Active control method and system for smoothing grid-connected power fluctuations for multi-source renewable energy

Abstract

The present invention discloses a method and system for actively smoothing grid-connected power fluctuations for multi-source renewable energy, and relates to the technical field of grid-connected power control. The method comprises: after accessing a multi-source renewable energy management device, reading device data; performing data collection of environmental data, establishing a time-series environmental data set, performing environmental prediction, and establishing environmental prediction results; performing collaborative relationship identification based on active identification of complementary windows, and establishing collaborative relationship identification results; performing initialization of a dynamic response virtual collaborative body; obtaining time-series load demand tasks, and generating collaborative scheduling results; and utilizing the collaborative scheduling results to actively smooth out grid-connected power fluctuations. The present invention solves the technical problems of the prior art in that multi-source renewable energy has large power fluctuations and low grid-connected system stability and reliability when multi-source renewable energy is connected to the grid, and achieves the technical effect of actively smoothing out grid-connected power fluctuations for multi-source renewable energy, and improving the stability and reliability of the grid-connected system.

Description

Technical Field

[0001] The present invention relates to the technical field of grid-connected power control, and in particular to a method and system for actively smoothing grid-connected power fluctuations for multi-source renewable energy. Background Art

[0002] In the field of renewable energy grid integration, the output power of renewable energy sources such as wind and photovoltaics is subject to fluctuations and intermittentity due to environmental factors. This leads to large power fluctuations during grid integration, seriously affecting the stability and reliability of the power grid. Existing technologies for addressing power fluctuations in multi-source renewable energy grid integration suffer from problems such as a lack of dynamic adaptability in coordinated scheduling strategies and insufficient response to complex environmental changes, making it difficult to effectively and proactively control grid-connected power fluctuations.

[0003] The existing technology has technical problems such as large power fluctuations when multiple sources of renewable energy are connected to the grid, and low stability and reliability of the grid-connected system. Summary of the Invention

[0004] The present application provides a method and system for actively smoothing power fluctuations in grid-connected multi-source renewable energy, which is used to solve the technical problems in the prior art of large power fluctuations and low stability and reliability of the grid-connected system when multi-source renewable energy is grid-connected.

[0005] In view of the above problems, the present application provides a method and system for actively smoothing grid-connected power fluctuations for multi-source renewable energy.

[0006] In a first aspect, the present application provides a method for actively smoothing grid-connected power fluctuations for multi-source renewable energy, the method comprising:

[0007] After accessing the multi-source new energy management equipment, the device data of the wind power equipment and the photovoltaic equipment are read; data collection of the environmental data is performed to establish a time series environmental data set, environmental prediction is performed based on the time series environmental data set, and an environmental prediction result is established; collaborative relationship identification based on the active identification complementary window is performed based on the device data and the environmental prediction result, and a collaborative relationship identification result is established; the collaborative relationship identification result is input into the dynamic response virtual collaborative body, and the dynamic response virtual collaborative body is initialized; the time series load demand task is obtained, and the time series load demand task and the environmental prediction result are used as input data and input into the initialized dynamic response virtual collaborative body to generate a collaborative scheduling result; the collaborative scheduling result is used to perform active smoothing control management of grid-connected power fluctuations.

[0008] A second aspect of the present application provides a grid-connected power fluctuation active smoothing control system for multi-source renewable energy, the system comprising:

[0009] The device data reading module is used to read the device data of wind power equipment and photovoltaic equipment after connecting to the multi-source new energy management equipment; the environmental prediction result establishment module is used to perform data acquisition of environmental data, establish a time-series environmental data set, perform environmental prediction based on the time-series environmental data set, and establish an environmental prediction result; the collaborative relationship identification result establishment module is used to perform collaborative relationship identification based on the active identification complementary window according to the device data and the environmental prediction result, and establish a collaborative relationship identification result; the initialization module is used to input the collaborative relationship identification result into the dynamic response virtual collaborative body and perform the initialization of the dynamic response virtual collaborative body; the collaborative scheduling result generation module is used to obtain the time-series load demand task, use the time-series load demand task and the environmental prediction result as input data, input them into the initialized dynamic response virtual collaborative body, and generate a collaborative scheduling result; the active smoothing control management module is used to use the collaborative scheduling result to perform active smoothing control management of grid-connected power fluctuations.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] After accessing the multi-source new energy management device, the system reads the device data of wind power equipment and photovoltaic equipment; performs data collection of environmental data, establishes a time-series environmental data set, performs environmental prediction, and establishes environmental prediction results; performs collaborative relationship identification based on active identification complementary windows based on the device data and the environmental prediction results, and establishes collaborative relationship identification results; performs initialization of a dynamic response virtual collaborative body; obtains time-series load demand tasks, uses the time-series load demand tasks and the environmental prediction results as input data, and inputs them into the initialized dynamic response virtual collaborative body to generate collaborative scheduling results; and uses the collaborative scheduling results to actively control and manage grid-connected power fluctuations. This achieves the technical effect of actively controlling and controlling the grid-connected power fluctuations of multi-source new energy and improving the stability and reliability of the grid-connected system. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A flow chart of a method for actively smoothing grid-connected power fluctuations for multi-source renewable energy provided in an embodiment of the present application;

[0014] Figure 2This is a schematic diagram of the structure of an active grid-connected power fluctuation smoothing control system for multi-source renewable energy provided in an embodiment of the present application.

[0015] Explanation of the accompanying reference numerals: device data reading module 10 , environment prediction result establishing module 20 , collaborative relationship identification result establishing module 30 , initialization module 40 , collaborative scheduling result generating module 50 , active leveling control management module 60 . DETAILED DESCRIPTION

[0016] This application provides a method and system for actively smoothing grid-connected power fluctuations for multi-source renewable energy, aiming to solve the technical problems in the prior art of large power fluctuations and low grid-connected system stability and reliability when multi-source renewable energy is connected to the grid.

[0017] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work are within the scope of protection of this application.

[0018] Example 1, as Figure 1 As shown, the present application provides a method for actively smoothing grid-connected power fluctuations for multi-source renewable energy, the method comprising:

[0019] Step S100: After accessing the multi-source new energy management device, the device data of the wind power equipment and the photovoltaic equipment are read.

[0020] Specifically, after connecting to multi-source new energy management equipment, the device operating status data is read in real time from wind energy equipment (such as wind speed sensors and power converters of wind turbines) and photovoltaic equipment (such as current and voltage monitoring modules and temperature sensors of solar panels) through standardized data interfaces, including but not limited to real-time values ​​of wind speed / direction, wind turbine output power curves, photovoltaic panel temperatures, photovoltaic array voltage / current parameters and other equipment data, providing basic data support for subsequent collaborative relationship identification based on active identification of complementary windows and initialization of dynamic response virtual collaborative bodies.

[0021] Step S200: performing data collection of environmental data, establishing a time series environmental data set, performing environmental prediction based on the time series environmental data set, and establishing an environmental prediction result.

[0022] Specifically, by deploying meteorological sensors (such as light intensity meters, anemometers, and temperature and humidity sensors) to collect environmental data, a time-series environmental dataset is constructed, encompassing parameters such as light intensity changes, wind speed / direction fluctuations, and temperature and humidity curves. Based on the collaborative contribution calculation logic of a multi-scale time window, an LSTM neural network is used to analyze the time-series environmental dataset, predicting future trends in environmental parameters (such as the light intensity fluctuation range and wind speed probability distribution over the next hour). This environmental prediction result is then generated and used as the input variable of the universal collaborative contribution function (influencing the power forecast volatility index σ and the dispatchability index R).

[0023] Step S300: performing collaborative relationship identification based on active identification of complementary windows according to the device data and the environmental prediction result, and establishing a collaborative relationship identification result.

[0024] Specifically, based on device data (such as the output power and operating status parameters of wind and photovoltaic equipment) and environmental forecast results (such as future trends in light intensity and wind speed), multi-scale time windows (such as minute and hour levels) are created by actively identifying complementary windows and configuring window trust factors. A universal collaborative contribution function is used to calculate the collaborative contribution values ​​of each energy source in different time windows. This function comprehensively considers power forecast volatility indicators, dispatchability indicators, and historical reliability indicators and assigns corresponding weight coefficients. Next, a prediction trust indicator is calculated based on the environmental forecast results to correct the window trust factor. The collaborative contribution values ​​are weighted to generate a first collaborative relationship identification result. At the same time, cloud-shared scenario similarity matching is performed based on the device data to establish a second collaborative relationship identification result. After collaborative authentication, the collaborative relationship identification is completed, forming a collaborative relationship mapping for each new energy device in different time windows.

[0025] Step S400: inputting the cooperative relationship identification result into the dynamic response virtual cooperative body, and executing the initialization of the dynamic response virtual cooperative body.

[0026] Specifically, the established collaborative relationship identification results (including the collaborative contribution matrix and collaborative relationship mapping of each new energy device under multi-scale time windows) are input into the dynamic response virtual collaborative body. The three-layer architecture of the virtual collaborative body is parameterized through the initialization module, and the virtual layer is called to virtually decompose the time-series load demand tasks based on the dispatchable capability index (R) of the equipment in the collaborative relationship to generate N virtual control task units; through the supply response layer, the volatility index (σ) in the collaborative contribution function and the environmental prediction results are combined to construct a supply source response feature pool containing the output characteristics of wind power / photovoltaic equipment; the matching layer is used to initialize the task-supply matching algorithm parameters according to the historical reliability index (η) in the collaborative relationship identification results to complete the initialization of the dynamic response virtual collaborative body, providing a virtual computing framework with energy collaborative optimization capabilities for the generation of collaborative scheduling results.

[0027] Step S500: obtaining a time-series load demand task, and inputting the time-series load demand task and the environment prediction result as input data into the initialized dynamic response virtual cooperative body to generate a cooperative scheduling result.

[0028] Specifically, the time-series load demand tasks (such as the electricity power demand curves in different time periods) arranged in chronological order are obtained, and are used together with the established environmental prediction results (such as the prediction of future changes in light intensity and wind speed) as input data, and input into the dynamic response virtual coordination body that has been initialized. First, the virtual layer in the virtual coordination body is called to decompose the time-series load demand tasks into N virtual control task units, each unit corresponding to the power control target in a specific time period; at the same time, the supply response layer is called to analyze the response capabilities of wind energy, photovoltaic and other supply sources in each time period based on the environmental prediction results, and construct a supply source response feature pool including parameters such as output power fluctuation range and dispatchable potential; finally, the matching layer initialized by the collaborative relationship identification result is used to dynamically adapt and match the virtual control task units and the supply source response feature pool, and combine the volatility index, dispatchable index and historical reliability index in the collaborative contribution function to generate the collaborative dispatch results of each new energy equipment in different time periods, and clarify the output power allocation plan and dispatch timing of wind energy and photovoltaic equipment.

[0029] Step S600: utilizing the coordinated scheduling result to perform active smoothing control management of grid-connected power fluctuations.

[0030] Specifically, the generated collaborative dispatch results are used to implement active smoothing control management for grid-connected power fluctuations. Based on the output power allocation plan of each new energy device at different time periods in the collaborative dispatch results, the operating status of wind power and photovoltaic equipment is adjusted in real time, and grid-connected power fluctuations are smoothed by adjusting the generated power. At the same time, the smoothing control effect is recorded to generate a record data set containing parameters such as power fluctuation amplitude and smoothing efficiency. Environmental features are extracted from the time-series environmental data set, and a sudden weather response feedback mechanism is established based on this to update the window trust factor for actively identifying complementary windows to optimize the identification and management of collaborative relationships. In addition, supply monitoring is performed on time-series load demand tasks, generating a supply monitoring data set to identify failure states in the collaborative dispatch results. If failure is found, an abnormal alarm is issued, and the dispatch strategy is switched based on the energy storage unit status and task adjustment analysis results. At the same time, visual management is implemented based on the collaborative contribution graph and dispatch return trajectory to ensure that the grid-connected power is stable within the target range.

[0031] In one possible implementation, step S300 further includes:

[0032] Step S310: creating a multi-scale time window according to the active identification complementary window, and configuring a window trust factor of the multi-scale time window.

[0033] Step S320: Utilizing the multi-scale time windows respectively to perform collaborative contribution calculations based on the environmental prediction results, and generating collaborative contribution values ​​mapped by the multi-scale time windows.

[0034] Step S330: Complete collaborative relationship identification based on the window trust factor, the collaborative contribution value, and the device data.

[0035] Specifically, based on the active identification of complementary windows, multi-scale time windows covering different time granularities, such as minutes, hours, and days, are created to adapt to the power output characteristics of wind and photovoltaic equipment under short-term fluctuations (such as sudden changes in wind speed at the minute level) and long-term trends (such as daily light cycles). Furthermore, based on parameters such as the accuracy of historical environmental forecast data (such as the root mean square error of hourly wind speed forecasts over the past seven days) and the frequency of device data collection (such as PV panel temperature sampling every 10 minutes), an initial window trust factor (ranging from 0.1 to 1) is assigned to each time window. For example, a trust factor of 0.8 is assigned to the minute-level window due to its rapid data updates but high noise, while a trust factor of 0.9 is assigned to the hour-level window due to its high prediction stability. This forms a weighting system that reflects the reliability of the time window data.

[0036] For the created multi-scale time windows (such as minute level, hour level, etc.), based on the environmental prediction results (such as the future trend of light intensity and wind speed), a general collaborative contribution function is used. , calculate the collaborative contribution. Among them, the first Energy in the time window The Power Forecast Volatility Index (such as the standard deviation of power fluctuation of photovoltaic equipment affected by cloud changes within the hourly window) and dispatchable indicators Such as the power adjustment ratio of wind turbines within the wind speed forecast range), and obtain historical reliability indicators of participation in collaboration based on historical equipment operation data (such as the successful execution rate of equipment collaborative scheduling in the past 30 days), through the weight coefficient of the volatility impact factor , the weight coefficient β of the regulation capability factor, and the weight coefficient γ of the reliability factor are weighted to generate the synergistic contribution value corresponding to each time window, and realize the quantitative mapping of the synergistic value of different energy sources in each time period.

[0037] First, the prediction trust identifier of each multi-scale time window is calculated based on the environmental prediction results, and the initially configured window trust factor is used to correct it so that it can reflect the actual reliability of the data in each time window under the current environmental conditions; then, the synergy contribution value is weightedly calculated using the corrected window trust factor to generate the first synergy relationship identification result, which reflects the weight difference of energy synergy contribution under different time windows; at the same time, based on the equipment data (such as the output power curve of wind energy and photovoltaic equipment, operating status parameters, etc.), cloud-shared scene similarity matching is performed to find the synergy relationship pattern under similar environmental and equipment status combinations in history, and establish the second synergy relationship identification result; finally, the first synergy relationship identification result and the second synergy relationship identification result are synergistically authenticated, integrating the time window weight and historical scenario experience, so as to complete the accurate identification of the synergy relationship between each new energy device, and provide a key synergy relationship matrix for the initialization of the subsequent dynamic response virtual synergy body.

[0038] In one possible implementation, step S320 further includes:

[0039] Step S321: Establish a general collaborative contribution function for the time window as follows:

[0040] ;in, Characterization Energy in the time window The collaborative contribution value of is the weight coefficient of the volatility impact factor, Characterization Energy in the time window The power of the volatility indicator, Characterization Energy in the time window The dispatchable index, is the weight coefficient of the control capability factor, For the The reliability index of each energy source's historical participation in coordination, is the weight coefficient of the reliability factor; the collaborative contribution calculation is completed according to the general collaborative contribution function.

[0041] Specifically, the process of calculating collaborative contribution based on multi-scale time windows is described as follows: First, a general collaborative contribution function for time windows is established: , in this function, represents the collaborative contribution, Characterization Energy in the time window The synergistic contribution value is used to quantify the contribution of this energy when working in conjunction with other energy sources within a specific time window. Represents the time window. Different values ​​correspond to multi-scale time windows such as minute, hour, and day levels, which are used to analyze the synergistic contribution of energy from different time dimensions. Indicates the serial number of energy, used to identify the Energy equipment that participates in the collaboration, such as wind power equipment or photovoltaic equipment. It is the weight coefficient of the volatility influencing factor, which is used to measure the weight of the power forecast volatility index in the calculation of the synergy contribution. Its value range is usually determined according to the actual application scenario and requirements. represents the power forecast volatility index, Characterization Energy in the time window The power forecast volatility reflects the instability of the power output of the energy source within a specific time window, such as the standard deviation of the power fluctuation of photovoltaic equipment affected by cloud changes within an hourly window. It is the weight coefficient of the control capability factor, which is used to reflect the importance of the dispatchable index in the calculation of the collaborative contribution. Its value is set according to the demand for energy control capability. represents the schedulable index, Characterization Energy in the time window The dispatchability of energy, such as the adjustable power ratio of wind turbines within the wind speed forecast range, reflects the ability of the energy to be dispatched within a specific time window. It is the weight coefficient of the reliability factor, which is used to determine the weight ratio of the reliability index of historical participation in collaboration in the calculation of collaborative contribution, and is set according to the system's requirements for energy reliability. is the reliability index of historical participation in collaboration, Indicates the The reliability of an energy source's historical participation in coordination, such as the successful execution rate of equipment coordinated scheduling in the past 30 days, reflects the reliability of the energy source in the historical coordination process.

[0042] When completing the synergy contribution calculation based on the general synergy contribution function, it is necessary to first determine the first Energy in the time window The Power Forecast Volatility Index (such as the standard deviation of power fluctuation of photovoltaic equipment affected by clouds within the hourly window) and dispatchable indicators (such as the power adjustment ratio of wind turbines within the wind speed forecast range), and then obtain historical reliability indicators based on the historical operation data of the equipment (Such as the successful execution rate of collaborative scheduling in the past 30 days). Then, according to the weight coefficients (volatility factor weight), (weight of control capability factor), (Reliability factor weight), weighted calculation of each indicator, and finally get the Energy in a specific time window The collaborative contribution value , to achieve quantitative calculation of the synergistic value of different energy sources in multi-scale time windows, and provide data support for the identification of synergistic relationships.

[0043] In one possible implementation, step S330 further includes:

[0044] Step S331: Calculate the prediction confidence of each of the multi-scale time windows according to the environmental prediction results to generate a prediction confidence identifier.

[0045] Step S332: After correcting the window trust factor using the predicted trust identifier, weighted calculation is performed on the collaborative contribution value to generate a first collaborative relationship identification result.

[0046] Step S333: performing scene similarity matching under cloud sharing based on the device data, and establishing a second collaborative relationship recognition result based on the scene similarity matching result.

[0047] Step S334: After collaborative authentication of the first collaborative relationship identification result and the second collaborative relationship identification result, collaborative relationship identification is completed.

[0048] Specifically, based on environmental prediction results (such as the predicted values ​​of parameters such as light intensity and wind speed within multi-scale time windows), prediction confidence is calculated for different time windows (e.g., minute-level and hour-level). The environmental prediction data for each time window is compared with historically collected data, and the root mean square error (RMSE) and mean absolute error (MAE) metrics are used to quantify the deviation between the predicted and measured values. Combined with the confidence parameters of the prediction model (such as the accuracy and confidence interval during model training), a standardized prediction confidence indicator is generated. For example, for a wind speed forecast within a 10-minute window, if the RMSE is 1.2 m / s and the model confidence is 90%, a prediction confidence indicator of 0.92 is obtained through normalization. This indicator, using a value in the interval [0, 1], represents the reliability of the environmental prediction results for each time window, providing a quantitative basis for the dynamic correction of the confidence factor in subsequent windows.

[0049] The specific implementation method for using the prediction trust identifier to correct the window trust factor and generate the first synergy relationship identification result is as follows: First, based on the obtained multi-scale time window prediction trust identifier (i.e., the reliability value of the environmental prediction result in each time window, such as 0.8 for the minute-level window prediction trust and 0.9 for the hour-level window), the initial window trust factor is adjusted according to the preset correction rules. For example, if the initial trust factor of an hour-level window is 0.8 and its prediction trust identifier is 0.9, the initial trust factor and the prediction trust factor are weighted and combined in a certain proportion (e.g., 50% each) to obtain a corrected trust factor of 0.85, which is more consistent with the reliability of the current prediction scenario. Next, the corrected trust factors for each time window are used to perform a weighted calculation of the synergy contribution value for the corresponding window. The synergy contribution value of each energy device in different time windows is weighted according to the corrected trust factor of the window in which it is located. The synergy contribution value of the window with the higher trust factor is calculated with a greater weight. The first collaborative relationship identification result is generated by multiplying the collaborative contribution value of each time window by the corresponding trust factor and then accumulating them. This result reflects the differences in the collaborative contribution weights of each energy device in different time windows in the form of quantitative data. For example, it presents the collaborative contribution ranking of each energy source in minute-level and hour-level windows in matrix form, providing a quantitative basis for the reliability of the time dimension for subsequent collaborative relationship certification.

[0050] Based on the equipment data of wind power equipment and photovoltaic equipment (such as real-time power output curves, operating status parameters, environmental sensor data, etc.), scenario similarity matching is performed through a historical scenario database shared in the cloud. Using similarity algorithms (such as Euclidean distance and cosine similarity), the current equipment status data is matched with the historical scenario data for feature vector matching, and the historical scenario that is most similar to the current environmental prediction results (such as light intensity and wind speed) and equipment operating status is found (for example, scheduling scenarios under the same wind speed range and similar light intensity in history). Based on the matching results, the synergistic relationship patterns of each energy device in the historical scenario are extracted (such as the power complementarity relationship between photovoltaic and wind power under specific weather conditions), and a second synergistic relationship identification result is established. The historical synergistic experience is used as a reference for the current synergistic relationship identification, providing empirical support in the scenario dimension for subsequent synergistic relationship certification.

[0051] When performing collaborative authentication on the first collaborative relationship identification result (synergy contribution analysis based on time window weights) and the second collaborative relationship identification result (synergy pattern result based on historical scenario similarity matching), both are first converted into feature vectors of uniform dimension, for example, mapped into matrices representing the synergy strengths between energy devices. Cross-validation is used to check the consistency of the energy device collaborative relationships between the two results. For example, if the time windows with high synergy contributions between photovoltaic and wind power in the first result also have similar synergy patterns in similar historical scenarios in the second result, the consistency between the two is high (e.g., the overlap between the synergy relationships exceeds a preset threshold of 70%), the two are fused according to preset weights (e.g., 60% for the first result and 40% for the second result) to generate the final collaborative relationship matrix. If the consistency is low, an exception handling mechanism is activated, rechecking the environmental prediction data or historical scenario matching parameters until both pass collaborative authentication. Ultimately, the collaborative relationships between the new energy devices are accurately identified, providing a reliable basis for the initialization of the dynamic response virtual collaborative entity.

[0052] In one possible implementation, step S500 further includes:

[0053] Step S510: calling the virtual layer in the dynamic response virtual coordination body to perform task virtualization of the time sequence load demand task and generate N virtual control task units.

[0054] Step S520: calling the supply response layer in the dynamic response virtual collaborative body, performing supply source response analysis under the environmental prediction result, and establishing a supply source response feature pool.

[0055] Step S530: using the matching layer initialized by the collaborative relationship identification result to perform adaptive matching of the N virtual control task units and the supply source response feature pool, and generating the collaborative scheduling result according to the adaptive matching result.

[0056] Specifically, the virtual layer within the dynamic response virtual collaborative body is invoked to virtualize time-series load demand tasks (e.g., electricity load change curves in different time periods). The virtual layer uses time slicing and power granularity to split continuous load demand into N discrete virtual control task units at preset time granularities (e.g., 10 minutes, 30 minutes). Each virtual control task unit contains key attributes such as target power value, time window, and task priority. For example, a 150kW load demand from 9:00-10:00 AM on a certain day is divided into four task units at a 15-minute time granularity, each corresponding to a 25kW power control target and a corresponding time interval. This converts actual load demand into standardized, quantifiable virtual tasks, providing a unified input format for subsequent matching and scheduling with new energy supply sources.

[0057] The supply response layer within the dynamic response virtual collaborative body analyzes supply source responses based on environmental forecast results (e.g., future trends in parameters such as light intensity, wind speed, and temperature), thereby establishing a supply source response feature pool. The supply response layer first simulates the power output characteristics of each supply source under different forecasted environmental conditions based on physical models of wind and photovoltaic equipment (e.g., the IV characteristic curve of photovoltaic cells and the power-wind speed curve of wind turbines) and historical operating data. For example, the output power range and fluctuation amplitude of photovoltaic equipment in each time window are calculated based on the forecasted light intensity, while the dispatchable power range and response delay of wind turbines are determined based on the forecasted wind speed curve. Next, the response capability characteristics of each supply source in different time windows (e.g., power adjustable range, output stability index, switching loss, minimum adjustment step size, etc.) are extracted to form standardized feature vectors, which are then organized into a supply source response feature pool in a time series. This feature pool integrates the dynamic response capabilities of each renewable energy supply source under the forecasted environment, providing a comprehensive supply-side feature description for the subsequent matching layer to perform virtual control task unit and supply source adaptation.

[0058] A matching layer initialized with the synergistic relationship identification results is used to perform adaptive matching between virtual control task units and the supply source response feature pool, generating a coordinated scheduling result. The specific implementation method is as follows: First, based on the synergistic relationship identification results (such as the synergistic weight matrix and complementary relationship strength between each new energy device), the matching rules and parameters of the matching layer are initialized. For example, the synergistic complementary weights of photovoltaic and wind power in a specific time window are used as matching priority parameters. Next, a distributed computing framework (such as Spark) is used to parallelize the processing of N virtual control task units (each unit contains attributes such as target power, time window, and priority) and the supply source response feature pool (which stores feature vectors such as the power adjustable range, response delay, and switching loss of each supply source in different time windows). Through an improved Hungarian or greedy algorithm, each virtual task unit is optimally matched with the corresponding time window features in the supply source feature pool. The matching process incorporates reliability indicators (such as the historical collaborative success rate) and scheduling cost constraints (such as the equipment start-stop loss function) in the collaborative relationship. For example, priority is given to supply sources that are complementary to the task demand fluctuation trend and have a high collaborative weight (when the photovoltaic power forecast increases, priority is given to matching intervals where the wind power forecast decreases and has a high collaborative weight). Finally, a dynamic programming algorithm is used to adjust the power allocation ratio and scheduling sequence of each supply source, generating a collaborative scheduling result that includes the scheduling time points, power output values, and collaborative coordination strategies of each new energy device, achieving precise matching of load demand and new energy supply, and actively smoothing grid-connected power fluctuations.

[0059] In one possible implementation, step S600 further includes:

[0060] Step S610: Record the active stabilization control management effect and generate a record data set.

[0061] Step S620: extracting environmental features of the time series environmental data set, and establishing a sudden weather change response feedback based on the recorded data set and the environmental features.

[0062] Step S630: performing window trust factor update of the active identification complementary window according to the sudden weather response feedback, and performing collaborative relationship identification management of the active identification complementary window according to the window trust factor update result.

[0063] Specifically, after using the coordinated dispatch results to actively smooth out grid-connected power fluctuations, the effectiveness of the active smoothing control is recorded to generate a record data set. Data acquisition devices are used to obtain real-time data on the actual fluctuation amplitude of the grid-connected power, power deviation values ​​before and after smoothing control, the actual output power of each new energy device, the charge and discharge status of the energy storage unit, and the execution of the dispatch strategy. This data is then organized and stored in chronological order to form a record data set containing fields such as timestamps, power parameters, and device operating status. This provides data support for subsequent analysis of smoothing control effects, optimization of control strategies, and response to sudden weather changes.

[0064] Environmental features (such as the magnitude of sudden changes in light intensity, the frequency of sudden changes in wind speed, and the value of sharp changes in temperature) are extracted from time-series environmental datasets and correlated with active leveling control management performance data in the recorded datasets (such as the magnitude of grid-connected power fluctuations, leveling response time, and equipment adjustment accuracy). By identifying the changing patterns of leveling control effectiveness under different environmental characteristics, especially sudden weather conditions (such as strong winds, heavy rain, and thunderstorms), (for example, when wind speed increases suddenly, wind power output fluctuations increase, leading to increased difficulty in grid-connected power leveling), sudden weather response feedback is established. This feedback is used to indicate deficiencies in active leveling control management under specific sudden weather conditions (such as delayed dispatch strategy response and insufficient equipment adjustment range) or areas that need adjustment, providing a basis for subsequent optimization of control strategies.

[0065] Based on the feedback from sudden weather events (such as the deviation analysis of the smoothing control effect under abnormal weather conditions such as strong winds and heavy rain), the window trust factor of the active identification complementary window is dynamically updated. If the feedback shows that the power forecast volatility of a certain time window increases significantly under sudden weather events (such as the prediction error of the minute-level window exceeds the threshold during thunderstorms), the trust factor of the window is reduced, for example, from the initial 0.8 to 0.5, to weaken its weight in the identification of collaborative relationships. After the update, the collaborative relationship identification management process is re-executed based on the new window trust factor: first, a multi-scale time window is created based on the active identification complementary window, and the parameters of each window are configured using the corrected trust factor. Then, the collaborative contribution value of each energy source in different windows is calculated through the universal collaborative contribution function, and the collaborative relationship identification is completed in combination with the equipment data, thereby optimizing the accuracy of collaborative relationship identification under sudden weather events and improving the dynamic adaptability of grid-connected power fluctuation smoothing.

[0066] In one possible implementation, step S600 further includes:

[0067] Step S640: Execute supply monitoring of the time-series load demand task to generate a supply monitoring data set.

[0068] Step S650: using the supply monitoring data set to identify the failure status of the collaborative scheduling result.

[0069] Step S660: Report an exception based on the failure status identification result and switch the scheduling strategy.

[0070] Specifically, supply monitoring of time-series load demand tasks is performed to generate a supply monitoring data set. Specifically, by deploying sensors and data acquisition devices on various new energy devices and grid-connected nodes, supply-side data such as the actual output power of wind power equipment and photovoltaic equipment, the charge and discharge status of energy storage units, and the power supply of the power grid are monitored in real time. At the same time, real-time load data related to time-series load demand tasks is collected, such as the deviation value between the actual power load and the scheduling demand in each time period. These real-time monitoring data are structured and stored in a time series to form a supply monitoring data set containing fields such as timestamp, equipment output power, energy storage status, and load deviation, providing data support for the subsequent identification of failure states of collaborative scheduling results.

[0071] To identify failure states in collaborative dispatch results using a supply monitoring dataset, a machine learning algorithm combining ensemble learning and time series analysis is employed. First, the isolation forest algorithm performs unsupervised anomaly detection on numerical features in the supply monitoring dataset, such as power deviation and energy storage status anomalies. This algorithm identifies sample points that significantly deviate from the normal distribution (e.g., PV output power in a certain period deviates by more than 20% from dispatch expectations). A time series prediction model is constructed using a long-short-term memory network (LSTM) to predict the normal output range for the current period based on historical supply data (e.g., wind power output sequences from the previous 12 hours). Any observed value outside the predicted confidence interval (e.g., a 95% confidence interval) is flagged as a potential failure point. For multi-dimensional features (e.g., power, load, and environmental parameters), an XGBoost algorithm is used to train a classification model. Inputs include real-time power deviation rate, dispatch execution delay, and historical failure mode features. The model outputs failure state probabilities (e.g., "serious failure," "mild anomaly," and "normal"). The SHAP value is used to interpret the contribution of each feature to failure identification (e.g., the power fluctuation index contributes 45%). Finally, the results of multiple algorithms are combined to generate a failure state identification report, providing data support for scheduling strategy switching.

[0072] Based on the failure status identification results from the coordinated scheduling (such as power deviation exceeding a threshold, equipment anomaly, etc.), anomalies are reported through audible and visual alarms, SMS notifications, or platform pop-up windows. Information such as the failure type, impact scope, and timeliness is also displayed. Simultaneously, the real-time energy storage status of the energy storage unit (such as remaining capacity and charge / discharge power limits) is obtained and combined with the failure status to calculate the compensatory dispatch capability. For example, if the energy storage capacity exceeds 30% and a failure results in a 50kW power shortfall, the available compensatory power and duration of the energy storage are evaluated to establish the first switching constraint. Next, the scalability of sequential load demand tasks is analyzed, distinguishing between critical and non-critical loads. The power adjustment range and time flexibility of each task are determined to establish the second switching constraint. Finally, based on these two types of constraints, an optimization algorithm is used to generate a new scheduling strategy, such as prioritizing energy storage to compensate for power shortfalls and adjusting the power consumption time periods of non-critical loads if energy storage is insufficient, to ensure grid-connected power fluctuations are effectively smoothed.

[0073] In one possible implementation, step S660 further includes:

[0074] Step S661: Acquire the energy storage state of the energy storage unit, perform compensation scheduling evaluation based on the energy storage state and the failure state identification result, and establish a first switching constraint.

[0075] Step S662: performing task adjustment analysis of the sequential load demand task and establishing a second switching constraint.

[0076] Step S663: Complete the switching of the scheduling strategy according to the first switching constraint and the second switching constraint.

[0077] Specifically, the battery management system (BMS) first collects energy storage unit status data, including remaining capacity (SOC), charge and discharge power limits, and battery state of health (SOH), in real time and transmits it to the central control system. Next, these real-time energy storage states, along with failure state identification results from the coordinated scheduling (e.g., power shortfall size and failure duration), are input into the compensation scheduling evaluation model. The model performs calculations based on preset rules (e.g., energy storage is considered capable of compensation when SOC > 30% and charge and discharge power ≥ 50% of the failure power shortfall). Finally, a first switching constraint is established based on the evaluation results. For example, when the energy storage's compensatory power ≥ 80% of the failure power shortfall, the constraint is that energy storage peak shaving is enabled and the charge and discharge power ≤ 70% of the rated power, ensuring that the energy storage can safely and effectively participate in scheduling strategy switching.

[0078] Sequential load demand tasks are prioritized into critical loads (e.g., medical equipment, communication base stations) and non-critical loads (e.g., commercial lighting, energy storage charging). A load elasticity matrix is ​​constructed by analyzing the priority tags and power characteristics (e.g., power demand, duration, and allowed interruption time) in the load task files. A linear programming algorithm is then used to calculate the adjustable range for each load. For example, non-critical loads can delay their power onset by ±30 minutes and reduce their power by 20%-50%, while critical loads must maintain a constant power supply. A second switching constraint is then established, combining failure state identification results (e.g., failure duration and power shortfall). When the failure duration is less than one hour, the constraint is that non-critical loads reduce their power by 30%, while critical loads remain unchanged. When the failure duration is ≥1 hour, the constraint is that non-critical loads delay their power onset, while critical loads utilize backup power sources. This entire process, through load characteristic analysis, ensures that the switching constraints adapt to supply-side fluctuations while maximizing the demand for critical loads.

[0079] The scheduling strategy is switched based on the first and second switching constraints. Specifically, the system first integrates the first switching constraint of energy storage compensation capacity (e.g., energy storage capacity must be greater than 30% and the available peak-shaving power must meet at least 50% of the power shortfall) and the second switching constraint of load adjustment flexibility (e.g., non-critical loads can delay their power usage by ±30 minutes or reduce their power by 20%-50%). These constraints are then converted into constraints within a mathematical model. A genetic algorithm or linear programming algorithm is then used to search for the optimal strategy within the feasible solution space. If energy storage compensation capacity is sufficient (e.g., SOC > 30% and compensation power ≥ 80% of the shortfall), energy storage is prioritized for power balancing and maintaining constant power for critical loads. If energy storage is insufficient, the power usage period or power of non-critical loads is adjusted, and backup energy sources (e.g., gas generators) are activated. During the switchover, the new strategy is verified in real time to ensure that all constraints are met, ensuring that grid-connected power fluctuations are within acceptable limits. This allows for smooth switching of scheduling strategies and ensures stable system operation.

[0080] In one possible implementation, step S600 further includes:

[0081] Step S670: establishing a collaborative contribution graph according to the collaborative scheduling result, and generating a scheduling return trajectory.

[0082] Step S680: Perform visual management based on the collaborative contribution graph and the scheduling return track.

[0083] Specifically, when establishing a collaborative contribution graph based on the collaborative scheduling results, the collaborative contribution values ​​of each new energy device (wind energy, photovoltaics, etc.) in different time windows (calculated by a general collaborative contribution function) are first mapped to the size or color depth of the nodes in the graph. The lines between the nodes represent the strength of the collaborative relationship between the devices, thereby intuitively displaying the contribution ratio of each energy source in smoothing power fluctuations and their mutual collaborative relationship; at the same time, a scheduling return trajectory is generated to record the execution process of the scheduling strategy, the power output changes of each device, the charging and discharging status of the energy storage unit and other key parameters in chronological order to form traceable historical trajectory data.

[0084] Visual management is performed based on the collaborative contribution graph and scheduling return trajectory. The collaborative contribution graph is displayed in the form of a dynamic chart through the web front-end or dedicated monitoring interface, allowing users to zoom in and out and click to view the details of each node and understand the contribution of each energy source in different time periods. The scheduling return trajectory is presented in the form of a timeline, and the scheduling process can be replayed in a specific time period. This allows operation and maintenance personnel to intuitively grasp the overall effect of grid-connected power leveling, strategy execution efficiency and equipment coordination, providing a visual decision-making basis for optimizing scheduling strategies and equipment management.

[0085] The second embodiment is based on the same inventive concept as the method for actively smoothing grid-connected power fluctuations for multi-source renewable energy in the previous embodiment. Figure 2 As shown, the present application provides a grid-connected power fluctuation active smoothing control system for multi-source renewable energy. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0086] The device data reading module 10 is used to read the device data of wind power equipment and photovoltaic equipment after connecting to the multi-source new energy management equipment.

[0087] The environment prediction result establishing module 20 is used to perform data collection of environmental data, establish a time series environment data set, perform environment prediction based on the time series environment data set, and establish an environment prediction result.

[0088] The collaborative relationship identification result establishing module 30 is used to perform collaborative relationship identification based on active identification of complementary windows according to the device data and the environmental prediction result, and establish a collaborative relationship identification result.

[0089] The initialization module 40 is configured to input the cooperative relationship identification result into the dynamic response virtual cooperative body and perform initialization of the dynamic response virtual cooperative body.

[0090] The collaborative scheduling result generating module 50 is used to obtain a time sequence load demand task, and input the time sequence load demand task and the environment prediction result as input data into the initialized dynamic response virtual collaborative body to generate a collaborative scheduling result.

[0091] The active smoothing control management module 60 is used to perform active smoothing control management of grid-connected power fluctuations using the coordinated scheduling results.

[0092] Furthermore, the system is also used to implement the following functions:

[0093] A multi-scale time window is created based on the active identification complementary window, and a window trust factor of the multi-scale time window is configured; the multi-scale time window is used to calculate the collaborative contribution based on the environmental prediction result, and a collaborative contribution value of the multi-scale time window mapping is generated; and collaborative relationship identification is completed based on the window trust factor, the collaborative contribution value, and the device data.

[0094] Furthermore, the system is also used to implement the following functions:

[0095] The general collaborative contribution function of the time window is established as follows:

[0096] ;in, Characterization Energy in the time window The collaborative contribution value of is the weight coefficient of the volatility impact factor, Characterization Energy in the time window The power of the volatility indicator, Characterization Energy in the time window The dispatchable index, is the weight coefficient of the control capability factor, For the The reliability index of each energy source's historical participation in coordination, is the weight coefficient of the reliability factor; the collaborative contribution calculation is completed according to the general collaborative contribution function.

[0097] Furthermore, the system is also used to implement the following functions:

[0098] According to the environmental prediction results, the prediction trust calculation of the multi-scale time window is performed respectively to generate a prediction trust identifier; after correcting the window trust factor using the prediction trust identifier, the collaborative contribution value is weightedly calculated to generate a first collaborative relationship identification result; according to the device data, scene similarity matching is performed under cloud sharing, and a second collaborative relationship identification result is established based on the scene similarity matching result; after collaborative authentication of the first collaborative relationship identification result and the second collaborative relationship identification result, the collaborative relationship identification is completed.

[0099] Furthermore, the system is also used to implement the following functions:

[0100] Call the virtual layer in the dynamic response virtual collaborative body to perform task virtualization of the time-series load demand task and generate N virtual control task units; call the supply response layer in the dynamic response virtual collaborative body to perform supply source response analysis under the environmental prediction result and establish a supply source response feature pool; use the matching layer initialized by the collaborative relationship identification result to perform adaptation and matching of the N virtual control task units and the supply source response feature pool, and generate the collaborative scheduling result according to the adaptation and matching result.

[0101] Furthermore, the system is also used to implement the following functions:

[0102] The effects of active smoothing control management are recorded to generate a record data set; environmental features of the time series environmental data set are extracted, and sudden weather response feedback is established based on the record data set and the environmental features; window trust factors of the active identification complementary windows are updated based on the sudden weather response feedback, and collaborative relationship identification management of the active identification complementary windows is performed based on the window trust factor update results.

[0103] Furthermore, the system is also used to implement the following functions:

[0104] Execute supply monitoring of the time-series load demand task to generate a supply monitoring data set; use the supply monitoring data set to identify the failure state of the collaborative scheduling result; report an exception based on the failure state identification result and switch the scheduling strategy.

[0105] Furthermore, the system is also used to implement the following functions:

[0106] Acquire the energy storage status of the energy storage unit, perform compensation scheduling evaluation based on the energy storage status and the failure status identification results, and establish a first switching constraint; perform task adjustment analysis of the time-series load demand task and establish a second switching constraint; complete the switching of the scheduling strategy based on the first switching constraint and the second switching constraint.

[0107] Furthermore, the system is also used to implement the following functions:

[0108] A collaborative contribution graph is established according to the collaborative scheduling result, and a scheduling return trajectory is generated; and visual management is performed based on the collaborative contribution graph and the scheduling return trajectory.

[0109] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0110] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0111] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for actively smoothing grid-connected power fluctuations for multi-source renewable energy, characterized in that: The method comprises: After accessing multi-source new energy management equipment, read the equipment data of wind power equipment and photovoltaic equipment; Performing data collection of environmental data, establishing a time series environmental data set, performing environmental prediction based on the time series environmental data set, and establishing an environmental prediction result; Performing collaborative relationship identification based on active identification of complementary windows according to the device data and the environmental prediction result, and establishing a collaborative relationship identification result; Inputting the cooperative relationship identification result into the dynamic response virtual cooperative body, and executing the initialization of the dynamic response virtual cooperative body; Obtaining a time-series load demand task, and inputting the time-series load demand task and the environmental prediction result as input data into the initialized dynamic response virtual collaborative body to generate a collaborative scheduling result; The coordinated dispatching result is used to perform active smoothing control management of grid-connected power fluctuations.

2. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 1, characterized in that: The step of performing collaborative relationship identification based on active identification of complementary windows according to the device data and the environmental prediction result, and establishing a collaborative relationship identification result, includes: Creating a multi-scale time window according to the active identification complementary window, and configuring a window trust factor of the multi-scale time window; Utilizing the multi-scale time windows to calculate collaborative contribution based on the environmental prediction results, respectively, to generate collaborative contribution values ​​mapped by the multi-scale time windows; The collaborative relationship is identified based on the window trust factor, the collaborative contribution value, and the device data.

3. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 2, characterized in that: The calculating of the collaborative contribution based on the environmental prediction results by respectively utilizing the multi-scale time windows includes: The general collaborative contribution function of the time window is established as follows: ； in, Characterization Energy in the time window The collaborative contribution value of is the weight coefficient of the volatility impact factor, Characterization Energy in the time window The power of the volatility indicator, Characterization Energy in the time window The dispatchable index, is the weight coefficient of the control capability factor, For the The reliability index of each energy source's historical participation in coordination, is the weight coefficient of the reliability factor; The synergy contribution calculation is completed according to the general synergy contribution function.

4. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 2, characterized in that: The completing the collaborative relationship identification according to the window trust factor, the collaborative contribution value, and the device data includes: Calculate the prediction confidence of the multi-scale time window according to the environmental prediction results, and generate a prediction confidence identifier; After correcting the window trust factor using the predicted trust identifier, weighted calculation is performed on the collaborative contribution value to generate a first collaborative relationship identification result; Performing scene similarity matching under cloud sharing based on the device data, and establishing a second collaborative relationship recognition result based on the scene similarity matching result; After collaborative authentication is performed on the first collaborative relationship identification result and the second collaborative relationship identification result, collaborative relationship identification is completed.

5. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 1, characterized in that: The step of inputting the time-series load demand task and the environmental prediction result as input data into the initialized dynamic response virtual collaborative entity to generate a collaborative scheduling result includes: Calling the virtual layer in the dynamic response virtual collaborative body to perform task virtualization of the time sequence load demand task and generate N virtual control task units; Calling the supply response layer in the dynamic response virtual collaborative body to perform supply source response analysis under the environmental prediction results and establish a supply source response feature pool; Adaptive matching of the N virtual control task units and the supply source response feature pool is performed using a matching layer initialized by the collaborative relationship identification result, and the collaborative scheduling result is generated according to the adaptive matching result.

6. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 1, characterized in that: After the coordinated scheduling result is used to perform active control and management of grid-connected power fluctuations, the method includes: Record the effects of active leveling control management and generate record data sets; Extracting environmental features of a time series environmental data set, and establishing a sudden weather response feedback based on the recorded data set and the environmental features; The window trust factor of the active identification complementary window is updated according to the sudden weather response feedback, and the collaborative relationship identification management of the active identification complementary window is performed according to the window trust factor update result.

7. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 1, characterized in that: The active smoothing control management of grid-connected power fluctuations using the coordinated scheduling results also includes: Executing supply monitoring of the time-series load demand task to generate a supply monitoring data set; Using the supply monitoring data set to identify failure states of the collaborative scheduling results; Report exceptions based on the failure status identification results and switch scheduling strategies.

8. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 7, characterized in that: The switching scheduling strategy includes: Acquiring an energy storage state of the energy storage unit, performing compensation scheduling evaluation according to the energy storage state and the failure state identification result, and establishing a first switching constraint; Performing task adjustment analysis on the time sequence load demand task and establishing a second switching constraint; The switching of the scheduling strategy is completed according to the first switching constraint and the second switching constraint.

9. The method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as claimed in claim 1, characterized in that: The active smoothing control management of grid-connected power fluctuations using the coordinated scheduling results also includes: Establishing a collaborative contribution graph based on the collaborative scheduling results and generating a scheduling return trajectory; Visual management is performed based on the collaborative contribution graph and the scheduling return visit trajectory.

10. A grid-connected power fluctuation active smoothing control system for multi-source renewable energy, characterized by: The system is used to implement the method for actively smoothing grid-connected power fluctuations for multi-source renewable energy as described in any one of claims 1 to 9, and the system includes: The device data reading module is used to read the device data of wind power equipment and photovoltaic equipment after connecting to the multi-source new energy management equipment; An environmental prediction result establishment module is used to perform data collection of environmental data, establish a time series environmental data set, perform environmental prediction based on the time series environmental data set, and establish an environmental prediction result; A collaborative relationship identification result establishing module, configured to perform collaborative relationship identification based on active identification of complementary windows according to the device data and the environmental prediction result, and establish a collaborative relationship identification result; an initialization module, configured to input the cooperative relationship identification result into the dynamic response virtual cooperative body and perform initialization of the dynamic response virtual cooperative body; A collaborative scheduling result generation module is used to obtain a time-series load demand task, input the time-series load demand task and the environmental prediction result as input data into the initialized dynamic response virtual collaborative body, and generate a collaborative scheduling result; An active smoothing control management module is used to use the collaborative scheduling results to perform active smoothing control management of grid-connected power fluctuations.

Images