Photovoltaic unit transient stability online evaluation method and system

Abstract

The invention discloses a photovoltaic unit transient stability online evaluation method and system, and relates to the technical field of photovoltaic equipment, and the method comprises the steps: S1, data collection and basic model establishment; s2, deploying a real-time monitoring system; s3, data preprocessing and anomaly detection; s4, dynamic simulation and model calibration; s5, performing real-time performance evaluation; s6, transient stability analysis; s7, performing prediction and trend analysis; s8, risk assessment and management; the system is constructed based on the method. According to the photovoltaic unit transient stability on-line evaluation method and system, comprehensive, real-time and dynamic transient stability evaluation of the photovoltaic unit is realized by integrating a plurality of functional modules based on the evaluation method, potential operation risks can be found and warned in time, the power generation efficiency and stability of the photovoltaic unit are improved, the maintenance cost is reduced, and the economic benefit is increased. And a powerful guarantee is provided for safe and efficient operation of a photovoltaic system. And meanwhile, timely feedback and strategy optimization of an evaluation result are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic equipment, and in particular to an online evaluation method and system for transient stability of a photovoltaic unit. Background Art

[0002] In the rapidly developing field of renewable energy, photovoltaic units are an important part of solar power generation, and their operating stability and efficiency directly affect the power generation capacity and economic benefits of the entire photovoltaic system.

[0003] However, the existing photovoltaic unit evaluation methods mostly rely on offline data analysis or simple real-time monitoring, lacking a comprehensive, dynamic and real-time transient stability evaluation mechanism. These methods often fail to accurately capture the real-time operating conditions of photovoltaic units under complex and changing environmental conditions, and fail to promptly warn of potential stability problems, resulting in high system maintenance costs, slow fault response, and limited power generation efficiency. In addition, traditional evaluation systems lack intelligent and automated data analysis capabilities, making it difficult to conduct long-term and continuous tracking and optimization of photovoltaic unit performance. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for online evaluation of transient stability of a photovoltaic unit to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for online evaluation of transient stability of a photovoltaic unit, comprising the following steps: S1. Data collection and basic model establishment Collect detailed specifications and operating data of photovoltaic units, extract behavioral patterns from existing operating data, identify operating characteristics under different environmental conditions, and establish basic data models; S2. Real-time monitoring system deployment Install sensors and monitoring equipment on the photovoltaic unit, such as power meters, temperature sensors, and light intensity meters, to collect real-time operating data of the photovoltaic unit for real-time evaluation of the performance and stability of the photovoltaic unit; S3. Data preprocessing and anomaly detection Preprocess the collected data, including cleaning, standardization, and normalization, to remove noise and inconsistent data points; design anomaly detection algorithms to identify anomalies in the data, such as equipment failures or environmental impact anomalies; S4. Dynamic simulation and model calibration Calibrate and validate the model of the PV unit using the collected historical data, and build a simulation model to simulate the performance of the PV unit under different operating conditions and environmental influences, and dynamically adjust the parameters to match the actual observed data to improve the accuracy and reliability of the prediction; S5. Real-time performance evaluation Based on the calibrated model, the current performance of the PV unit is evaluated, using algorithms to calculate key performance indicators such as power output stability, efficiency and loss rate; S6. Transient stability analysis Use small signal stability analysis and transient stability simulation technology to evaluate the responsiveness of photovoltaic units in the face of large-scale environmental changes, and predict the stability problems of photovoltaic systems by simulating different environments and operating scenarios; S7. Forecast and trend analysis Conduct in-depth analysis of the collected data using statistical and machine learning methods to predict the performance and risks of PV units over a period of time in the future; S8. Risk Assessment and Management Based on the assessment and prediction results, identify the main risk points faced by the PV units and formulate corresponding risk management strategies, including adjusting operating parameters, planning maintenance and upgrades, and formulating emergency response measures.

[0006] Further optimizing the technical solution, in step S1, the detailed specifications and operating data of the photovoltaic unit include the type, size, maximum power, efficiency, temperature coefficient of the photovoltaic panel, and the model, power range and control strategy of the inverter; The geographical location of the installation site, climate conditions such as average sunshine duration, temperature changes, and system historical data are collected; Behavioral patterns are extracted from existing operating data through data mining technology, and operating characteristics under different environmental conditions are identified through cluster analysis.

[0007] To further optimize the technical solution, in step S1, the basic data model established integrates multi-source data and environmental factors, and the data model is as follows:

[0008] in, ( ) is the predicted power output; ( ) is the temperature variable, which comes from the temperature sensor and affects the output through the polynomial regression model; ( ) is the light intensity, obtained from a light intensity meter; ( ) is the operating voltage of the inverter; ( ) is the historical power output data; ( ) is the seasonal factor from the GIS; ( ) are model parameters, which are obtained through training with historical data.

[0009] To further optimize the technical solution, in step S2, when evaluating the performance and stability of the photovoltaic unit in real time, a real-time response model of the photovoltaic unit is constructed, and the real-time response model is as follows:

[0010] in, ( ) is the real-time power output; ( ) is the real-time light intensity, measured by a light intensity meter; ( ) is the real-time temperature, measured by a temperature sensor; ( ) is the real-time voltage, provided by the inverter sensor; ( ) are model parameters that are adjusted through historical data and experiments to best fit the device performance.

[0011] To further optimize the technical solution, in step S3, the anomaly detection algorithm is based on a hybrid model of statistical learning, combined with time series analysis and outlier recognition technology, and the hybrid model is as follows:

[0012] in, ( ) means at time ( )’s anomaly score; ( ) is at time ( ) real-time data points, such as power output or temperature readings; ( )yes( ) is used to smooth historical data and reduce the impact of noise; ( )express( )and( ) ( ) are model parameters, where ( ) Adjusted baseline abnormality score, ( ) adjusts for the influence of historical errors, while ( ) is a nonlinear factor used to increase the sensitivity of the model to extreme values.

[0013] To further optimize the technical solution, in step S4, when performing calibration and verification, the interaction between electrical characteristics and environmental factors is considered, a comprehensive model of circuit and thermal forces is constructed, and the influence of temperature on electrical performance is quantified. The comprehensive model is as follows:

[0014] in, ( ) is at time ( )’s output power; ( ) is at time ( ) input power, which is related to the sunshine intensity; ( ) is at time ( ) ambient temperature; ( ) is the reference temperature, set to 25°C; ( ) is the temperature coefficient, which indicates the percentage of efficiency decrease for every 1°C increase in temperature; ( ) is an efficiency coefficient that is dynamically adjusted based on equipment aging and maintenance status.

[0015] To further optimize the technical solution, in step S5, a composite model is used to dynamically evaluate the performance of the photovoltaic unit, and the composite model is as follows:

[0016] in, ( ) represents the time ( )’s actual power output; ( ) is based on the expected power output under ideal conditions and the intensity of sunlight; ( ) is the difference between the actual temperature and the optimum operating temperature; ( ) is a coefficient that describes the sensitivity of the performance to temperature changes; ( ) is a decay factor that reflects the aging or maintenance status of the equipment.

[0017] To further optimize the technical solution, in step S6, the small signal stability analysis is applied in the photovoltaic system to predict the impact of small-scale but frequent environmental changes such as cloud shadows on system stability; Transient stability simulation technology is used in photovoltaic systems to evaluate the behavior of photovoltaic systems after encountering large-scale disturbances such as sudden weather changes and equipment failures. Through dynamic simulation in the time domain, the process of photovoltaic systems transitioning from one stable state to another possible stable state is examined.

[0018] Further optimizing the technical solution, in step S7, the statistical and machine learning methods include time series analysis, regression model analysis, machine learning methods and integrated learning methods; Time series analysis to predict future performance indicators such as power output and efficiency; Regression model analysis is used to predict or estimate the relationship between a response variable, such as photovoltaic power output, and one or more explanatory variables, such as temperature and light intensity; Machine learning methods to learn from historical data and predict the stability and performance of photovoltaic cells; Ensemble learning methods, which improve prediction accuracy and robustness by combining multiple models, are used to predict the stability and performance of photovoltaic cells.

[0019] A photovoltaic unit transient stability online evaluation system is constructed based on the above photovoltaic unit transient stability online evaluation method, including a data collection module, a basic model building module, a real-time monitoring deployment module, a data preprocessing module, an anomaly detection module, a simulation calibration module, a real-time performance evaluation module, a transient stability analysis module, a trend analysis module and a risk assessment module; The system also includes a report generation module and a continuous optimization module; The report generation module is used to formulate standardized report templates and regularly generate performance and stability evaluation reports for reference by system operation and maintenance and management personnel; The continuous optimization module is used to establish a continuous feedback loop to continuously optimize the monitoring, analysis and response strategies based on the evaluation results and the operating experience of the photovoltaic unit.

[0020] Compared with the prior art, the present invention provides a method and system for online evaluation of transient stability of photovoltaic units, which has the following beneficial effects: The online evaluation method and system for transient stability of photovoltaic units integrates multiple functional modules and realizes a comprehensive, real-time and dynamic transient stability evaluation of photovoltaic units based on the evaluation method. It can timely discover and warn of potential operating risks, improve the power generation efficiency and stability of photovoltaic units, reduce maintenance costs, and provide strong guarantees for the safe and efficient operation of photovoltaic systems. At the same time, it realizes timely feedback of evaluation results and strategy optimization, further improving the operation and management level of photovoltaic units. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of a process of an online evaluation method for transient stability of a photovoltaic unit proposed by the present invention; Figure 2 This is a schematic diagram of the analysis of step S6 in the online evaluation method for transient stability of a photovoltaic unit proposed by the present invention; Figure 3 This is a schematic diagram of the analysis of step S7 in the online evaluation method for transient stability of a photovoltaic unit proposed by the present invention; Figure 4 This is a structural flow chart of an online evaluation system for transient stability of a photovoltaic unit proposed by the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Example 1: Please refer to Figure 1 , a method for online evaluation of transient stability of a photovoltaic unit, comprising the following steps: S1. Data collection and basic model establishment Collect detailed specifications and operating data of photovoltaic units, extract behavioral patterns from existing operating data, identify operating characteristics under different environmental conditions, and establish basic data models.

[0024] In this embodiment, the detailed specifications and operating data of the photovoltaic unit include the type, size, maximum power, efficiency, and temperature coefficient of the photovoltaic panel, and the model, power range, and control strategy of the inverter; The geographical location of the installation site, climate conditions such as average sunshine duration, temperature changes, and system historical data are collected; Behavioral patterns are extracted from existing operating data through data mining technology, and operating characteristics under different environmental conditions are identified through cluster analysis.

[0025] At the same time, the basic data model established integrates multi-source data and environmental factors. The data model is as follows:

[0026] in, ( ) is the predicted power output; ( ) is the temperature variable, which comes from the temperature sensor and affects the output through the polynomial regression model; ( ) is the light intensity, obtained from a light intensity meter; ( ) is the operating voltage of the inverter; ( ) is the historical power output data; ( ) is the seasonal factor from the GIS; ( ) are model parameters, which are obtained through training with historical data.

[0027] The model uses machine learning algorithms for parameter optimization, especially random forest and gradient boosting machine methods, which excel in handling nonlinear relationships and high-dimensional data. Through continuous real-time data training, the model is able to self-learn and adjust to adapt to environmental changes and the effects of equipment aging.

[0028] In addition, the model combines cluster analysis to classify the operating conditions of PV units into multiple clusters through historical data, each cluster representing a specific combination of environmental and operating conditions. This classification helps identify performance patterns under specific conditions, thereby more accurately predicting the stability performance of PV units under similar conditions.

[0029] As new real-time data is collected, the model is used to predict power output and stability under upcoming conditions. For example, if the model indicates a possible performance degradation based on a sudden change in environmental parameters (such as rapid changes in temperature or light intensity), the system will automatically adjust the inverter parameters or notify maintenance personnel to check.

[0030] S2. Real-time monitoring system deployment Sensors and monitoring equipment, such as power meters, temperature sensors, and light intensity meters, are installed on the photovoltaic units to collect real-time operating data of the photovoltaic units for real-time evaluation of the performance and stability of the photovoltaic units.

[0031] In this embodiment, when evaluating the performance and stability of the photovoltaic unit in real time, a real-time response model of the photovoltaic unit is constructed. The real-time response model is as follows:

[0032] in, ( ) is the real-time power output; ( ) is the real-time light intensity, measured by a light intensity meter; ( ) is the real-time temperature, measured by a temperature sensor; ( ) is the real-time voltage, provided by the inverter sensor; ( ) are model parameters that are adjusted through historical data and experiments to best fit the device performance.

[0033] The model updates its parameters in real time with real-time data, which enables the model to dynamically adjust to changing environmental and operating conditions. For example, by analyzing the interactive effects of light intensity and temperature, power output fluctuations can be more accurately predicted, and the inverter operation strategy can be adjusted to optimize efficiency.

[0034] By applying this model to a large number of PV units, performance can be compared and standardized across the system, helping operations teams identify underperforming units and carry out targeted maintenance and optimization.

[0035] In practical applications, this model can be used in the following situations: Performance monitoring: Real-time monitoring of the deviation between power output and model predictions to detect equipment failures or efficiency declines.

[0036] Environmental impact analysis: Evaluate the performance of PV cells under different environmental conditions, such as analyzing the change in power output under different temperatures and light intensities.

[0037] Decision support: Provide data to support the decision-making system, adjust the operating parameters of the photovoltaic unit according to the model prediction results, and optimize the overall performance and stability.

[0038] S3. Data preprocessing and anomaly detection Preprocess the collected data, including cleaning, standardization, and normalization, to remove noise and inconsistent data points. Design anomaly detection algorithms to identify anomalies in the data, such as equipment failures or environmental impact anomalies.

[0039] In this embodiment, the anomaly detection algorithm is based on a hybrid model of statistical learning, combined with time series analysis and outlier identification technology. The hybrid model is as follows:

[0040] in, ( ) means at time ( )’s anomaly score; ( ) is at time ( ) real-time data points, such as power output or temperature readings; ( )yes( ) is used to smooth historical data and reduce the impact of noise; ( )express( )and( ) ( ) are model parameters, where ( ) Adjusted baseline abnormality score, ( ) adjusts for the influence of historical errors, while ( ) is a nonlinear factor used to increase the sensitivity of the model to extreme values.

[0041] This model allows adaptive adjustment of its parameters through historical data to best fit the operating characteristics of a specific PV cell. By applying machine learning techniques such as gradient descent, the parameters ( ) can be optimized based on actual data to improve the model’s prediction accuracy and anomaly detection capabilities.

[0042] When the model is used: For each newly received data point ( ) Calculate the anomaly score ( ).if( ) exceeds the preset threshold, an alarm is triggered.

[0043] Based on the statistical distribution of historical anomaly scores, the alarm thresholds are dynamically adjusted to adapt to environmental changes and system aging.

[0044] High anomaly scores may indicate specific types of equipment failures or environmental impacts, and analyzing these scores in relation to environmental variables and equipment parameters can help maintenance teams accurately diagnose problems.

[0045] S4. Dynamic simulation and model calibration The model of the PV unit is calibrated and verified using the collected historical data. By building a simulation model, the performance of the PV unit under different operating conditions and environmental influences is simulated, and the parameters are dynamically adjusted to match the actual observed data to improve the accuracy and reliability of the prediction.

[0046] In this embodiment, during calibration and verification, the interaction between electrical characteristics and environmental factors is considered, a comprehensive model of circuit and thermal forces is constructed, and the influence of temperature on electrical performance is quantified. The comprehensive model is as follows:

[0047] in, ( ) is at time ( )’s output power; ( ) is at time ( ) input power, which is related to the sunshine intensity; ( ) is at time ( ) ambient temperature; ( ) is the reference temperature, set to 25°C; ( ) is the temperature coefficient, which indicates the percentage of efficiency decrease for every 1°C increase in temperature; ( ) is an efficiency coefficient that is dynamically adjusted based on equipment aging and maintenance status.

[0048] Based on the above model, we use historical data to estimate ( )and( ) to make the model output as close as possible to the actual measured output power ( ).

[0049] The accuracy of the model is verified by comparing the model's predictions with newly collected data. This is usually done by calculating the mean squared error (MSE) between the predicted and actual values.

[0050] At the same time, the model needs to have its parameters updated regularly to reflect the aging of the equipment or the impact of any maintenance activities. This process can be automated to continuously optimize the parameters using real-time data.

[0051] S5. Real-time performance evaluation Based on the calibrated model, the current performance of the PV unit is evaluated, using algorithms to calculate key performance indicators such as power output stability, efficiency and loss rate.

[0052] In this embodiment, a composite model is used to dynamically evaluate the performance of the photovoltaic unit, and the composite model is as follows:

[0053] in, ( ) represents the time ( )’s actual power output; ( ) is based on the expected power output under ideal conditions and the intensity of sunlight; ( ) is the difference between the actual temperature and the optimum operating temperature; ( ) is a coefficient that describes the sensitivity of the performance to temperature changes; ( ) is a decay factor that reflects the aging or maintenance status of the equipment.

[0054] The composite model uses historical data to set ( )and( ) and continuously optimize these parameters through real-time data to ensure the accuracy of performance evaluation.

[0055] Performance indicators are calculated in real time during system operation and compared with thresholds to quickly identify performance degradation or device anomalies.

[0056] Combined with a real-time dashboard, performance indices are displayed in real time, allowing operators to intuitively see the system status and any changes that require attention.

[0057] S6. Transient stability analysis Small signal stability analysis and transient stability simulation techniques are used to evaluate the responsiveness of photovoltaic units in the face of large-scale environmental changes. By simulating different environments and operating scenarios, stability problems in photovoltaic systems are predicted.

[0058] like Figure 2 As shown, in this embodiment, small signal stability analysis is applied in a photovoltaic system to predict the impact of small-scale but frequent environmental changes such as cloud shadows on system stability. When small signal stability analysis is applied: Linearized model: The nonlinear dynamic model of the PV cell is first linearized, usually around a certain operating point.

[0059] Eigenvalue analysis: Calculate the eigenvalues ​​and eigenvectors of the linearized model, and analyze the real and imaginary parts of these eigenvalues ​​to determine the stability of the system. If the real part of the eigenvalue is negative, the system is stable at that operating point.

[0060] Transient stability simulation technology is used in photovoltaic systems to evaluate the behavior of photovoltaic systems after encountering large-scale disturbances such as sudden weather changes and equipment failures. Through dynamic simulation in the time domain, the process of photovoltaic systems transitioning from one stable state to another possible stable state is examined.

[0061] Transient stability simulation technology is used when: Time stepping: Advancing the state of a system in the time domain using numerical integration methods such as the Runge-Kutta method by setting initial conditions and perturbations.

[0062] Steady-state detection: monitors whether the system can return to a stable state after a disturbance, or whether the system dynamics diverge.

[0063] In this embodiment, simulation testing and verification methods can also be used to ensure that the model matches the behavior of the actual photovoltaic unit, including: Scenario simulation: Create a variety of possible operating and environmental scenarios, such as extreme temperature changes, sudden changes in light, etc., and simulate the system response under these scenarios.

[0064] Data comparison: Compare simulation results with historical or experimental data and calibrate model parameters to ensure high consistency and prediction accuracy.

[0065] In this embodiment, numerical sensitivity analysis can also be used to understand the impact of changes in different parameters on the stability of the photovoltaic system, thereby optimizing the system design and adjustment strategy.

[0066] The numerical sensitivity analysis is applied as follows: Parameter perturbations: Systematically changing key parameters of a model and observing the effects of these changes on the stability of the system.

[0067] Result analysis: Use statistical methods (such as analysis of variance) to determine which parameters are key factors affecting system stability.

[0068] S7. Forecast and trend analysis The collected data is analyzed in depth using statistical and machine learning methods to predict the performance and risks of PV units over future periods of time.

[0069] like Figure 3 As shown, in this embodiment, the statistical and machine learning methods include time series analysis, regression model analysis, machine learning methods, and integrated learning methods.

[0070] Time series analysis to predict future performance metrics such as power output and efficiency.

[0071] Among them, the ARIMA model (autoregressive integrated moving average model) is used to predict continuous data points, such as photovoltaic output power.

[0072] Seasonal decomposition, using the ETS (error, trend, seasonality) model or STL (seasonal and trend decomposition) to analyze seasonal patterns and long-term trends in the data.

[0073] Regression model analysis is used to predict or estimate the relationship between a response variable, such as photovoltaic power output, and one or more explanatory variables, such as temperature and light intensity.

[0074] Among them, linear regression is suitable for situations where the relationship between variables is relatively straight.

[0075] Multiple regression, which considers multiple influencing factors simultaneously, is more suitable for complex systems such as photovoltaic systems, where multiple variables may affect the output simultaneously.

[0076] Ridge regression and Lasso regression can help improve the stability and predictive ability of the model when there is multicollinearity in the data.

[0077] Machine learning methods are used to learn from historical data and predict the stability and performance of photovoltaic cells.

[0078] Among them, decision trees and random forests are very effective in dealing with classification and regression problems.

[0079] Support Vector Machines (SVMs) are suitable for classification and regression problems with high-dimensional data, especially when there are few data points but the dimensionality is high.

[0080] Neural networks, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), are able to process large datasets and abstract complex patterns and features from them.

[0081] Ensemble learning methods, which improve prediction accuracy and robustness by combining multiple models, are used to predict the stability and performance of photovoltaic cells.

[0082] Among them, Boosting, which gradually improves a single model, such as AdaBoost and Gradient Boosting Machine (GBM), enhances model performance by focusing on cases that the previous model failed to predict correctly.

[0083] Bagging, such as random forest, reduces the variance of the model and improves the stability of prediction by building multiple independent decision trees and voting or averaging them.

[0084] S8. Risk Assessment and Management Based on the assessment and prediction results, identify the main risk points faced by the PV units and formulate corresponding risk management strategies, including adjusting operating parameters, planning maintenance and upgrades, and formulating emergency response measures.

[0085] Example 2: Please refer to Figure 4 A photovoltaic unit transient stability online evaluation system is constructed based on the photovoltaic unit transient stability online evaluation method described in Example 1, including a data collection module, a basic model building module, a real-time monitoring deployment module, a data preprocessing module, an anomaly detection module, a simulation calibration module, a real-time performance evaluation module, a transient stability analysis module, a trend analysis module, and a risk assessment module. The uses of the above functional modules correspond to the use process in each step of the photovoltaic unit transient stability online evaluation method described in Example 1.

[0086] The system also includes a report generation module and a continuous optimization module.

[0087] The report generation module is used to formulate standardized report templates and regularly generate performance and stability evaluation reports for reference by system operation and maintenance and management personnel.

[0088] Among them, the continuous optimization module is used to establish a continuous feedback loop to continuously optimize the monitoring, analysis and response strategies based on the evaluation results and the operating experience of the photovoltaic unit.

[0089] In summary, the system starts with collecting data from the PV unit and the related environment, and builds a basic data model. Sensors and monitoring equipment are deployed on the PV unit to start real-time data collection. The real-time collected data is cleaned and standardized, and anomaly detection is performed to ensure data quality. Using the cleaned and calibrated data, dynamic simulation and model calibration are performed to adapt to actual observations. Based on the calibrated model, the performance of the current PV unit is evaluated. The stability of the PV unit under different environmental and operating conditions is analyzed. Future performance and risks are predicted using machine learning and statistical methods. Based on the prediction results, potential risks are identified and management strategies are formulated.

[0090] The beneficial effects of the present invention are: The online evaluation method and system for transient stability of photovoltaic units integrates multiple functional modules and realizes a comprehensive, real-time and dynamic transient stability evaluation of photovoltaic units based on the evaluation method. It can timely discover and warn of potential operating risks, improve the power generation efficiency and stability of photovoltaic units, reduce maintenance costs, and provide strong guarantees for the safe and efficient operation of photovoltaic systems. At the same time, it realizes timely feedback of evaluation results and strategy optimization, further improving the operation and management level of photovoltaic units.

[0091] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0092] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for online evaluation of transient stability of a photovoltaic unit, characterized in that: The following steps are involved: S1. Data collection and basic model establishment Collect detailed specifications and operating data of photovoltaic units, extract behavioral patterns from existing operating data, identify operating characteristics under different environmental conditions, and establish basic data models; S2. Real-time monitoring system deployment Install sensors and monitoring equipment on the photovoltaic unit, such as power meters, temperature sensors, and light intensity meters, to collect real-time operating data of the photovoltaic unit for real-time evaluation of the performance and stability of the photovoltaic unit; S3. Data preprocessing and anomaly detection Preprocessing of the collected data, including cleaning, standardization, and normalization, to remove noise and inconsistent data points; Design anomaly detection algorithms to identify anomalies in the data, such as equipment failures or environmental impact anomalies; S4. Dynamic simulation and model calibration Use the collected historical data to calibrate and validate the model of the PV unit, and build a simulation model to simulate the performance of the PV unit under different operating conditions and environmental influences, and dynamically adjust the parameters to match the actual observed data to improve the accuracy and reliability of the prediction; S5. Real-time performance evaluation Based on the calibrated model, the current performance of the PV unit is evaluated, using algorithms to calculate key performance indicators such as power output stability, efficiency and loss rate; S6. Transient stability analysis Use small signal stability analysis and transient stability simulation technology to evaluate the responsiveness of photovoltaic units in the face of large-scale environmental changes, and predict the stability problems of photovoltaic systems by simulating different environments and operating scenarios; S7. Forecast and trend analysis Conduct in-depth analysis of the collected data using statistical and machine learning methods to predict the performance and risks of PV units over a period of time in the future; S8. Risk Assessment and Management Based on the assessment and prediction results, identify the main risk points faced by the PV units and formulate corresponding risk management strategies, including adjusting operating parameters, planning maintenance and upgrades, and formulating emergency response measures.

2. A photovoltaic unit transient stability online evaluation method according to claim 1, characterized in that: In step S1, the detailed specifications and operating data of the photovoltaic unit include the type, size, maximum power, efficiency, temperature coefficient of the photovoltaic panel, and the model, power range and control strategy of the inverter; The geographical location of the installation site, climate conditions such as average sunshine duration, temperature changes, and system historical data are collected; Behavioral patterns are extracted from existing operating data through data mining technology, and operating characteristics under different environmental conditions are identified through cluster analysis.

3. A photovoltaic unit transient stability online evaluation method according to claim 1, characterized in that: In step S1, the basic data model established integrates multi-source data and environmental factors, and the data model is as follows: ; in, ( ) is the predicted power output; ( ) is the temperature variable, which comes from the temperature sensor and affects the output through the polynomial regression model; ( ) is the light intensity, obtained from a light intensity meter; ( ) is the operating voltage of the inverter; ( ) is the historical power output data; ( ) is the seasonal factor from the GIS; ( ) are model parameters, which are obtained through training with historical data.

4. A photovoltaic unit transient stability online evaluation method according to claim 1, characterized in that: In step S2, when evaluating the performance and stability of the photovoltaic unit in real time, a real-time response model of the photovoltaic unit is constructed. The real-time response model is as follows: ; in, ( ) is the real-time power output; ( ) is the real-time light intensity, measured by a light intensity meter; ( ) is the real-time temperature, measured by a temperature sensor; ( ) is the real-time voltage, provided by the inverter sensor; ( ) are model parameters that are adjusted through historical data and experiments to best fit the device performance.

5. The method for online evaluation of transient stability of a photovoltaic unit according to claim 1, characterized in that: In step S3, the anomaly detection algorithm is based on a hybrid model of statistical learning, combined with time series analysis and outlier recognition technology. The hybrid model is as follows: ; in, ( ) means at time ( )’s anomaly score; ( ) is at time ( ) real-time data points, such as power output or temperature readings; ( )yes( ) is used to smooth historical data and reduce the impact of noise; ( )express( )and( ) ( ) are model parameters, where ( ) Adjusted baseline abnormality score, ( ) adjusts for the influence of historical errors, while ( ) is a nonlinear factor used to increase the sensitivity of the model to extreme values.

6. A photovoltaic unit transient stability online evaluation method according to claim 1, characterized in that: In step S4, when performing calibration and verification, the interaction between electrical characteristics and environmental factors is considered, and a comprehensive model of circuit and thermal forces is constructed to quantify the impact of temperature on electrical performance. The comprehensive model is as follows: ; in, ( ) is at time ( )’s output power; ( ) is at time ( ) input power, which is related to the sunshine intensity; ( ) is at time ( ) ambient temperature; ( ) is the reference temperature, set to 25°C; ( ) is the temperature coefficient, which indicates the percentage of efficiency decrease for every 1°C increase in temperature; ( ) is an efficiency coefficient that is dynamically adjusted based on equipment aging and maintenance status.

7. A photovoltaic unit transient stability online evaluation method according to claim 1, characterized in that: In step S5, a composite model is used to dynamically evaluate the performance of the photovoltaic unit. The composite model is as follows: ; in, ( ) represents the time ( )’s actual power output; ( ) is based on the expected power output under ideal conditions and the intensity of sunlight; ( ) is the difference between the actual temperature and the optimum operating temperature; ( ) is a coefficient that describes the sensitivity of the performance to temperature changes; ( ) is a decay factor that reflects the aging or maintenance status of the equipment.

8. The method for online evaluation of transient stability of a photovoltaic unit according to claim 1, characterized in that: In step S6, small signal stability analysis is applied in the photovoltaic system to predict the impact of small-scale but frequent environmental changes such as cloud shadows on system stability; Transient stability simulation technology is used in photovoltaic systems to evaluate the behavior of photovoltaic systems after encountering large-scale disturbances such as sudden weather changes and equipment failures. Through dynamic simulation in the time domain, the process of photovoltaic systems transitioning from one stable state to another possible stable state is examined.

9. The method for online evaluation of transient stability of a photovoltaic unit according to claim 1, characterized in that: In step S7, the statistical and machine learning methods include time series analysis, regression model analysis, machine learning methods and integrated learning methods; Time series analysis to predict future performance indicators such as power output and efficiency; Regression model analysis is used to predict or estimate the relationship between a response variable, such as photovoltaic power output, and one or more explanatory variables, such as temperature and light intensity; Machine learning methods to learn from historical data and predict the stability and performance of photovoltaic cells; Ensemble learning methods, which improve prediction accuracy and robustness by combining multiple models, are used to predict the stability and performance of photovoltaic cells.

10. A photovoltaic unit transient stability online evaluation system, constructed based on the photovoltaic unit transient stability online evaluation method according to any one of claims 1 to 9, characterized in that: It includes data collection module, basic model building module, real-time monitoring deployment module, data preprocessing module, anomaly detection module, simulation calibration module, real-time performance evaluation module, transient stability analysis module, trend analysis module and risk assessment module; The system also includes a report generation module and a continuous optimization module; The report generation module is used to formulate standardized report templates and regularly generate performance and stability evaluation reports for reference by system operation and maintenance and management personnel; The continuous optimization module is used to establish a continuous feedback loop to continuously optimize the monitoring, analysis and response strategies based on the evaluation results and the operating experience of the photovoltaic unit.

Images