degree of deviation is related to the severity of dust accumulation. Therefore, the irradiance correction coefficient is a coefficient that changes according to weather type (e.g., sunny, cloudy, rainy, snowy). The irradiance correction coefficient is used to improve the accuracy of the PV power generation fitting model under different weather conditions. The PV power generation fitting model provides a benchmark for dust accumulation monitoring and is the basis for accurate dust accumulation judgment and cleaning decisions. In one embodiment, the irradiance correction coefficient can be 0.3 for thick cloud cover, 0.25 for light rain, 0.15 for moderate rain, and 0.20 for snowy days.

[0055] Preferably, the dust accumulation degree satisfies the following relationship: ; in, Indicates the expected photovoltaic power generation capacity. Indicates the actual power generation of photovoltaic power; Power generation attenuation .

[0056] Understandably, dust accumulation level provides a quantitative indicator for assessing the impact of dust accumulation on photovoltaic (PV) panels. The principle is to reflect the degree of power generation reduction caused by dust accumulation by comparing the actual power generation of the PV panels with the expected power generation. Specifically, when dust accumulates on the surface of PV panels, it reduces light transmittance and decreases power generation. The dust accumulation level is calculated by dividing the difference between the actual and expected power generation by the actual power generation, thus providing a direct quantification of the impact of dust accumulation on power generation efficiency. Dust accumulation level helps maintenance personnel to promptly identify dust accumulation problems and take maintenance measures such as cleaning to optimize the power generation performance of the PV power plant.

[0057] Preferably, the calculation of revenue satisfies the following relationship: ; in, Indicates the power generation attenuation. Indicates the value of electricity. It is represented as time.

[0058] Understandably, evaluating the economic benefits of cleaning photovoltaic panels to determine the necessity of such cleaning work involves estimating the power generation loss caused by dust accumulation based on dust levels, and combining this with factors such as sunlight conditions and electricity prices to calculate the recoverable power generation and corresponding economic benefits after cleaning. The recoverable power generation is related to the power attenuation. By calculating the power attenuation and combining it with time and electricity price parameters, the increase in power generation revenue after cleaning can be obtained: power attenuation × time = change in power generation after cleaning. If the increase in revenue exceeds the cleaning cost, the cleaning operation is economically worthwhile and can bring net benefits to the photovoltaic power station; otherwise, cleaning may not be recommended to avoid unnecessary economic expenditure. This process provides a scientific economic decision-making basis for the operation and maintenance management of photovoltaic power stations, ensuring the rationality and effectiveness of cleaning operations.

[0059] Based on the above description, the following is a partial cleaning strategy of one embodiment: First, based on real-time collected irradiance, outdoor temperature, and weather type, a photovoltaic power generation fitting model is used. Calculate the expected power under the current environment. And compare it with the real-time collected actual photovoltaic power generation. Compare and calculate the current dust accumulation level. Based on weather forecast data, hourly irradiance is obtained for the next 15 days. Outdoor temperature and weather type (e.g., sunny, heavy rain, light rain, snowfall, etc.), determine the irradiance correction factor according to the weather type. .

[0060] Subsequently, the hourly expected photovoltaic power generation was calculated. And then through the formula The power generation attenuation is obtained, where At the current dust accumulation level The correction value is based on the weather impact adjustment, i.e. =min(max( +Dust accumulation correction factor, 0), 100), for example, light rain (<10mm / day): dust accumulation +15% (forms a bonding layer), heavy rain (>38mm / day): dust accumulation -80% (effective cleaning), sunny day: dust accumulation +0%; and the correction value for dust accumulation affected by weather is continuous. If the weather does not change significantly, it continues the previous state. For example, if the future weather is light rain, light rain, light rain, heavy rain, heavy rain, sunny, sunny, then the dust accumulation will be... +15% +15% +15%, min(max( +15%-80%,0),100),min(max( +15%-80%,0),100),min(max( +15%-80%,0),100),min(max( +15%-80%,0),100).

[0061] By combining the power loss due to dust accumulation at various times with the corresponding electricity price, the estimated revenue is obtained by summing the results. (That is, the additional power generation revenue obtained by eliminating dust accumulation degradation through photovoltaic cleaning). For example, after obtaining the total expected power generation loss due to dust accumulation in the next 15 days as 85 yuan, the system compares this value with the preset single cleaning cost of 200 yuan: the calculated net revenue increment of cleaning is 85 yuan - 200 yuan = -115 yuan; since this value is less than zero, it is determined that this cleaning behavior is not economically feasible, and the conclusion of "cleaning is not recommended for the time being" is drawn, and monitoring continues. If dust accumulation intensifies or the weather forecast shows that there will be continuous sunny weather in the future, the system will re-trigger the evaluation.

[0062] Preferably, step S5 includes: S51: Collect historical fault tag data corresponding to photovoltaics, and perform data preprocessing on the operating status data, including outlier identification and screening, missing value filling, time series synchronization and data normalization. S52: Based on the preprocessed data, construct derived features reflecting the operating status of the photovoltaic panel; based on the derived features and the historical fault label data, construct and train a machine learning fault diagnosis model. S53: Utilize the trained machine learning fault diagnosis model to perform real-time assessment and fault warning of photovoltaic panel status.

[0063] The fault diagnosis model refers to a model trained on a large amount of data, used for real-time assessment and fault early warning of the photovoltaic (PV) panel's operating status. Historical fault label data refers to PV panel fault information recorded by past maintenance personnel, including fault type and occurrence time. This data carries clear labels and is used to train the model to identify different fault modes. Derived features are new feature variables extracted from the PV panel's operating data, such as the power-to-light intensity ratio, which can more sensitively reflect the equipment's operating status. Machine learning algorithms, such as XGBoost, are the technical methods for building and training fault diagnosis models. These algorithms can learn the mapping relationship between features and faults from large amounts of labeled data, thereby achieving automatic fault diagnosis.

[0064] Understandably, the purpose of step S5 is to achieve real-time monitoring of the photovoltaic (PV) panel status and early warning of faults. The principle is to collect a large amount of tagged historical fault data and real-time PV panel operating data, and then use machine learning algorithms to mine the correlation patterns between features and faults. The constructed fault diagnosis model can quickly and accurately determine whether the current state of the PV panel is normal when new operating data is input; if an anomaly is found, it further identifies the specific fault type. This process helps maintenance personnel to promptly identify potential faults, take preventative maintenance measures, avoid escalation of faults, reduce power generation losses, and improve the reliability and operating efficiency of the PV power plant.

[0065] A management system for a photovoltaic four-way gateway, such as Figure 2 As shown, the management system for the photovoltaic four-way gateway is applied to the photovoltaic four-way gateway management method described above. The management system for the photovoltaic four-way gateway includes: The data acquisition module is used to collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; The photovoltaic short-term forecasting module is used to construct a photovoltaic power generation forecasting model based on the electrical parameters, equipment status parameters, and environmental parameters, and output the photovoltaic power generation forecast value for future periods. The MPPT adaptive adjustment module is used to calculate the fluctuation of the photovoltaic power generation forecast value based on the photovoltaic power generation forecast value for the future period, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation to track the maximum power point of the photovoltaic module. The photovoltaic panel operation and maintenance optimization module is used to construct a photovoltaic power generation fitting model based on the environmental and electrical parameters. By comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, the module determines the dust accumulation status of the photovoltaic panels and generates cleaning suggestions based on future weather forecasts. The fault diagnosis module is used to construct a fault diagnosis model based on the electrical parameters, equipment status parameters, and environmental parameters, and to perform real-time assessment and fault warning of the photovoltaic panel status.

[0066] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0067] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A management method for a photovoltaic four-way gateway, characterized in that, The management method includes the following steps: S1: Collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; S2: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a photovoltaic power generation prediction model and output the predicted photovoltaic power generation value for future periods; S3: Based on the predicted photovoltaic power generation value for the future period, calculate the fluctuation level of the predicted photovoltaic power generation value, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation level to track the maximum power point of the photovoltaic module. S4: Based on the environmental and electrical parameters, a photovoltaic power generation fitting model is constructed. By comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, the dust accumulation status of the photovoltaic panels is determined, and cleaning suggestions are generated based on future weather forecasts. S5: Based on the electrical parameters, equipment status parameters, and environmental parameters, construct a fault diagnosis model to perform real-time assessment and fault warning of the photovoltaic panel status.

2. The management method for a photovoltaic four-way gateway according to claim 1, characterized in that, Step S1 includes: The electrical parameters include the input and output voltage, input and output current, input and output power, active power, reactive power, and power generation of the photovoltaic modules and inverters. The equipment status parameters include the inverter's operating mode and temperature, the photovoltaic module status, and the photovoltaic switch status. The environmental parameters include wind speed, wind direction, light intensity, air pressure, humidity, outdoor temperature, direct radiation, diffuse radiation, solar altitude angle, solar azimuth angle, and weather conditions.

3. The management method for a photovoltaic four-way gateway according to claim 1, characterized in that, Step S2 includes: S21: Perform data preprocessing on the electrical parameters, equipment status parameters and environmental parameters. The data preprocessing includes data cleaning, handling of missing and outlier values ​​and data normalization. S22: Based on the preprocessed data, feature extraction and construction are performed. The feature extraction and construction includes extracting meteorological features from the environmental parameters and constructing statistical and signal analysis features from the electrical parameters and meteorological features, including mean, root mean square value, peak value, maximum difference, peak index, waveform index, impulse index, margin index, skewness, and kurtosis. S23: Based on the results of feature extraction and construction, a long short-term memory network model is constructed and used to predict photovoltaic power generation, outputting the predicted photovoltaic power generation values ​​at several future time granularities.

4. The management method for a photovoltaic four-way gateway according to claim 1, characterized in that, Step S3 includes: S31: Obtain the predicted photovoltaic power generation for future periods, calculate the fluctuation of the predicted photovoltaic power generation, and obtain the rolling standard deviation of the predicted photovoltaic power. Satisfying the relation ,in Indicates the size of the time window. Indicates the current time point, This represents the predicted photovoltaic power generation at a specific time point i. This represents the average value of the predicted photovoltaic power generation from t+1 to t+w+1. S32: Set a threshold between the calculated volatility and the standard deviation. The comparison is performed, and the perturbation voltage step size of the MPPT algorithm is adaptively adjusted based on the comparison result, where: If the volatility is less than or equal to the standard deviation threshold Then the first disturbance voltage is used. Adjust the output voltage of the photovoltaic panel; If the volatility exceeds the standard deviation threshold Then the second perturbation voltage is used. Adjusting the output voltage of the photovoltaic panel, the second disturbance voltage Less than the first disturbance voltage ; S33: After applying a disturbance voltage, measure the change in output power of the photovoltaic system, and adjust the subsequent disturbance direction based on the change in output power: if the output power increases, maintain the current disturbance direction; if the output power decreases, reverse the disturbance direction.

5. The management method for a photovoltaic four-way gateway according to claim 1, characterized in that, Step S4 includes: S41: Select a historical period when the photovoltaic panels are in a clean state, and obtain the irradiance, outdoor temperature and photovoltaic power generation of the selected period as modeling data; S42: Perform data preprocessing on the modeling data, including outlier identification and removal, missing value filling and data normalization. Based on the preprocessed data, fit the outdoor temperature, irradiance and photovoltaic power generation to construct a photovoltaic power generation fitting model. S43: Obtain real-time irradiance and outdoor temperature, and calculate the expected photovoltaic power generation based on the photovoltaic power generation fitting model; S44: Compare the actual photovoltaic power generation with the expected photovoltaic power generation. If the actual photovoltaic power generation is consistently less than the expected photovoltaic power generation and the relative deviation exceeds the preset deviation threshold, then the photovoltaic panel is determined to be dusty, and the dust accumulation degree is calculated. S45: Based on dust accumulation, predicted power generation, predicted power generation attenuation, electricity price, and cleaning cost, quantify net revenue and generate cleaning recommendations: If the expected increase in revenue after cleaning is greater than the cleaning cost, generate a cleaning recommendation that suggests cleaning; otherwise, generate a cleaning recommendation that does not suggest cleaning.

6. The management method for a photovoltaic four-way gateway according to claim 5, characterized in that, The photovoltaic power generation fitting model satisfies the following relationship: ； in, , and Represents the fitting coefficient. Indicates the expected photovoltaic power generation capacity. Indicates irradiance, Indicates outdoor temperature. This represents the irradiance correction factor.

7. The management method for a photovoltaic four-way gateway according to claim 6, characterized in that, Dust accumulation satisfies the following relationship: ； in, Indicates the expected photovoltaic power generation capacity. Indicates the actual power generation of photovoltaic power; Power generation attenuation .

8. The management method for a photovoltaic four-way gateway according to claim 7, characterized in that, The calculation of revenue satisfies the following relationship: ； in, Indicates the power generation attenuation. Indicates the value of electricity. It is represented as time.

9. The management method for a photovoltaic four-way gateway according to claim 1, characterized in that, Step S5 includes: S51: Collect historical fault tag data corresponding to photovoltaics, and perform data preprocessing on the operating status data, including outlier identification and screening, missing value filling, time series synchronization and data normalization. S52: Based on the preprocessed data, construct derived features reflecting the operating status of the photovoltaic panel; based on the derived features and the historical fault label data, construct and train a machine learning fault diagnosis model. S53: Utilize the trained machine learning fault diagnosis model to perform real-time assessment and fault warning of photovoltaic panel status.

10. A management system for a photovoltaic four-way gateway, characterized in that, The management system of the photovoltaic four-way gateway is applied to the management method of the photovoltaic four-way gateway according to any one of claims 1-9, and the management system of the photovoltaic four-way gateway includes: The data acquisition module is used to collect electrical parameters, equipment status parameters, and environmental parameters of the photovoltaic power station; The photovoltaic short-term forecasting module is used to construct a photovoltaic power generation forecasting model based on the electrical parameters, equipment status parameters, and environmental parameters, and output the photovoltaic power generation forecast value for future periods. The MPPT adaptive adjustment module is used to calculate the fluctuation of the photovoltaic power generation forecast value based on the photovoltaic power generation forecast value for the future period, and adaptively adjust the perturbation step size of the MPPT algorithm based on the fluctuation to track the maximum power point of the photovoltaic module. The photovoltaic panel operation and maintenance optimization module is used to construct a photovoltaic power generation fitting model based on the environmental and electrical parameters, determine the dust accumulation status of the photovoltaic panel by comparing the deviation between the real-time power generation and the fitting power of the photovoltaic power generation fitting model, and generate cleaning suggestions based on future weather forecasts. The fault diagnosis module is used to construct a fault diagnosis model based on the electrical parameters, equipment status parameters, and environmental parameters, and to perform real-time assessment and fault warning of the photovoltaic panel status.

Images