Distributed Photovoltaic Energy Efficiency Dynamic Monitoring Method Based on Time Series Analysis

Abstract

The present application provides a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis, which relates to the technical field of photovoltaic power generation and includes: deploying edge computing nodes, regularly receiving the electrical operation parameters of photovoltaic inverters, and recording monitoring nodes; predicting the photovoltaic output power according to the electrical operation parameters to obtain the predicted output power; analyzing the power supply log to obtain the actual output power under the monitoring nodes, forming energy efficiency analysis data based on the monitoring nodes, the predicted output power and the actual output power, and storing it in the energy efficiency analysis database; using the energy efficiency analysis database to perform power prediction deviation analysis and constructing a time period - deviation comparison table; when the distributed photovoltaic equipment performs grid connection scheduling, performing power output compensation based on the time period - deviation comparison table. The present application solves the technical problem that the traditional time series analysis method lacks precise analysis for different time periods, resulting in insufficient power prediction accuracy, and improves the accuracy of power prediction and the quality of grid connection scheduling.

Description

Technical Field

[0001] This application relates to the technical field of photovoltaic power generation, and particularly to a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis. Background Art

[0002] Distributed photovoltaic systems, as an important part of clean energy, are increasingly becoming a key driving force for the global energy transition. Distributed photovoltaic systems can generate electricity directly at the user side, reducing transmission losses, and at the same time support grid-connected power generation, optimizing the energy structure of the power system. However, photovoltaic power generation is vulnerable to factors such as weather and time, and the power generation power fluctuates greatly. Accurately predicting the power output of distributed photovoltaics is not only crucial for optimizing the operation efficiency of the photovoltaic system itself, but also a key step in ensuring the stable and efficient operation of the power system.

[0003] Time series analysis methods are often used in the dynamic monitoring of distributed photovoltaic energy efficiency. By analyzing historical data, the changing trend of photovoltaic output power over time is analyzed, and future power is predicted and grid-connected scheduling is carried out based on these trends. However, the current time series analysis methods mainly focus on the prediction of macro trends and lack refined analysis of power fluctuations in short time periods. This coarseness in time period analysis makes it difficult to capture the immediate impact of rapidly changing weather conditions on photovoltaic power, thereby affecting the accuracy of power prediction. In grid-connected scheduling, relying on such rough predictions for power output adjustment not only fails to respond to grid demands in a timely manner, but may also cause grid fluctuations due to large deviations between predictions and actual outputs, affecting the overall operation quality and stability of the grid. Summary of the Invention

[0004] This application provides a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis, which solves the technical problem that traditional time series analysis methods usually perform power prediction based on historical data and macro trends, lack precise analysis for different time periods, resulting in insufficient power prediction accuracy, and thus affecting the accuracy of grid-connected scheduling. It achieves the technical effect of improving the accuracy of photovoltaic power generation prediction to improve the quality of grid-connected scheduling and enhance the stability of grid operation.

[0005] In view of the above problems, the present application provides a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis. The method includes: deploying edge computing nodes at distributed photovoltaic devices, regularly receiving electrical operation parameters of photovoltaic inverters, and recording monitoring nodes; predicting photovoltaic output power according to the electrical operation parameters to obtain predicted output power; analyzing power supply logs to obtain actual output power under the monitoring nodes, and forming energy efficiency analysis data based on the monitoring nodes, predicted output power, and actual output power, and storing the data in an energy efficiency analysis database, where the energy efficiency analysis database is embedded in the edge computing nodes; using the energy efficiency analysis database to perform power prediction deviation analysis according to multiple energy efficiency analysis data, and constructing a time period - deviation comparison table; when the distributed photovoltaic devices perform grid connection scheduling, performing power output compensation based on the time period - deviation comparison table.

[0006] Preferably, the electrical operation parameters at least include DC input voltage, DC input current, power factor, and inverter temperature, where the inverter temperature is monitored by a temperature sensor deployed on the photovoltaic inverter.

[0007] Preferably, predicting photovoltaic output power according to the electrical operation parameters to obtain predicted output power includes: using the device attribute characteristics and service life of the photovoltaic inverter as device constraints, using the electrical operation parameters as keyword constraints, and using output power prediction as a guide, retrieving a sample data set based on grid big data; performing differential screening on the sample data set according to a predetermined strategy to obtain a sample training set, and using the sample training set to perform supervised training on a BP neural network to obtain a power predictor; using the power predictor to predict photovoltaic output power for the electrical operation parameters to obtain the predicted output power.

[0008] Preferably, performing differential screening on the sample data set according to a predetermined strategy to obtain a sample training set includes: randomly selecting first sample data in the sample data set, where the first sample data includes a first voltage, a first current, a first power factor, a first temperature, and a first power; performing deviation analysis on the first voltage and other voltages in the sample data set respectively, and counting the data volume with a deviation greater than a predetermined voltage threshold, denoted as a first random coefficient; if the first random coefficient is greater than a coefficient scalar, then performing deviation analysis on the first current, the first power factor, and the first temperature in sequence to obtain a second random coefficient, a third random coefficient, and a fourth random coefficient; if the second random coefficient, the third random coefficient, and the fourth random coefficient are all greater than the coefficient scalar, then adding the first sample data to the sample training set.

[0009] Preferably, power prediction deviation analysis is performed based on multiple energy efficiency analysis data, and a time period - deviation comparison table is constructed, including: dividing the multiple energy efficiency analysis data according to a predetermined period, performing time alignment and data extraction on the division results to obtain multiple energy efficiency data sets at multiple identical nodes, where the predetermined period is 24 hours; performing power prediction deviation analysis on multiple monitoring nodes respectively according to the multiple energy efficiency data sets to determine multiple prediction deviation coefficients; and constructing the time period - deviation comparison table according to the multiple monitoring nodes and multiple prediction deviation coefficients.

[0010] Preferably, determining multiple prediction deviation coefficients includes: randomly selecting a first monitoring node and a first energy efficiency data set of the first monitoring node; calculating power prediction deviation according to the first energy efficiency data set to obtain multiple first prediction deviations, where the first prediction deviation is the ratio of the difference between the actual output power and the predicted output power to the actual output power; calculating the mean value of the multiple first prediction deviations to obtain a first prediction deviation coefficient, and adding it to the multiple prediction deviation coefficients.

[0011] Preferably, constructing the time period - deviation comparison table according to the multiple monitoring nodes and multiple prediction deviation coefficients includes: arranging the multiple monitoring nodes in the order of the nodes to construct a monitoring node sequence; selecting a first deviation coefficient of a first node and a second deviation coefficient of a second node, and calculating a first coefficient difference between the first deviation coefficient and the second deviation coefficient, where the first node is the initial node of the monitoring node sequence and the second node is the adjacent node of the first node; if the first coefficient difference is less than a predetermined coefficient threshold, then merging the first node and the second node to construct a first time period, where the deviation coefficient of the first time period is the mean value of the first deviation coefficient and the second deviation coefficient; if the first coefficient difference is greater than or equal to the predetermined coefficient threshold, then setting a first node dividing line and performing iterative analysis until all nodes of the monitoring node sequence are analyzed to obtain multiple node dividing lines; determining multiple monitoring time periods and multiple deviation coefficients corresponding to the multiple monitoring time periods based on the multiple node dividing lines, and constructing the time period - deviation comparison table according to the multiple monitoring time periods and multiple deviation coefficients.

[0012] Preferably, performing power output compensation based on the time period - deviation comparison table includes: receiving real - time electrical operation parameters of a current node, using the power predictor to perform photovoltaic output power prediction on the real - time electrical operation parameters to obtain a real - time predicted output power; inputting the current node into the time period - deviation comparison table for matching to output an adapted deviation coefficient; compensating the real - time predicted output power according to the adapted deviation coefficient to obtain a real - time output power, and performing grid - connection scheduling of the distributed photovoltaic device.

[0013] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0014] By deploying edge computing nodes at distributed photovoltaic devices, regularly receiving the electrical operation parameters of photovoltaic inverters, recording monitoring nodes, and performing local data processing, data transmission delay and cost are reduced, and the real-time performance and response speed of data processing are improved, providing fast and accurate basic data for subsequent power prediction. According to the electrical operation parameters, photovoltaic output power prediction is performed to obtain the predicted output power, and the power demand and power generation situation are predicted, providing data reference for subsequent deviation analysis and compensation. Analyze the power supply log to obtain the actual output power under the monitoring node, and based on the monitoring node, predicted output power, and actual output power, form energy efficiency analysis data and store it in the energy efficiency analysis database to monitor the operation status of distributed photovoltaic devices, providing complete and continuous data for subsequent deviation analysis. Using the energy efficiency analysis database, perform power prediction deviation analysis based on multiple energy efficiency analysis data, construct a time period - deviation comparison table, capture the prediction error characteristics in different time periods, refine the power error analysis, and identify the power fluctuations within a short time, providing a key basis for subsequent power output compensation. When the distributed photovoltaic device is connected to the grid for scheduling, perform power output compensation based on the time period - deviation comparison table to eliminate the impact of prediction errors on grid-connected power generation, ensure the stability of the power system, reduce the power fluctuations during the grid connection process, and improve the quality and stability of grid connection scheduling.

[0015] In summary, this application combines edge computing, power prediction, and actual power monitoring to accurately capture the power prediction errors in different time periods, and through dynamic output power compensation, reduces the grid fluctuations caused by power prediction errors. This solution significantly improves the accuracy and real-time performance of the energy efficiency dynamic monitoring of the distributed photovoltaic system, enhances the flexibility and stability of grid connection scheduling, strengthens the fast response ability of the distributed photovoltaic power generation system to grid demands, and ensures the efficient operation of the distributed photovoltaic system.

[0016] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0017] Figure 1 It is a schematic flowchart of the method for dynamically monitoring the energy efficiency of a distributed photovoltaic based on time series analysis provided by an embodiment of this application;

[0018] Figure 2 It is a schematic flowchart of obtaining the predicted output power in the method for dynamically monitoring the energy efficiency of a distributed photovoltaic based on time series analysis provided by an embodiment of this application;

[0019] Figure 3 This is a schematic flow chart for constructing a time period - deviation comparison table in the distributed photovoltaic energy efficiency dynamic monitoring method provided by the embodiments of the present application based on time series analysis. Detailed implementation manners

[0020] By providing a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis, the embodiments of the present application solve the technical problem that traditional time series analysis methods usually perform power prediction based on historical data and macro trends, lack precise analysis for different time periods, resulting in insufficient power prediction accuracy, and further affecting the accuracy of grid connection scheduling, and achieve the technical effect of improving the accuracy of photovoltaic power prediction to improve the quality of grid connection scheduling and enhance the stability of power grid operation.

[0021] As Figure 1 shown, the embodiments of the present application provide a distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis, and the method includes:

[0022] Step S1: Deploy edge computing nodes at the distributed photovoltaic devices, regularly receive the electrical operation parameters of the photovoltaic inverters, and record the monitoring nodes.

[0023] Specifically, an edge computing node is a computing infrastructure deployed close to the data source, responsible for real - time processing, analysis, and storage of data, thereby reducing data transmission latency. Compared with cloud computing, edge computing can process data locally and reduce the response time. In the embodiments of the present application, the edge computing device can be a small computing unit installed in the photovoltaic system, such as a single - board computer or a customized industrial computing device. First, deploy multiple edge computing nodes near the distributed photovoltaic devices. Through these edge computing nodes, regularly receive the electrical operation parameters provided by the photovoltaic inverters and mark the monitoring time points. Among them, the photovoltaic inverter is responsible for converting the direct current generated by the photovoltaic modules into alternating current, and its electrical parameters include voltage, current, power factor, inverter temperature, etc.

[0024] Through step S1, data can be quickly collected and analyzed locally, reducing the dependence on the remote cloud platform, reducing the data transmission latency, and providing real - time and accurate data support for subsequent photovoltaic power prediction and energy efficiency analysis.

[0025] Step S2: Perform photovoltaic output power prediction according to the electrical operation parameters to obtain the predicted output power.

[0026] Specifically, the prediction of photovoltaic output power is mainly based on training a prediction model with historical data, such as linear regression, random forest, neural network, etc. The previously obtained electrical operation parameters are input into the pre-trained prediction model to obtain the predicted output power for a future period of time, and the power generation capacity of photovoltaic equipment is estimated in advance to provide reference data for subsequent grid connection scheduling.

[0027] Step S3: Analyze the power supply log to obtain the actual output power under the monitoring node. Based on the monitoring node, predicted output power, and actual output power, form energy efficiency analysis data and store it in the energy efficiency analysis database, where the energy efficiency analysis database is embedded in the edge computing node.

[0028] Specifically, the power supply log records the power supply data during the operation of the photovoltaic power generation system, including information such as actual output power and timestamp. By analyzing the power supply log, the actual output power data under the same monitoring node as the predicted output power data in Step S2 is extracted. Combine each monitoring node with its corresponding predicted output power and actual output power to form a set of energy efficiency analysis data. This energy efficiency analysis data not only includes the predicted and actual output powers, but also auxiliary information such as timestamp and device status, providing comprehensive data support for subsequent energy efficiency analysis. These data reflect the performance of the photovoltaic system and the accuracy of the prediction model. Store these energy efficiency analysis data in the energy efficiency analysis database, which is embedded in the edge computing node, to facilitate data processing and storage near the device, and to respond, analyze, and adjust the power generation strategy in a timely manner. The energy efficiency analysis database can not only store data for a single period, but also accumulate historical records over a long time to help with more accurate power prediction and deviation analysis in the future.

[0029] Step S4: Use the energy efficiency analysis database to conduct power prediction deviation analysis based on multiple energy efficiency analysis data and construct a time period - deviation comparison table.

[0030] Specifically, extract the energy efficiency analysis data of multiple monitoring nodes from the energy efficiency analysis database, and these data include predicted output power and actual output power. Group the extracted data according to time information (such as time periods of hours, days, months, etc.) to analyze the prediction deviations in different time periods. For each time period, calculate the deviation between the predicted output power and the actual output power. Organize the calculated deviations according to time periods to construct a comparison table, that is, the time period - deviation comparison table, which records the power prediction deviation situations under different time periods and provides a reference basis for power generation scheduling and compensation decisions. Analyze the data in the comparison table to identify the regularity and periodicity of the deviations. Through systematic deviation analysis, the power prediction deviations can be more accurately identified and quantified, providing an important basis for optimizing the grid connection scheduling of the photovoltaic system, reducing power fluctuations caused by prediction errors, and improving the overall power generation efficiency and grid stability.

[0031] Step S5: When the distributed photovoltaic device performs grid connection scheduling, perform power output compensation based on the time period - deviation comparison table.

[0032] Specifically, when the distributed photovoltaic device performs grid connection scheduling, power output compensation refers to adjusting the output power of the photovoltaic device according to the real - time demand of the power grid and the characteristics of distributed photovoltaic power generation to ensure the stability of the power grid and the power quality. This compensation mechanism can reduce the volatility and uncertainty of photovoltaic power generation and avoid impacting the power grid.

[0033] When performing grid connection scheduling, key parameters such as the load demand, frequency, and voltage of the power grid are monitored in real - time to timely understand the operating conditions of the power grid. Using the time period - deviation comparison table constructed in step S4, according to the time period in which the current time is located, query the corresponding predicted deviation coefficient. Adjust the output power of the photovoltaic device according to the deviation coefficient to compensate for the predicted deviation and ensure that the photovoltaic power generation is more matched with the power grid demand. Through this compensation mechanism, the accuracy and reliability of grid connection scheduling can be greatly improved, and stable power output can be achieved.

[0034] Furthermore, in step S1 of the embodiment of the present application, the electrical operating parameters at least include the DC input voltage, DC input current, power factor, and inverter temperature, where the inverter temperature is monitored by a temperature sensor deployed on the photovoltaic inverter.

[0035] Specifically, in step S1, the electrical operating parameters to be monitored at least include the DC input voltage, DC input current, power factor, and inverter temperature. The DC input voltage and DC input current are the core indicators of the photovoltaic system, directly reflecting the power generated by the solar panels. The fluctuations of voltage and current can indicate the working state of the photovoltaic panels. These parameters may change drastically during weather changes or equipment failures. For example, on sunny days, the photovoltaic panels may generate higher voltage and current, and conversely, the voltage and current will decrease significantly on cloudy days. The power factor is used for the power utilization efficiency of the distributed photovoltaic power generation system. If the power factor decreases, there may be reactive power consumption in the inverter, and adjustments are needed to improve the overall power generation efficiency.

[0036] The inverter temperature is an important indicator for the stable operation of the equipment. Excessive temperature may cause the performance of the inverter to decline or even be damaged. Therefore, by deploying a temperature sensor on the inverter, the inverter temperature is obtained in real - time to detect the operating state of the photovoltaic device. The temperature sensor can measure the temperature condition of the inverter in real - time through a thermistor integrated inside the inverter or an externally installed temperature probe. When the working temperature of the inverter exceeds a certain threshold, cooling measures may need to be taken or the power output may need to be reduced to avoid equipment damage.

[0037] For example, in a distributed photovoltaic system installed on the roof of a factory, the DC input voltage of the inverter is 400V, the DC current is 12A, the power factor is 0.98, and the inverter temperature is 60°C. In this state, the photovoltaic system is operating normally and efficiently. However, if the inverter temperature suddenly rises to 80°C, the photovoltaic system may activate an automatic protection mechanism to reduce power output to avoid overheating.

[0038] Furthermore, as Figure 2 shown, step S2 of the embodiment of the present application further includes:

[0039] Taking the device attribute characteristics and device service life of the photovoltaic inverter as device constraints, taking the electrical operation parameters as keyword constraints, and taking the output power prediction as a guide, retrieve a sample data set based on the power grid big data; screen the sample data set differently according to a predetermined strategy to obtain a sample training set, and use the sample training set to supervise and train a BP neural network to obtain a power predictor; use the power predictor to predict the photovoltaic output power of the electrical operation parameters to obtain the predicted output power.

[0040] Specifically, the device attribute characteristics refer to the basic technical parameters and performance indicators of the photovoltaic inverter, such as rated power, efficiency, rated input voltage and current, etc. These characteristics help to identify the type and applicable range of the inverter. The device service life refers to the time since the photovoltaic inverter was installed. This factor affects the performance and efficiency of the inverter. Generally, older devices have lower efficiency. First, screen according to the device attribute characteristics and service life of the photovoltaic inverter in the power grid big data. Then, use the obtained electrical operation parameters as keyword constraints to retrieve a sample data set of similar devices and similar operation parameters from the power grid big data. This sample data set contains historical output power data under similar device attribute characteristics and service life but different electrical operation parameters, which is used for the training and testing of the power predictor.

[0041] Then, according to a predetermined strategy, differentially screen the sample data set to determine the data set for the power predictor, that is, the sample training set. Use these sample training sets to supervise and train the BP neural network. The BP neural network is a machine learning model that can learn complex non-linear relationships. Through the backpropagation algorithm, the network will continuously adjust its internal weights to improve the prediction ability of the photovoltaic power generation. During the training process, set the structure of the BP neural network (input layer, hidden layer, output layer), randomly initialize the weights and biases, input the historical electrical operation parameters, and calculate the predicted output power through the neural network. Compare the predicted value with the actual value and calculate the loss function. According to the loss function, adjust the weights and biases backward to reduce the prediction error. Repeat the forward and backward propagation processes until the model converges.

[0042] The trained BP neural network is the power predictor, which is used to predict the photovoltaic output power according to the input electrical operating parameters. This predictor can be applied to new electrical operating parameters in real time for photovoltaic output power prediction. When new electrical parameters are input, the power predictor can quickly calculate the expected photovoltaic output power under the current conditions.

[0043] Through the above steps, based on the sample data set with equipment constraints and keyword constraints, combined with the supervised training of the BP neural network, a high-precision power predictor can be obtained to provide accurate photovoltaic output efficiency prediction data.

[0044] Furthermore, in step S2 of the embodiment of the present application, the sample data set is differentially screened according to a predetermined strategy to obtain a sample training set, including:

[0045] Randomly select the first sample data in the sample data set, where the first sample data includes the first voltage, the first current, the first power factor, the first temperature, and the first power; perform deviation analysis on the first voltage and other voltages in the sample data set respectively, and count the data volume with a deviation greater than a predetermined voltage threshold, denoted as the first random coefficient; if the first random coefficient is greater than the coefficient scalar, then perform deviation analysis on the first current, the first power factor, and the first temperature in sequence to obtain the second random coefficient, the third random coefficient, and the fourth random coefficient; if the second random coefficient, the third random coefficient, and the fourth random coefficient are all greater than the coefficient scalar, then add the first sample data to the sample training set.

[0046] Specifically, before training the model, it is necessary to differentially screen the sample data set to improve the difference of the training data, thereby saving the training time, enabling the BP neural network to quickly converge, and improving the training quality. First, randomly select a set of the first sample data in the sample data set, and this data includes electrical operating parameters such as the first voltage, the first current, the first power factor, the first temperature, and the first power.

[0047] Next, perform deviation analysis on the first voltage: compare the first voltage of the first sample data with all other voltage values in the sample data set, and count the data volume with a deviation greater than a predetermined voltage threshold, and this data volume is called the first random coefficient. Among them, the predetermined voltage threshold is a preset acceptable voltage deviation range, usually a specific value, and the data exceeding this range will be regarded as having a large deviation. For example, if the current first voltage is 350V and the predetermined voltage threshold is 10V, find all voltage values in the sample data set and count the number of voltage data with a difference from 350V exceeding 10V, and use this number as the first random coefficient.

[0048] If the first random coefficient is greater than a preset coefficient scalar, the same deviation analysis is performed on other electrical parameters. First, a deviation analysis is performed on the first current to obtain a second random coefficient. If the second random coefficient is greater than the preset coefficient scalar, a power deviation analysis is continued. A deviation analysis is performed on the first power factor to obtain a third random coefficient; if the third random coefficient is greater than the preset coefficient scalar, a temperature deviation analysis is continued. A deviation analysis is performed on the first temperature to obtain a fourth random coefficient. If the fourth random coefficient is greater than the preset coefficient scalar, the first sample data is added to the sample training set. The above deviation analysis process is similar to the voltage deviation analysis. Refer to the voltage deviation analysis for details. Among them, the second random coefficient, the third random coefficient, and the fourth random coefficient respectively correspond to the deviation analysis results of the first current, the first power factor, and the first temperature, and are used to comprehensively evaluate the overall stability of the current sample. The coefficient scalar is a preset value used to compare with the random coefficient to determine whether the sample meets the conditions for inclusion in the training set.

[0049] If any one of the first random coefficient, the second random coefficient, the third random coefficient, and the fourth random coefficient is less than or equal to the system scalar, the deviation analysis is terminated, and the first sample data is removed from the dataset. A new set of sample data is reselected and the above steps are repeated until all the sample data in the sample dataset are traversed, and finally the sample training set is obtained.

[0050] Through the above differential screening strategy, the diversity and representativeness of the sample training set can be ensured, the overfitting problem during model training can be avoided, the generalization ability and prediction accuracy of the model can be improved, and high-quality training data can be provided for the subsequent training of the power predictor based on the BP neural network.

[0051] Furthermore, as Figure 3 shown, step S4 of the embodiment of the present application further includes:

[0052] The multiple energy efficiency analysis data are divided according to a predetermined period, and time alignment and data extraction are performed on the division results to obtain multiple energy efficiency datasets at multiple identical nodes, where the predetermined period is 24 hours; according to the multiple energy efficiency datasets, power prediction deviation analysis is respectively performed on multiple monitoring nodes to determine multiple prediction deviation coefficients; and a time period - deviation comparison table is constructed according to the multiple monitoring nodes and the multiple prediction deviation coefficients.

[0053] Specifically, first, divide multiple energy efficiency analysis data according to a predetermined period, i.e., 24 hours. Organize and process the photovoltaic power generation data of each day as an independent period. Align the divided energy efficiency analysis data according to the detection nodes to avoid analysis errors caused by time deviation and ensure the consistency and comparability of data in different time periods. Subsequently, extract the energy efficiency data of specific nodes from each time period, including the predicted output power and the actual output power, to form multiple energy efficiency data sets. For example, extract the energy efficiency data at 9:00 within a continuous week to obtain 7 energy efficiency data sets.

[0054] For each monitoring node, use multiple energy efficiency data sets extracted from the database to conduct power prediction deviation analysis. Calculate the deviation between the predicted output power and the actual output power. The deviation calculation can adopt various methods such as absolute deviation, relative deviation, or root mean square error. According to the deviation statistical results, determine the prediction deviation coefficient of each monitoring node. The prediction deviation coefficient reflects the consistency level between the predicted output and the actual output of this node and can be used as a quantitative index to evaluate the prediction performance of the model.

[0055] Summarize the prediction deviation coefficients of all monitoring nodes and associate them with the corresponding time period information. According to the summary information, construct a time period - deviation comparison table. This table clearly shows the deviation situation between the predicted output power and the actual output power under different monitoring nodes and different time periods. By analyzing the comparison table, the monitoring nodes and time periods with large deviations can be identified, and then targeted scheduling compensation can be carried out.

[0056] Furthermore, determining multiple prediction deviation coefficients in step S4 of the embodiment of the present application includes:

[0057] Randomly select a first monitoring node and the first energy efficiency data set of the first monitoring node; calculate the power prediction deviation according to the first energy efficiency data set to obtain multiple first prediction deviations, where the first prediction deviation is the ratio of the difference between the actual output power and the predicted output power to the actual output power; calculate the mean of the multiple first prediction deviations to obtain the first prediction deviation coefficient and add it to the multiple prediction deviation coefficients.

[0058] Specifically, randomly select a node from multiple monitoring nodes as the first monitoring node, and extract multiple energy efficiency data sets of this node as the first energy efficiency data set. Each data set contains the predicted output power and actual output power data of this node in different periods.

[0059] For each dataset in the first energy efficiency dataset, calculate the power prediction deviation according to the formula: First prediction deviation = (actual output power - predicted output power) / actual output power, to obtain multiple first prediction deviations. Then, calculate the average value of the multiple first prediction deviations to obtain a value representing the overall prediction accuracy of the node, that is, the first prediction deviation coefficient. If the coefficient value is high, it indicates that the prediction error of the node is large and the prediction accuracy of the model is low; conversely, if the coefficient value is low, it indicates that the prediction of the node is relatively accurate. Add this first prediction deviation coefficient to the set of multiple prediction deviation coefficients. This set contains the prediction deviation coefficients of multiple monitoring nodes and is used for subsequent system optimization and scheduling decisions.

[0060] Through the above steps, not only a specific prediction deviation quantification index is provided for the selected node, but also data support is provided for comparing the prediction performance of different monitoring nodes in subsequent analysis, which helps to identify potential improvement points of the prediction model and optimize the operation strategy of the photovoltaic system.

[0061] Further, in step S4 of the embodiment of the present application, constructing the time period - deviation comparison table according to the multiple monitoring nodes and multiple prediction deviation coefficients includes:

[0062] Arrange the multiple monitoring nodes in the order of the nodes to construct a monitoring node sequence; select the first deviation coefficient of the first node and the second deviation coefficient of the second node, and calculate the first coefficient difference between the first deviation coefficient and the second deviation coefficient, where the first node is the initial node of the monitoring node sequence and the second node is the adjacent node of the first node; if the first coefficient difference is less than the predetermined coefficient threshold, then merge the first node and the second node to construct the first time period, where the deviation coefficient of the first time period is the average value of the first deviation coefficient and the second deviation coefficient; if the first coefficient difference is greater than or equal to the predetermined coefficient threshold, set the first node dividing line and perform iterative analysis until all nodes in the monitoring node sequence are analyzed, to obtain multiple node dividing lines; determine multiple monitoring time periods and multiple deviation coefficients corresponding to the multiple monitoring time periods based on the multiple node dividing lines, and construct the time period - deviation comparison table according to the multiple monitoring time periods and multiple deviation coefficients.

[0063] Specifically, a list of monitoring nodes arranged in the order of nodes forms a monitoring node sequence. The first node (i.e., the initial node) in the sequence and its adjacent second node are selected, and the first deviation coefficient of the first node and the difference between the second deviation coefficient of the second node are calculated, denoted as the first coefficient difference. If the first coefficient difference is less than a predetermined coefficient threshold, the first node and the second node are merged into the same time period, denoted as the first time period, and the deviation coefficient of this time period is the average of the first deviation coefficient and the second deviation coefficient. Among them, the predetermined coefficient threshold is a set critical value used to determine whether the difference in deviation coefficients between two monitoring nodes is close enough.

[0064] If the first coefficient difference is greater than or equal to the predetermined coefficient threshold, a node dividing line is set between the first node and the second node, and the analysis of the first node is marked as ended. Subsequently, taking the second node as the new starting node, following the same steps as above, a new coefficient difference is calculated and it is determined whether to merge time periods, and iterative analysis is carried out until all nodes in the sequence have been analyzed.

[0065] Through the above iterative process, multiple node dividing lines, as well as multiple monitoring time periods and their corresponding deviation coefficients, can be determined, thereby dividing the entire monitoring node sequence into multiple time periods with similar deviation characteristics. Based on the monitoring time periods determined by multiple node dividing lines and the deviation coefficients corresponding to each time period, a time period - deviation comparison table is constructed to visually display the predicted deviation conditions in different time periods.

[0066] Exemplarily, there are 5 monitoring nodes numbered from 001 to 005, and the corresponding predicted deviation coefficients are 0.02, 0.03, 0.04, 0.06, and 0.07 respectively, and the predetermined coefficient threshold is 0.015.

[0067] Starting from node 001 (deviation coefficient 0.02), comparing with node 002 (deviation coefficient 0.03), the first coefficient difference is 0.01, which is less than the threshold, so node 001 and node 002 are merged to form the first time period, and the time period deviation coefficient is 0.025.

[0068] Next, comparing with the subsequent node of node 002 (now the merged node of the first time period, deviation coefficient 0.025), that is, node 003 (deviation coefficient 0.04). The calculated coefficient difference is 0.015, which is equal to the threshold, so the first time period and node 003 can be merged, and the deviation coefficient of the first time period is updated to (0.02 + 0.03 + 0.04) / 3 = 0.03.

[0069] Compare with the subsequent nodes starting from node 003 (now the node after the first period merger, deviation coefficient 0.03), that is, node 004 (deviation coefficient 0.06). The calculated coefficient difference is 0.03, which is greater than the threshold. Therefore, a node dividing line is set between node 003 and node 004. Taking node 004 (deviation coefficient 0.06) as the new starting node, compare it with node 005 (deviation coefficient 0.07). The coefficient difference is 0.01, which is less than the threshold. Merge node 004 and node 005 to form the second period, and the period deviation coefficient is (0.06 + 0.07) / 2 = 0.065. Through the above steps, finally, the 5 monitoring nodes are divided into two periods, forming a period - deviation comparison table:

[0070] Further, step S5 of the embodiment of the present application further includes:

[0071] Receive the real - time electrical operation parameters of the current node, use the power predictor to predict the photovoltaic output power of the real - time electrical operation parameters, and obtain the real - time predicted output power; input the current node into the period - deviation comparison table for matching, and output the adapted deviation coefficient; compensate the real - time predicted output power according to the adapted deviation coefficient to obtain the real - time output power, and execute the grid - connection scheduling of the distributed photovoltaic equipment.

[0072] Specifically, when the distributed photovoltaic equipment performs grid - connection scheduling, first, the edge computing node receives the real - time electrical operation parameters of the current node, including the real - time DC input voltage, real - time DC input current, real - time power factor, and real - time inverter temperature. Input the real - time electrical operation parameters into the trained power predictor for photovoltaic output power prediction to obtain the real - time predicted output power. According to the current time node, query the period - deviation comparison table to find the adapted deviation coefficient for the corresponding period. This coefficient is selected from the period - deviation comparison table according to the current period and is the deviation coefficient that best matches the current period, reflecting the deviation degree between the predicted value and the actual value in the current period and is used to compensate for the prediction error. Use the adapted deviation coefficient to adjust the real - time predicted output power, by increasing or decreasing the output power to compensate for the prediction deviation. The adjusted power is the real - time output power. According to the real - time output power, adjust the output of the photovoltaic equipment to ensure that it matches the power supply and demand of the power grid, and at the same time optimize the operation efficiency and stability of the power grid.

[0073] Exemplarily, the current time is 12:00 noon. The electrically parameters monitored in real time show that the voltage is 380V, the current is 15A, the power factor is 0.98, and the temperature is 45°C. These data are input into the power predictor, and the predicted current power generation is 120kW. By referring to the time period - deviation comparison table, it is found that the deviation coefficient for the period from 12:00 to 13:00 is 3%. Then the distributed photovoltaic system increases the predicted power by 3% to obtain the adjusted real - time output power of 123.6kW. The 123.6kW is used as the power value for the current grid connection scheduling to guide the grid connection of photovoltaic equipment and ensure the stable operation of the power grid.

[0074] Through the above steps, the real - time power prediction and compensation of distributed photovoltaic equipment in grid connection scheduling are achieved, which helps to improve the accuracy of power dispatching and the grid connection quality, and ensure the stable operation of the power grid.

[0075] In summary, the distributed photovoltaic energy efficiency dynamic monitoring method based on time - series analysis provided by the embodiments of this application has the following technical effects:

[0076] By deploying edge computing nodes at distributed photovoltaic equipment, regularly receiving the electrical operation parameters of photovoltaic inverters, recording monitoring nodes, and performing local data processing, the data transmission delay and cost are reduced, and the real - time performance and response speed of data processing are improved, providing fast and accurate basic data for subsequent power prediction. According to the electrical operation parameters, the photovoltaic output power is predicted to obtain the predicted output power, and the power demand and power generation situation are predicted, providing data reference for subsequent deviation analysis and compensation. Analyze the power supply log to obtain the actual output power under the monitoring node, and based on the monitoring node, predicted output power and actual output power, form energy efficiency analysis data and store it in the energy efficiency analysis database to monitor the operation status of distributed photovoltaic equipment, providing complete and continuous data for subsequent deviation analysis. Using the energy efficiency analysis database, perform power prediction deviation analysis according to multiple energy efficiency analysis data, construct a time period - deviation comparison table, capture the prediction error characteristics in different time periods, refine the power error analysis, identify the power fluctuations within a short period of time, providing a key basis for subsequent power output compensation. When the distributed photovoltaic equipment is connected to the grid for scheduling, based on the time period - deviation comparison table, perform power output compensation to eliminate the impact of prediction errors on grid - connected power generation, ensure the stability of the power system, reduce the power fluctuations during the grid connection process, and improve the quality and stability of grid connection scheduling.

[0077] Overall, the embodiments of the present application combine edge computing, power prediction, and actual power monitoring to accurately capture the power prediction errors in different time periods, and reduce the grid fluctuations caused by power prediction errors through dynamic output power compensation. This solution significantly improves the accuracy and real-time performance of the energy efficiency dynamic monitoring of distributed photovoltaic systems, enhances the flexibility and stability of grid connection scheduling, strengthens the fast response ability of distributed photovoltaic power generation systems to grid demands, and ensures the efficient operation of distributed photovoltaic systems.

[0078] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A distributed photovoltaic energy efficiency dynamic monitoring method based on time series analysis, characterized in that, Including: Deploy edge computing nodes at distributed photovoltaic devices, regularly receive the electrical operation parameters of photovoltaic inverters, and record monitoring nodes; Perform photovoltaic output power prediction based on the electrical operation parameters to obtain predicted output power; Analyze the power supply log to obtain the actual output power under the monitoring node, and form energy efficiency analysis data based on the monitoring node, predicted output power and actual output power, and store it in the energy efficiency analysis database, where the energy efficiency analysis database is embedded in the edge computing node; Utilize the energy efficiency analysis database to perform power prediction deviation analysis based on multiple energy efficiency analysis data, and construct a time period - deviation comparison table; When the distributed photovoltaic device performs grid connection scheduling, perform power output compensation based on the time period - deviation comparison table; Performing photovoltaic output power prediction based on the electrical operation parameters to obtain predicted output power, including: Taking the device attribute characteristics and device usage years of the photovoltaic inverter as device constraints, taking the electrical operation parameters as keyword constraints, and taking the output power prediction as a guide, retrieve the sample data set based on the power grid big data; Differentially screen the sample data set according to a predetermined strategy to obtain a sample training set, and use the sample training set to perform supervised training on a BP neural network to obtain a power predictor; Use the power predictor to perform photovoltaic output power prediction on the electrical operation parameters to obtain the predicted output power; Differentially screen the sample data set according to a predetermined strategy to obtain a sample training set, including: Randomly select the first sample data in the sample data set, where the first sample data includes the first voltage, first current, first power factor, first temperature and first power; Perform deviation analysis on the first voltage and other voltages in the sample data set respectively, and count the data volume with deviation greater than the predetermined voltage threshold, denoted as the first random coefficient; If the first random coefficient is greater than the coefficient scalar, then perform deviation analysis on the first current, first power factor and first temperature in sequence to obtain the second random coefficient, third random coefficient and fourth random coefficient; If the second random coefficient, third random coefficient and fourth random coefficient are all greater than the coefficient scalar, then add the first sample data to the sample training set; Performing power prediction deviation analysis based on multiple energy efficiency analysis data to construct a time period - deviation comparison table, including: Divide the multiple energy efficiency analysis data according to a predetermined period, perform time alignment and data extraction on the division results to obtain multiple energy efficiency data sets under multiple same nodes, where the predetermined period is 24 hours; According to the multiple energy efficiency data sets, perform power prediction deviation analysis on multiple monitoring nodes respectively to determine multiple prediction deviation coefficients; Construct the time period - deviation comparison table according to the multiple monitoring nodes and multiple prediction deviation coefficients; Determining multiple prediction deviation coefficients, including: Randomly select the first monitoring node and the first energy efficiency data set of the first monitoring node; Calculate the power prediction deviation according to the first energy efficiency dataset to obtain a plurality of first prediction deviations, where the first prediction deviation is the ratio of the difference between the actual output power and the predicted output power to the actual output power; Calculate the mean of the plurality of first prediction deviations to obtain a first prediction deviation coefficient and add it to the plurality of prediction deviation coefficients; Construct the time period - deviation comparison table according to the plurality of monitoring nodes and the plurality of prediction deviation coefficients, including: Arrange the plurality of monitoring nodes in the order of the nodes to construct a monitoring node sequence; Select the first deviation coefficient of the first node and the second deviation coefficient of the second node, and calculate the first coefficient difference between the first deviation coefficient and the second deviation coefficient, where the first node is the initial node of the monitoring node sequence, and the second node is the adjacent node of the first node; If the first coefficient difference is less than a predetermined coefficient threshold, merge the first node and the second node to construct a first time period, where the deviation coefficient of the first time period is the mean of the first deviation coefficient and the second deviation coefficient; If the first coefficient difference is greater than or equal to the predetermined coefficient threshold, set a first node dividing line and perform iterative analysis until all nodes in the monitoring node sequence are analyzed to obtain a plurality of node dividing lines; Determine a plurality of monitoring time periods and a plurality of deviation coefficients corresponding to the plurality of monitoring time periods based on the plurality of node dividing lines, and construct the time period - deviation comparison table according to the plurality of monitoring time periods and the plurality of deviation coefficients.

2. The distributed photovoltaic energy efficiency dynamic monitoring method based on timing analysis according to claim 1, characterized in that The electrical operating parameters at least include DC input voltage, DC input current, power factor, and inverter temperature, where the inverter temperature is monitored by a temperature sensor deployed on the PV inverter.

3. The distributed photovoltaic energy efficiency dynamic monitoring method based on timing analysis according to claim 1, characterized in that, Perform power output compensation based on the time period - deviation comparison table, including: Receive the real - time electrical operating parameters of the current node, and use the power predictor to predict the PV output power of the real - time electrical operating parameters to obtain a real - time predicted output power; Input the current node into the time period - deviation comparison table for matching and output an adapted deviation coefficient; Compensate the real - time predicted output power according to the adapted deviation coefficient to obtain a real - time output power, and perform grid - connected scheduling of the distributed PV equipment.

Images