Virtual Power Plant Integrated Energy Management System Based on Photovoltaic-Storage Co-operation and Control

Abstract

This invention discloses a virtual power plant integrated energy management system based on photovoltaic-storage co-regulation control, belonging to the field of virtual power plant technology. It includes: a data acquisition module for collecting multi-source data from the virtual power plant; a data processing module for processing the multi-source data to determine characteristic data of the virtual power plant; an analysis and prediction module for constructing a virtual power plant integrated energy management prediction model to analyze the characteristic data and determine the prediction results; and an energy management module for managing the integrated energy of the virtual power plant based on photovoltaic-storage co-regulation control. This invention solves the problem of existing systems failing to effectively manage the integrated energy of virtual power plants, resulting in poor management performance. This invention enables effective management of the integrated energy of virtual power plants based on photovoltaic-storage co-regulation control, achieving efficient energy utilization and flexible grid regulation, thus improving the management effect of virtual power plants.

Description

Technical Field

[0001] This invention relates to the field of virtual power plant technology, specifically to a virtual power plant integrated energy management system based on photovoltaic-storage joint control. Background Technology

[0002] A virtual power plant is a smart energy system that aggregates distributed energy resources (such as photovoltaic power generation, energy storage devices, and adjustable loads) through advanced information and communication technologies and digital means to achieve centralized and optimized management. Its core functions include: resource aggregation: integrating distributed power sources, energy storage, and controllable loads to form a virtual "power plant"; intelligent dispatch: dynamically adjusting energy production and consumption using artificial intelligence, big data analysis, and real-time monitoring technologies; and market participation: obtaining economic benefits through participation in electricity market transactions.

[0003] Existing technologies cannot effectively manage the integrated energy of virtual power plants based on photovoltaic-storage joint control, and cannot effectively achieve efficient energy utilization and flexible grid regulation, resulting in poor management performance of virtual power plants. Summary of the Invention

[0004] The purpose of this invention is to provide a virtual power plant integrated energy management system based on photovoltaic-storage joint control, which can effectively manage the integrated energy of the virtual power plant based on photovoltaic-storage joint control, effectively realize the efficient utilization of energy and the flexible regulation of the power grid, improve the management effect of the virtual power plant, and solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] The integrated energy management system for virtual power plants based on photovoltaic-storage co-operation and control includes:

[0007] The data acquisition module is used to collect multi-source data from the virtual power plant based on IoT devices;

[0008] The data processing module is used to process multi-source data from the virtual power plant and determine the characteristic data of the virtual power plant.

[0009] The analysis and prediction module is used to construct a prediction model for the integrated energy management of virtual power plants, analyze the characteristic data of virtual power plants, and determine the prediction results for the integrated energy management of virtual power plants.

[0010] The energy management module is used to manage the integrated energy of the virtual power plant based on the photovoltaic-storage joint control.

[0011] Preferably, a virtual power plant integrated energy management prediction model is constructed to analyze the characteristic data of the virtual power plant, and the following operations are performed:

[0012] The virtual power plant integrated energy management prediction model is deployed in a virtual power plant integrated energy management prediction environment based on photovoltaic-storage joint control.

[0013] The characteristic data of the virtual power plant is input into the virtual power plant integrated energy management prediction model. The virtual power plant characteristic data is analyzed according to the virtual power plant integrated energy management prediction model to identify the integrated energy management status of the virtual power plant, and the supply and demand status of the virtual power plant is predicted to determine the prediction result of the virtual power plant integrated energy management.

[0014] Preferably, based on the prediction results of the integrated energy management of the virtual power plant, energy storage equipment is used to smooth the volatility of photovoltaic power generation and achieve peak shaving and valley filling and demand response, and the integrated energy of the virtual power plant is effectively managed based on the photovoltaic-storage joint regulation and control.

[0015] In cases where there is a surplus of electricity after the distributed photovoltaic power generation meets the electricity load demand, energy storage devices are used to store the surplus electricity.

[0016] When the distributed photovoltaic power generation cannot meet the electricity load demand, energy storage devices are used to release the stored electrical energy to meet the electricity load demand.

[0017] Preferably, collect multi-source data from the virtual power plant and perform the following operations:

[0018] Real-time monitoring and collection of power generation data, environmental data, and equipment status data from the power generation side are achieved using IoT devices to obtain distributed photovoltaic power generation data.

[0019] Real-time monitoring and collection of energy storage device charging and discharging status and battery pack data are performed using IoT devices to obtain energy storage device data.

[0020] Real-time monitoring and collection of real-time electricity consumption data and adjustability data on the electricity consumption side are carried out using IoT devices to obtain controllable load data;

[0021] Based on distributed photovoltaic power generation data, energy storage equipment data, and controllable load data, multi-source data for the virtual power plant is determined.

[0022] Preferably, the real-time power generation data includes photovoltaic array output power, DC side voltage and current, AC side voltage, current and frequency, inverter efficiency, and power factor;

[0023] Environmental data includes light intensity, ambient temperature, photovoltaic module temperature, wind speed, and wind direction;

[0024] Equipment status data includes inverter operating status, photovoltaic module operating status, and grid connection point voltage status;

[0025] The charge / discharge status includes the current charge / discharge power, charge / discharge current, charge / discharge efficiency, remaining battery capacity, battery health status, and depth of discharge.

[0026] Battery pack data includes individual cell voltage, total battery pack voltage, battery temperature, and internal resistance;

[0027] Real-time electricity consumption data includes load power, voltage, current, power factor, and load curve;

[0028] Adjustability data includes maximum power reduction, minimum operating power, interruptible time, and recovery time.

[0029] Preferably, the multi-source data of the virtual power plant is processed by performing the following operations:

[0030] Clean the multi-source data of the virtual power plant to remove noisy data that is not valuable for the integrated energy management of the virtual power plant, thereby reducing the interference of noise on the integrated energy management of the virtual power plant.

[0031] Real-time detection of multi-source data from virtual power plants, identification of outliers in the multi-source data of virtual power plants, and processing of outliers in the multi-source data of virtual power plants;

[0032] Among them, outliers in the multi-source data of virtual power plants are evaluated, and the contribution rate is used to determine whether outliers in the multi-source data of virtual power plants are valuable for the integrated energy management of virtual power plants.

[0033] When outliers in the multi-source data of the virtual power plant are valuable for the integrated energy management of the virtual power plant, the outliers in the multi-source data of the virtual power plant are corrected.

[0034] When outliers in the multi-source data of a virtual power plant are of no value to the integrated energy management of the virtual power plant, they are deleted.

[0035] Preferably, the multi-source data of the virtual power plant is processed by performing the following operations:

[0036] Normalize the multi-source data of virtual power plants to convert it into a unified data format, remove the differences in dimensions in the multi-source data of virtual power plants, and form standardized multi-source data of virtual power plants.

[0037] The system integrates multi-source data from virtual power plants, combining data from different sources into a unified data view, and stores and backs up the integrated multi-source data.

[0038] Feature extraction is performed on multi-source data of virtual power plants. Feature vectors related to the integrated energy management of virtual power plants are extracted from the multi-source data of virtual power plants, and the extracted feature vectors are weighted and fused to form virtual power plant feature data.

[0039] Preferably, a virtual power plant integrated energy management prediction model is constructed, and the following operations are performed:

[0040] Collect historical data of virtual power plants and divide the collected historical data of virtual power plants into training set and test set.

[0041] Based on deep learning technology, a training set is used to train the deep learning model, enabling the deep learning model to autonomously learn the prediction behavior of virtual power plant integrated energy management from the training set, and to predict the supply and demand of virtual power plant, thus determining the deep learning-based prediction model for virtual power plant integrated energy management.

[0042] The test set is input into the deep learning-based virtual power plant integrated energy management prediction model. The deep learning-based virtual power plant integrated energy management prediction model is tested based on the test set to evaluate its performance, determine whether it can achieve the expected effect of virtual power plant integrated energy management prediction, and determine the model test evaluation results.

[0043] When the deep learning-based virtual power plant integrated energy management prediction model fails to achieve the expected results of virtual power plant integrated energy management prediction, the parameters of the deep learning-based virtual power plant integrated energy management prediction model are adjusted and iteratively optimized until the deep learning-based virtual power plant integrated energy management prediction model can achieve the expected results of virtual power plant integrated energy management prediction, thereby determining the optimal virtual power plant integrated energy management prediction model.

[0044] Preferably, real-time detection is performed on multi-source data of the virtual power plant to identify outliers in the multi-source data, including:

[0045] The virtual power plant multi-source data is sorted and aligned based on time series to obtain the aligned virtual power plant multi-source data.

[0046] Take any one data source from the aligned multi-source data of the virtual power plant as the data to be identified;

[0047] Obtain the extreme value of each data point in the neighborhood range in the data to be identified;

[0048] The data to be identified is divided into regions based on continuous data points with the same extreme values ​​to determine several time series segments;

[0049] Calculate the correlation coefficient between any two time series segments;

[0050] The correlation coefficient is compared with a preset correlation coefficient threshold, and time series segments with correlation coefficients greater than or equal to the preset correlation coefficient threshold are grouped into time series segment groups to obtain several time series segment groups.

[0051] Calculate the entropy and mean of the data values ​​in each time series segment within each time series segment group; determine the evaluation coefficient for each time series segment based on the entropy and mean;

[0052] Calculate the mean of the evaluation coefficients of all time series segments in each time series segment group; determine the deviation of the evaluation coefficient of each time series segment from the mean based on the mean;

[0053] Time series segments with deviations from the mean greater than or equal to a preset deviation threshold are identified as anomalous time series segments; all time series segment groups are iterated through to obtain several anomalous time series segments;

[0054] Take any abnormal time series segment as the first abnormal time series segment; take several normal time series segments in the time series segment group containing the first abnormal time series segment as the second time series segment;

[0055] Calculate the mean value of the data points at the corresponding positions of each data point in several second time series segments;

[0056] The difference between the data points in the first abnormal time series segment and the mean of the data is calculated to obtain the deviation value of each data point in the first abnormal time series segment;

[0057] Data points with deviation values ​​greater than or equal to a preset deviation threshold are considered abnormal data points.

[0058] Traverse all the abnormal time series segments to obtain several abnormal data points;

[0059] Traverse all data sources in the virtual power plant's multi-source data to determine the outliers corresponding to the abnormal data points in the virtual power plant's multi-source data.

[0060] Preferably, the assessment of outliers in multi-source data of a virtual power plant based on contribution rate is valuable for the integrated energy management of the virtual power plant, including:

[0061] The outliers in the multi-source data are evaluated based on their contribution rates in the technical, economic, and risk dimensions to determine their evaluation value for the integrated energy management of the virtual power plant.

[0062]

[0063] in, This represents the evaluation value of outliers for the integrated energy management of the virtual power plant; , , These are the weighting coefficients; This represents the weight value of the s-th type of data source; S represents the total number of data source types. This represents the i-th outlier in the s-th data source; n represents the total number of outliers in the s-th data source. This represents the historical mean of data values ​​in the s-th data source; This represents the historical standard deviation of data values ​​in the s-th type of data source; This represents the sensitivity of the system's energy efficiency to the s-th type of data source, i.e.: ; This indicates the cost difference caused by abnormal data; Indicates the reference cost benchmark value; This represents the time decay factor; t represents the time of anomaly occurrence. Indicates the peak value of electricity prices; Indicates the time decay coefficient; Indicates the probability of system cascading failure; This represents the failure probability of the k-th subsystem; k represents the subsystem index. This indicates that an anomaly has occurred in the s-th type of data source. When, the failure probability of subsystem k; Indicates the risk tolerance threshold;

[0064]

[0065] The evaluation value Compare with the preset evaluation value threshold; if the evaluation value... If the outlier in the multi-source data is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant; if the evaluation value is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant. If the outliers in the multi-source data are less than the preset evaluation threshold, then they are of no value to the integrated energy management of the virtual power plant.

[0066] Compared with the prior art, the beneficial effects of the present invention are:

[0067] This invention collects multi-source data from a virtual power plant by real-time monitoring of distributed photovoltaic power generation data, energy storage device data, and controllable load data through IoT devices. By processing this multi-source data and extracting key features for weighted fusion, characteristic data of the virtual power plant is determined. A comprehensive energy management prediction model for the virtual power plant is constructed to analyze this characteristic data and predict the supply and demand situation, thus determining the comprehensive energy management prediction results. Based on these prediction results, energy storage devices are used to smooth the volatility of photovoltaic power generation, achieving peak shaving and valley filling, and demand response. Furthermore, based on photovoltaic-storage integrated control, the comprehensive energy of the virtual power plant is effectively managed. This approach enables efficient energy utilization and flexible grid regulation, improving the management effectiveness of the virtual power plant. Attached Figure Description

[0068] Figure 1 This is a block diagram of the virtual power plant integrated energy management system of the present invention;

[0069] Figure 2 This is a flowchart of the virtual power plant integrated energy management system of the present invention. Detailed Implementation

[0070] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] To address the current limitations in effectively managing the integrated energy resources of virtual power plants based on photovoltaic-storage integrated control, thus hindering efficient energy utilization and flexible grid regulation, resulting in poor virtual power plant management performance, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution:

[0072] The integrated energy management system for virtual power plants based on photovoltaic and energy storage joint control includes: a data acquisition module, a data processing module, an analysis and prediction module, and an energy management module.

[0073] Specifically, through the interaction between the data acquisition module, data processing module, analysis and prediction module, and energy management module, the integrated energy of the virtual power plant can be effectively managed based on the photovoltaic-storage joint control, which can effectively realize the efficient utilization of energy and the flexible regulation of the power grid, thereby improving the management effect of the virtual power plant.

[0074] The data acquisition module is used to collect multi-source data from the virtual power plant based on IoT devices.

[0075] In this embodiment, multi-source data from the virtual power plant is collected, and the following operations are performed:

[0076] Real-time monitoring and collection of power generation data, environmental data, and equipment status data from the power generation side are achieved using IoT devices to obtain distributed photovoltaic power generation data.

[0077] It should be noted that real-time power generation data includes photovoltaic array output power, DC side voltage and current, AC side voltage, current and frequency, inverter efficiency, and power factor.

[0078] Environmental data includes light intensity, ambient temperature, photovoltaic module temperature, wind speed, and wind direction;

[0079] Equipment status data includes inverter operating status, photovoltaic module operating status, and grid connection point voltage status;

[0080] Real-time monitoring and collection of energy storage device charging and discharging status and battery pack data are performed using IoT devices to obtain energy storage device data.

[0081] It should be noted that the charge / discharge status includes the current charge / discharge power, charge / discharge current, charge / discharge efficiency, remaining battery capacity, battery health status, and depth of discharge.

[0082] Battery pack data includes individual cell voltage, total battery pack voltage, battery temperature, and internal resistance;

[0083] Real-time monitoring and collection of real-time electricity consumption data and adjustability data on the electricity consumption side are carried out using IoT devices to obtain controllable load data;

[0084] It should be noted that real-time electricity consumption data includes load power, voltage, current, power factor, and load curve;

[0085] Adjustability data includes maximum power reduction, minimum operating power, interruptible time, and recovery time;

[0086] Based on distributed photovoltaic power generation data, energy storage equipment data, and controllable load data, multi-source data for the virtual power plant can be determined, which can provide data support for subsequent integrated energy management of the virtual power plant.

[0087] The data processing module is used to process multi-source data from the virtual power plant and determine the characteristic data of the virtual power plant.

[0088] In this embodiment, the multi-source data of the virtual power plant is processed by performing the following operations:

[0089] Clean the multi-source data of the virtual power plant to remove noisy data that is not valuable for the integrated energy management of the virtual power plant, reduce the interference of noise on the integrated energy management of the virtual power plant, and improve the accuracy of data analysis.

[0090] Real-time detection of multi-source data from virtual power plants, identification of outliers in the multi-source data of virtual power plants, and processing of outliers in the multi-source data of virtual power plants;

[0091] Among them, outliers in the multi-source data of virtual power plants are evaluated, and the contribution rate is used to determine whether outliers in the multi-source data of virtual power plants are valuable for the integrated energy management of virtual power plants.

[0092] When outliers in the multi-source data of the virtual power plant are valuable for the integrated energy management of the virtual power plant, the outliers in the multi-source data of the virtual power plant are corrected.

[0093] When outliers in the multi-source data of a virtual power plant are of no value to the integrated energy management of the virtual power plant, they are deleted.

[0094] It should be noted that by cleaning the multi-source data of the virtual power plant, removing noisy data and outliers that are not valuable for the integrated energy management of the virtual power plant, and correcting outliers that are valuable for the integrated energy management of the virtual power plant, the data quality of the multi-source data of the virtual power plant can be improved.

[0095] In this embodiment, the multi-source data of the virtual power plant is processed by performing the following operations:

[0096] Normalize the multi-source data of virtual power plants to convert it into a unified data format, remove the differences in dimensions in the multi-source data of virtual power plants, and form standardized multi-source data of virtual power plants.

[0097] The system integrates multi-source data from virtual power plants, combining data from different sources into a unified data view, and stores and backs up the integrated multi-source data.

[0098] Feature extraction is performed on multi-source data of virtual power plants. Feature vectors related to the integrated energy management of virtual power plants are extracted from the multi-source data of virtual power plants, and the extracted feature vectors are weighted and fused to form virtual power plant feature data.

[0099] It should be noted that by normalizing, integrating, and extracting features from multi-source data of virtual power plants, feature vectors related to the integrated energy management of virtual power plants can be extracted. The extracted feature vectors are then weighted and fused to form virtual power plant feature data, which facilitates better analysis of the integrated energy management of virtual power plants.

[0100] The analysis and prediction module is used to construct a virtual power plant integrated energy management prediction model, analyze the characteristic data of the virtual power plant, and determine the prediction results of the virtual power plant integrated energy management.

[0101] In this embodiment, a virtual power plant integrated energy management prediction model is constructed, and the following operations are performed:

[0102] Collect historical data of virtual power plants and divide the collected historical data of virtual power plants into training set and test set.

[0103] Based on deep learning technology, a training set is used to train the deep learning model, enabling the deep learning model to autonomously learn the prediction behavior of virtual power plant integrated energy management from the training set, and to predict the supply and demand of virtual power plant, thus determining the deep learning-based prediction model for virtual power plant integrated energy management.

[0104] The test set is input into the deep learning-based virtual power plant integrated energy management prediction model. The deep learning-based virtual power plant integrated energy management prediction model is tested based on the test set to evaluate its performance, determine whether it can achieve the expected effect of virtual power plant integrated energy management prediction, and determine the model test evaluation results.

[0105] When the deep learning-based virtual power plant integrated energy management prediction model fails to achieve the expected results of virtual power plant integrated energy management prediction, the parameters of the deep learning-based virtual power plant integrated energy management prediction model are adjusted and iteratively optimized until the deep learning-based virtual power plant integrated energy management prediction model can achieve the expected results of virtual power plant integrated energy management prediction, thereby determining the optimal virtual power plant integrated energy management prediction model.

[0106] In this embodiment, a virtual power plant integrated energy management prediction model is constructed to analyze the characteristic data of the virtual power plant and perform the following operations:

[0107] The optimal virtual power plant integrated energy management prediction model is deployed in a virtual power plant integrated energy management prediction environment based on photovoltaic-storage joint control.

[0108] The characteristic data of the virtual power plant is input into the virtual power plant integrated energy management prediction model. The virtual power plant characteristic data is analyzed according to the virtual power plant integrated energy management prediction model to identify the integrated energy management status of the virtual power plant, and the supply and demand status of the virtual power plant is predicted to determine the prediction result of the virtual power plant integrated energy management.

[0109] The energy management module is used to manage the integrated energy of the virtual power plant based on the photovoltaic-storage joint control.

[0110] In this embodiment, based on the prediction results of the integrated energy management of the virtual power plant, energy storage equipment is used to smooth the volatility of photovoltaic power generation and achieve peak shaving and valley filling and demand response. The integrated energy of the virtual power plant is effectively managed based on the photovoltaic-storage joint regulation and control.

[0111] In cases where there is a surplus of electricity after the distributed photovoltaic power generation meets the electricity load demand, energy storage devices are used to store the surplus electricity.

[0112] When the distributed photovoltaic power generation cannot meet the electricity load demand, energy storage devices are used to release the stored electrical energy to meet the electricity load demand.

[0113] In summary, by constructing a comprehensive energy management prediction model for virtual power plants, analyzing the characteristic data of virtual power plants, and predicting the supply and demand of virtual power plants, the prediction results for comprehensive energy management of virtual power plants are determined. Based on the prediction results, energy storage equipment is used to smooth the volatility of photovoltaic power generation and achieve peak shaving and valley filling, as well as demand response. Furthermore, the comprehensive energy of virtual power plants can be effectively managed based on photovoltaic-storage joint regulation control. This approach can effectively achieve efficient energy utilization and flexible grid regulation, thereby improving the management effect of virtual power plants.

[0114] In this embodiment, real-time detection of multi-source data from the virtual power plant is performed to identify outliers in the multi-source data, including:

[0115] The virtual power plant multi-source data is sorted and aligned based on time series to obtain the aligned virtual power plant multi-source data.

[0116] Take any one data source from the aligned multi-source data of the virtual power plant as the data to be identified;

[0117] Obtain the extreme value of each data point in the neighborhood range in the data to be identified;

[0118] The data to be identified is divided into regions based on continuous data points with the same extreme values ​​to determine several time series segments;

[0119] Calculate the correlation coefficient between any two time series segments;

[0120] The correlation coefficient is compared with a preset correlation coefficient threshold, and time series segments with correlation coefficients greater than or equal to the preset correlation coefficient threshold are grouped into time series segment groups to obtain several time series segment groups.

[0121] Calculate the entropy and mean of the data values ​​in each time series segment within each time series segment group; determine the evaluation coefficient for each time series segment based on the entropy and mean;

[0122] Calculate the mean of the evaluation coefficients of all time series segments in each time series segment group; determine the deviation of the evaluation coefficient of each time series segment from the mean based on the mean;

[0123] Time series segments with deviations from the mean greater than or equal to a preset deviation threshold are identified as anomalous time series segments; all time series segment groups are iterated through to obtain several anomalous time series segments;

[0124] Take any abnormal time series segment as the first abnormal time series segment; take several normal time series segments in the time series segment group containing the first abnormal time series segment as the second time series segment;

[0125] Calculate the mean value of the data points at the corresponding positions of each data point in several second time series segments;

[0126] The difference between the data points in the first abnormal time series segment and the mean of the data is calculated to obtain the deviation value of each data point in the first abnormal time series segment;

[0127] Data points with deviation values ​​greater than or equal to a preset deviation threshold are considered abnormal data points.

[0128] Traverse all the abnormal time series segments to obtain several abnormal data points;

[0129] Traverse all data sources in the virtual power plant's multi-source data to determine the outliers corresponding to the abnormal data points in the virtual power plant's multi-source data.

[0130] In this embodiment, the data to be identified is divided into regions based on continuous data points with the same extreme values ​​to determine several time series segments; specifically, extreme values ​​include maximum and minimum values; if the maximum or minimum values ​​of two adjacent data points are the same, then the two data points are divided into a time series segment.

[0131] In this embodiment, the evaluation coefficient of each time series segment is determined based on the entropy and the mean. Specifically, the product of the entropy and the mean is used as the evaluation coefficient of each time series segment. That is, the larger the entropy and the larger the mean, the more volatile and disordered the data values ​​of the data points in the time series segment are.

[0132] In this embodiment, an abnormal time series segment is randomly selected as the first abnormal time series segment; several normal time series segments in the time series segment group containing the first abnormal time series segment are selected as the second time series segment.

[0133] Calculate the mean value of the data points at the corresponding positions of each data point in several second time series segments;

[0134] The difference between the data points in the first abnormal time series segment and the mean of the data is calculated to obtain the deviation value of each data point in the first abnormal time series segment;

[0135] Data points with deviation values ​​greater than or equal to a preset deviation threshold are considered abnormal data points.

[0136] For example: First abnormal segment (t1-t5): [22, 23, 35, 24, 21]; within the segment group of the same device, select 3 known normal segments as the "second time series segment":

[0137] Normal segment 1: [21, 22, 23, 24, 22]

[0138] Normal segment 2: [23, 24, 22, 23, 24]

[0139] Normal segment 3: [22, 23, 24, 22, 23]

[0140] Calculate the mean of the data at the same time point (e.g., t1 corresponds to the first data point in all segments) in three normal segments to obtain the "mean sequence":

[0141] Mean of t1: (21 + 23 + 22) ÷ 3 = 22

[0142] Mean of t2: (22 + 24 + 23) ÷ 3 = 23

[0143] The mean of t3 is (23 + 22 + 24) ÷ 3 = 23

[0144] The mean of t4 is (24 + 23 + 22) ÷ 3 = 23

[0145] The mean of t5 is (22 + 24 + 23) ÷ 3 = 23

[0146] Mean sequence: [22, 23, 23, 23, 23]

[0147] The "deviation value" is obtained by subtracting the mean of the corresponding position in the mean sequence from each data point in the first outlier segment:

[0148] t1 deviation: 22 - 22 = 0

[0149] t2 deviation: 23 - 23 = 0

[0150] t3 deviation: 35 - 23 = 12

[0151] t4 deviation: 24 - 23 = 1

[0152] t5 deviation: 21 - 23 = -2

[0153] Deviation value sequence: [0, 0, 12, 1, -2]

[0154] The preset deviation threshold is "5" (meaning that an absolute deviation value ≥ 5 is considered abnormal):

[0155] t1 deviation 0 < 5 → normal

[0156] t2 deviation 0 < 5 → Normal

[0157] t3 deviation ≥ 5 → Abnormal

[0158] t4 Deviation 1 < 5 → Normal

[0159] t5 deviation - 2 (absolute value 2) < 5 → Normal

[0160] Conclusion: The t3 (3rd hour) data point in the first abnormal time series segment is an abnormal data point.

[0161] The working principle and beneficial effects of the above technical solution are as follows: Through detailed time series analysis, including sorting, alignment, and region division, the time series segments where abnormal data points are located can be accurately determined; since it is designed for processing multi-source data of virtual power plants, it comprehensively considers data from different data sources; by timely detection of these outliers, potentially problematic equipment or systems can be inspected and maintained in advance, preventing further development of faults, thereby ensuring the stable operation of virtual power plants and reducing the occurrence of power outages or unstable power supply due to equipment failures.

[0162] In this embodiment, determining whether outliers in the multi-source data of the virtual power plant are valuable for the integrated energy management of the virtual power plant based on the contribution rate includes:

[0163] The outliers in the multi-source data are evaluated based on their contribution rates in the technical, economic, and risk dimensions to determine their evaluation value for the integrated energy management of the virtual power plant.

[0164]

[0165] in, This represents the evaluation value of outliers for the integrated energy management of the virtual power plant; , , These are the weighting coefficients; This represents the weight value of the s-th type of data source; S represents the total number of data source types. This represents the i-th outlier in the s-th data source; n represents the total number of outliers in the s-th data source. This represents the historical mean of data values ​​in the s-th data source; This represents the historical standard deviation of data values ​​in the s-th type of data source; This represents the sensitivity of the system's energy efficiency to the s-th type of data source, i.e.: ; This indicates the cost difference caused by abnormal data; Indicates the reference cost benchmark value; This represents the time decay factor; t represents the time of anomaly occurrence. Indicates the peak value of electricity prices; Indicates the time decay coefficient; Indicates the probability of system cascading failure; This represents the failure probability of the k-th subsystem; k represents the subsystem index. This indicates that an anomaly has occurred in the s-th type of data source. When, the failure probability of subsystem k; Indicates the risk tolerance threshold;

[0166]

[0167] The evaluation value Compare with the preset evaluation value threshold; if the evaluation value... If the outlier in the multi-source data is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant; if the evaluation value is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant. If the outliers in the multi-source data are less than the preset evaluation threshold, then they are of no value to the integrated energy management of the virtual power plant.

[0168] In this embodiment, Indicates the technical dimension; Indicates the economic dimension; This indicates the risk dimension.

[0169] The working principle and beneficial effects of the above technical solution are as follows: by comprehensively considering the contribution rates of different dimensions to evaluate outliers, a more comprehensive decision-making basis can be provided for the integrated energy management of virtual power plants; accurately identifying the value of outliers helps to rationally allocate resources to deal with abnormal situations; managers can decide whether to intervene in abnormal situations and how to adjust energy management strategies based on the evaluation value of outliers, thereby improving the overall operating efficiency and benefits of virtual power plants.

[0170] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0171] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A virtual power plant integrated energy management system based on photovoltaic-storage co-regulation control, characterized in that, include: The data acquisition module is used to collect multi-source data from the virtual power plant based on IoT devices; The data processing module is used to process multi-source data from the virtual power plant and determine the characteristic data of the virtual power plant. Among them, real-time detection of multi-source data of virtual power plants is carried out to identify outliers in the multi-source data of virtual power plants, and the contribution rate is used to determine whether the outliers in the multi-source data of virtual power plants are valuable for the comprehensive energy management of virtual power plants. The analysis and prediction module is used to construct a virtual power plant integrated energy management prediction model, analyze the characteristic data of the virtual power plant, and determine the prediction results of the virtual power plant integrated energy management. The energy management module is used to manage the integrated energy of the virtual power plant based on the photovoltaic-storage joint control. Determining the value of outliers in multi-source data of a virtual power plant for integrated energy management based on contribution rate includes: The outliers in the multi-source data are evaluated based on their contribution rates in the technical, economic, and risk dimensions to determine their evaluation value for the integrated energy management of the virtual power plant. in, This represents the evaluation value of outliers for the integrated energy management of the virtual power plant; , , These are the weighting coefficients; This represents the weight value of the s-th type of data source; S represents the total number of data source types. This represents the i-th outlier in the s-th data source; n represents the total number of outliers in the s-th data source. This represents the historical mean of data values ​​in the s-th data source; This represents the historical standard deviation of data values ​​in the s-th type of data source; This represents the sensitivity of the system's energy efficiency to the s-th type of data source, i.e.: ; This indicates the cost difference caused by abnormal data; Indicates the reference cost benchmark value; This represents the time decay factor; t represents the time of anomaly occurrence. Indicates the peak value of electricity prices; Indicates the time decay coefficient; Indicates the probability of system cascading failure; This represents the failure probability of the k-th subsystem; k represents the subsystem index. This indicates that an anomaly has occurred in the s-th type of data source. When, the failure probability of subsystem k; Indicates the risk tolerance threshold; The evaluation value Compare with the preset evaluation value threshold; if the evaluation value... If the outlier in the multi-source data is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant; if the evaluation value is greater than or equal to the preset evaluation threshold, then the outlier is valuable for the integrated energy management of the virtual power plant. If the outlier is less than the preset evaluation threshold, then the outlier in the multi-source data is of no value for the integrated energy management of the virtual power plant. Construct a virtual power plant integrated energy management prediction model and perform the following operations: Collect historical data of virtual power plants and divide the collected historical data of virtual power plants into training set and test set. Based on deep learning technology, a training set is used to train the deep learning model, enabling the deep learning model to autonomously learn the prediction behavior of virtual power plant integrated energy management from the training set, and to predict the supply and demand of virtual power plant, thus determining the deep learning-based prediction model for virtual power plant integrated energy management. The test set is input into the deep learning-based virtual power plant integrated energy management prediction model. The deep learning-based virtual power plant integrated energy management prediction model is tested based on the test set to evaluate its performance, determine whether it can achieve the expected effect of virtual power plant integrated energy management prediction, and determine the model test evaluation results. When the deep learning-based virtual power plant integrated energy management prediction model fails to achieve the expected results of virtual power plant integrated energy management prediction, the parameters of the deep learning-based virtual power plant integrated energy management prediction model are adjusted and iteratively optimized until the deep learning-based virtual power plant integrated energy management prediction model can achieve the expected results of virtual power plant integrated energy management prediction, thereby determining the optimal virtual power plant integrated energy management prediction model.

2. The virtual power plant integrated energy management system based on photovoltaic-storage joint control as described in claim 1, characterized in that, Real-time monitoring of multi-source data from virtual power plants to identify outliers, including: The virtual power plant multi-source data is sorted and aligned based on time series to obtain the aligned virtual power plant multi-source data. Take any one data source from the aligned multi-source data of the virtual power plant as the data to be identified; Obtain the extreme value of each data point in the neighborhood range in the data to be identified; The data to be identified is divided into regions based on continuous data points with the same extreme values ​​to determine several time series segments; Calculate the correlation coefficient between any two time series segments; The correlation coefficient is compared with a preset correlation coefficient threshold, and time series segments with correlation coefficients greater than or equal to the preset correlation coefficient threshold are grouped into time series segment groups to obtain several time series segment groups. Calculate the entropy and mean of the data values ​​in each time series segment within each time series segment group; determine the evaluation coefficient for each time series segment based on the entropy and mean; Calculate the mean of the evaluation coefficients of all time series segments in each time series segment group; determine the deviation of the evaluation coefficient of each time series segment from the mean based on the mean; Time series segments with deviations from the mean greater than or equal to a preset deviation threshold are identified as anomalous time series segments; all time series segment groups are iterated through to obtain several anomalous time series segments; Take any abnormal time series segment as the first abnormal time series segment; take several normal time series segments in the time series segment group containing the first abnormal time series segment as the second time series segment; Calculate the mean value of the data points at the corresponding positions of each data point in several second time series segments; The difference between the data points in the first abnormal time series segment and the mean of the data is calculated to obtain the deviation value of each data point in the first abnormal time series segment; Data points with deviation values ​​greater than or equal to a preset deviation threshold are considered abnormal data points. Traverse all the abnormal time series segments to obtain several abnormal data points; Traverse all data sources in the virtual power plant's multi-source data to determine the outliers corresponding to the abnormal data points in the virtual power plant's multi-source data.

3. The virtual power plant integrated energy management system based on photovoltaic-storage joint control as described in claim 2, characterized in that, To construct a comprehensive energy management prediction model for virtual power plants, analyze the characteristic data of virtual power plants and perform the following operations: The virtual power plant integrated energy management prediction model is deployed in a virtual power plant integrated energy management prediction environment based on photovoltaic-storage joint control. The characteristic data of the virtual power plant is input into the virtual power plant integrated energy management prediction model. The virtual power plant characteristic data is analyzed according to the virtual power plant integrated energy management prediction model to identify the integrated energy management status of the virtual power plant, and the supply and demand status of the virtual power plant is predicted to determine the prediction result of the virtual power plant integrated energy management.

4. The virtual power plant integrated energy management system based on photovoltaic-storage joint control as described in claim 3, characterized in that, Based on the prediction results of the integrated energy management of the virtual power plant, energy storage equipment is used to smooth the volatility of photovoltaic power generation and achieve peak shaving and valley filling and demand response. The integrated energy of the virtual power plant is effectively managed based on the photovoltaic-storage joint regulation and control. In cases where there is a surplus of electricity after the distributed photovoltaic power generation meets the electricity load demand, energy storage devices are used to store the surplus electricity. When the distributed photovoltaic power generation cannot meet the electricity load demand, energy storage devices are used to release the stored electrical energy to meet the electricity load demand.

5. The virtual power plant integrated energy management system based on photovoltaic-storage joint control according to claim 4, characterized in that, Collect multi-source data from the virtual power plant and perform the following operations: Real-time monitoring and collection of power generation data, environmental data, and equipment status data from the power generation side are achieved using IoT devices to obtain distributed photovoltaic power generation data. Real-time monitoring and collection of energy storage device charging and discharging status and battery pack data are performed using IoT devices to obtain energy storage device data. Real-time monitoring and collection of real-time electricity consumption data and adjustability data on the electricity consumption side are carried out using IoT devices to obtain controllable load data; Based on distributed photovoltaic power generation data, energy storage equipment data, and controllable load data, multi-source data for the virtual power plant is determined.

6. The virtual power plant integrated energy management system based on photovoltaic-storage joint control as described in claim 5, characterized in that, Real-time power generation data includes photovoltaic array output power, DC side voltage and current, AC side voltage, current and frequency, inverter efficiency, and power factor; Environmental data includes light intensity, ambient temperature, photovoltaic module temperature, wind speed, and wind direction; Equipment status data includes inverter operating status, photovoltaic module operating status, and grid connection point voltage status; The charge / discharge status includes the current charge / discharge power, charge / discharge current, charge / discharge efficiency, remaining battery capacity, battery health status, and depth of discharge. Battery pack data includes individual cell voltage, total battery pack voltage, battery temperature, and internal resistance; Real-time electricity consumption data includes load power, voltage, current, power factor, and load curve; Adjustability data includes maximum power reduction, minimum operating power, interruptible time, and recovery time.

7. The virtual power plant integrated energy management system based on photovoltaic-storage joint control according to claim 6, characterized in that, Process the multi-source data of the virtual power plant by performing the following operations: Clean the multi-source data of the virtual power plant to remove noisy data that is not valuable for the integrated energy management of the virtual power plant, thereby reducing the interference of noise on the integrated energy management of the virtual power plant. Real-time detection of multi-source data from virtual power plants, identification of outliers in the multi-source data of virtual power plants, and processing of outliers in the multi-source data of virtual power plants; Among them, outliers in the multi-source data of virtual power plants are evaluated, and the contribution rate is used to determine whether outliers in the multi-source data of virtual power plants are valuable for the integrated energy management of virtual power plants. When outliers in the multi-source data of the virtual power plant are valuable for the integrated energy management of the virtual power plant, the outliers in the multi-source data of the virtual power plant are corrected. When outliers in the multi-source data of a virtual power plant are of no value to the integrated energy management of the virtual power plant, they are deleted.

8. The virtual power plant integrated energy management system based on photovoltaic-storage joint control according to claim 7, characterized in that, Process the multi-source data of the virtual power plant by performing the following operations: Normalize the multi-source data of virtual power plants to convert it into a unified data format, remove the differences in dimensions in the multi-source data of virtual power plants, and form standardized multi-source data of virtual power plants. The system integrates multi-source data from virtual power plants, combining data from different sources into a unified data view, and stores and backs up the integrated multi-source data. Feature extraction is performed on multi-source data of virtual power plants. Feature vectors related to the integrated energy management of virtual power plants are extracted from the multi-source data of virtual power plants, and the extracted feature vectors are weighted and fused to form virtual power plant feature data.

Images