A method and platform for coordinated power generation control of new energy power plants

Abstract

This invention provides a method and platform for coordinated power generation control of renewable energy power plants, relating to the field of power generation control technology. The method includes: acquiring a set of K power generation units of a target renewable energy power plant; traversing the operational data of the K power generation units, and time-series sensing and aggregation of the energy storage status and grid load demand of the target renewable energy power plant, to obtain aggregated operational data of the K power generation units, aggregated energy storage status data of the power plant, and aggregated grid load demand data; using a digital twin platform for the power plant to identify the operating condition prototype and determine an initial coordinated power generation control strategy; and performing twin simulation to correct the initial coordinated power generation control strategy and determine the target coordinated power generation control strategy. This invention solves the technical problem that existing technologies mainly rely on static historical data, which cannot respond promptly to dynamic changes in renewable energy power generation, leading to lags in the coordinated scheduling between renewable energy power plants, thus affecting grid stability and power generation efficiency.

Description

Technical Field

[0001] This invention relates to the field of power generation regulation technology, and specifically to a method and platform for coordinated power generation regulation of new energy power plants. Background Technology

[0002] A renewable energy cluster is a power generation system composed of multiple power stations, each consisting of multiple power generation units, including wind farms and photovoltaic power plants. The advantages of renewable energy lie in its abundant resources and clean, environmentally friendly nature; however, it suffers from volatility and intermittency, making the power output of renewable energy clusters difficult to predict and posing challenges to grid load dispatching and the stable operation of the power system. Many traditional power generation coordination and control methods rely on historical data and static models, but these methods fail to fully utilize real-time data for condition adaptation and cannot respond promptly to dynamic changes in grid load and renewable energy generation. This results in a lag in the coordinated dispatching of power generation units, energy storage systems, and grid load, leading to an inability to balance supply and demand in a timely manner, thus affecting grid stability and power generation efficiency. Summary of the Invention

[0003] This application provides a method and platform for coordinated power generation control of new energy power plants, aiming to solve the technical problem that existing technologies mainly rely on static historical data, which cannot respond to dynamic changes in new energy power generation in a timely manner, resulting in lag in the coordinated scheduling between new energy power plants, and thus affecting grid stability and power generation efficiency.

[0004] The first aspect disclosed in this application provides a method for coordinated power generation control of a new energy power plant cluster. The method includes: acquiring a set of K power generation units from K power plants in a target new energy power plant cluster; within a preset sensing window, traversing the operating data of the K power generation unit sets, as well as the energy storage status and grid load demand of the target new energy power plant cluster, and performing time-series sensing aggregation to obtain aggregated operating data of the K power generation units, aggregated energy storage status data of the power plant cluster, and aggregated grid load demand data; using a digital twin platform for the power plant cluster to perform condition prototype identification on the aggregated operating data of the K power generation units, the aggregated energy storage status data of the power plant cluster, and the aggregated grid load demand data, and determining an initial coordinated power generation control strategy based on the identification results; transmitting the initial coordinated power generation control strategy and the aggregated operating data of the K power generation units, the aggregated energy storage status data of the power plant cluster, and the aggregated grid load demand data to the digital twin platform for twin simulation, and correcting the initial coordinated power generation control strategy based on the twin simulation results to determine a target coordinated power generation control strategy.

[0005] The second aspect of this application discloses a power generation coordination and control platform for a new energy power plant cluster. The platform is used in the aforementioned power generation coordination and control method for a new energy power plant cluster. The platform includes: a power generation unit acquisition module, used to acquire a set of K power generation units from K power plants in a target new energy power plant cluster; and a time-series perception and aggregation module, used to traverse the operating data of the K power generation unit sets, as well as the energy storage status and grid load demand of the target new energy power plant cluster within a preset perception window, to perform time-series perception and aggregation, thereby obtaining aggregated operating data of the K power generation units, aggregated energy storage status data of the power plant cluster, and aggregated grid load demand data. The system includes: a prototype identification module for identifying operating conditions of the K power generation units, the energy storage status data, and the grid load demand data using a digital twin platform; and an initial power generation coordinated control strategy determined based on the identification results. A control strategy acquisition module transmits the initial power generation coordinated control strategy, the K power generation units' operating data, the energy storage status data, and the grid load demand data to the digital twin platform for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] By acquiring the set of K power generation units from K power plants within the target renewable energy cluster, a comprehensive understanding of the cluster's actual operation can be achieved, ensuring integrated monitoring of various power generation units and providing accurate input data for subsequent control strategies. Through time-series sensing aggregation, the operational data of the K power generation unit sets within a preset sensing window, along with the energy storage status and grid load demand of the target renewable energy cluster, can be effectively integrated to generate precise aggregated data. This not only provides a comprehensive view of the status of each power generation unit and the system, but also, through time-series data analysis within a preset sensing window, captures dynamic changes in power generation, energy storage, and grid load, improving the system's timeliness and responsiveness, and providing a data foundation for subsequent operating condition analysis and optimization of coordinated control strategies. Utilizing digital twin technology, the historical operating conditions of the target renewable energy cluster can be virtually reconstructed for efficient operating condition prototype identification. This is achieved by aggregating the operational data of the K power generation units and the energy storage status of the cluster. By aggregating data on aggregated data and grid load demand for prototype identification of operating conditions, the similarity between current and historical operating conditions can be identified. This allows for the rapid identification of the closest historical operating condition pattern when facing complex operating states, and the formulation of initial generation coordination and control strategies based on this pattern. This not only improves the rationality of the strategy but also accelerates the strategy generation process. In the digital twin platform of the power grid cluster, the initial generation coordination and control strategy is simulated based on the current system state. This simulation can verify the impact of the initial generation coordination and control strategy on grid load, generation unit operation, and energy storage status. The initial generation coordination and control strategy is iteratively revised based on the twin simulation results, and the target generation coordination and control strategy is finally determined. This process ensures that the target generation coordination and control strategy can adapt to the current grid demand and be flexibly adjusted according to real-time data, improving the automation and intelligence level of the control strategy and reducing risks and unnecessary costs during operation.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 This is a schematic flowchart of a power generation coordination and control method for a new energy power plant cluster, provided as an embodiment of this application.

[0010] Figure 2 This is a schematic diagram of a power generation coordination and control platform for new energy power plants, provided as an embodiment of this application.

[0011] Figure labeling: 10 for power generation unit acquisition module, 20 for time sequence perception aggregation module, 30 for operating condition prototype identification module, and 40 for control strategy acquisition module. Detailed Implementation

[0012] This application provides a method and platform for coordinated power generation control of new energy power plants, which solves the technical problem that the existing technology mainly relies on static historical data, which cannot respond to the dynamic changes of new energy power generation in a timely manner, resulting in lag in the coordinated scheduling between new energy power plants, and thus affecting the stability of the power grid and the efficiency of power generation.

[0013] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0014] Example 1, as Figure 1 As shown in the figure, this application provides a method for coordinated power generation control of new energy power plants, the method comprising:

[0015] Obtain the set of K power generation units from K power plants in the target new energy power cluster.

[0016] A target new energy power cluster refers to a regional-level power generation system comprising multiple power stations. The task of this target new energy power cluster is to generate and manage all wind and solar power resources within the region. The cluster consists of multiple power stations, each containing multiple power generation units. These power generation units can be wind power units (wind turbine generators) and solar power units (solar panels). This means that each power station will have both wind power and solar power generation units. The scale and type of each power station are different, but they are all part of the new energy power generation system as a whole. K represents the number of power stations in the cluster, and the set of K power generation units refers to the set of all power generation units in the entire cluster. These power generation units are used for coordinated control.

[0017] Within a preset sensing window, the operation data of the K power generation unit set, as well as the energy storage status and grid load demand of the target new energy field group, are traversed and time-series sensing data are aggregated to obtain the aggregated operation data of the K power generation units, the aggregated energy storage status data of the field group, and the aggregated grid load demand data.

[0018] The preset perception window is a time period, usually an hour or longer. It is the time window for the system to analyze and perceive data, representing the perception and collection of data within that time period. The size and time span of the preset perception window will affect the data processing and optimization results.

[0019] Within a preset sensing window, the operating data of K sets of power generation units are traversed. The operating data includes: wind power unit data, such as wind speed, output power of wind turbines, and operating status; photovoltaic unit data, such as solar radiation intensity, power output of photovoltaic panels, and temperature; and the time series of the operating data, that is, the power output and status data of each power generation unit at each moment during the sensing window.

[0020] Within the same preset sensing window, the energy storage status of the target new energy power plant cluster is collected. The energy storage system is a system in the target new energy power plant cluster used to store excess electrical energy, such as a battery energy storage system. The energy storage status includes information such as the charging and discharging status of the energy storage system and the remaining power. The energy storage status determines whether it can provide additional power support when demand is high. Within the same preset sensing window, the grid load demand is obtained. The grid load demand refers to the power demand of the target grid in a specific time period. It is affected by various factors such as weather, electricity consumption habits, and industrial production. By analyzing the grid load demand, the scheduling strategies of power generation units and energy storage devices can be adjusted to ensure the balance between power supply and demand.

[0021] Time-series sensing aggregation processes the operational data of the aforementioned K power generation units, the energy storage status of the target renewable energy cluster, and the grid load demand in chronological order. This data is continuous and needs to be aggregated based on the characteristics of time series data to form unified aggregated data. Specifically, the K power generation unit operational aggregation data includes the aggregated results of power output, operational status, and efficiency of all wind and solar power units within the sensing window; the cluster energy storage status aggregation data includes the charging and discharging status and remaining power of all energy storage devices in the entire target renewable energy cluster; and the grid load demand aggregation data includes the grid's power demand within the sensing window, helping to determine the changing trend of grid load.

[0022] Using the digital twin platform of the power generation cluster, the operating condition prototypes of the K power generation units, the power storage status, and the grid load demand are identified, and the initial power generation coordinated control strategy is determined based on the identification results.

[0023] A digital twin platform for a power plant cluster refers to a platform that virtually replicates a target renewable energy power plant cluster on a digital platform. Through real-time data, models, and algorithms, it simulates and predicts the behavior of the physical system based on the digital twin. This platform is used to simulate and analyze the actual operation of the target renewable energy power plant cluster. The platform continuously receives actual operation data for simulation, including aggregated data on the operation of K power generation units, aggregated data on the energy storage status of the power plant cluster, and aggregated data on grid load demand, to simulate the operation process of the target renewable energy power plant cluster.

[0024] The operating condition prototype refers to the typical performance of a target new energy power plant cluster under specific operating environments or conditions. It is usually a collection of historical data that reflects the system's working mode under certain load, grid demand, climate and other conditions. Specifically, using the digital twin platform of the power plant cluster, pattern recognition is performed on the aggregated data of the operation of K power generation units, the aggregated data of the energy storage status of the power plant cluster, and the aggregated data of grid load demand. By comparing the current data with historical operating conditions, the digital twin platform of the power plant cluster identifies the operating condition prototype of the current system. This process is similar to matching, that is, finding the historical operating condition that is most similar to the current system performance and using it as the reference pattern for the current system.

[0025] Once the operating condition prototype is identified, an initial power generation coordinated control strategy is formulated based on the known control strategies in historical operating conditions. This initial power generation coordinated control strategy is mainly based on the experience of historical patterns and may not be the most perfect, but it provides a basis for subsequent optimization and correction.

[0026] The initial power generation coordinated control strategy, the aggregated operation data of K power generation units, the aggregated energy storage status data of the power grid, and the aggregated grid load demand data are transmitted to the power grid digital twin platform for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy.

[0027] The initial power generation coordinated control strategy, along with the aggregated operational data of K power generation units, the aggregated data of the energy storage status of the power grid, and the aggregated data of the grid load demand, are transmitted to the power grid digital twin platform. By combining real-world data with digital models, the purpose of the twin simulation is to evaluate the performance of the current initial power generation coordinated control strategy in actual operation. The twin simulation results obtained through the twin simulation can reflect the actual operating performance of the system. The operating performance specifically includes: whether the power generation and grid load demand can be balanced, whether energy storage devices are used reasonably, whether each power generation unit works as expected, and whether there are any overload or inefficient operation situations.

[0028] Compare the twin simulation results with the expected control targets and assess the differences between them. If the twin simulation results show that the initial power generation coordinated control strategy cannot effectively meet the expected control targets or is unstable under certain conditions, it will be corrected. The correction process includes adjusting the output power of each power generation unit, optimizing energy storage scheduling, and modifying the scheduling strategy. After correction, the new strategy will be transmitted back to the digital twin platform of the power plant cluster, and twin simulation will be performed again until the expected control targets are achieved.

[0029] Through iterative twin simulation and correction processes, the final target power generation coordinated control strategy is output, which is the optimized final strategy. This target power generation coordinated control strategy achieves the best balance between various power generation units, energy storage devices and grid load demand, and can ensure the high efficiency and stability of system operation.

[0030] Furthermore, within a preset sensing window, the operational data of the K power generation unit set, as well as the energy storage status and grid load demand of the target new energy cluster, are traversed and time-series sensing aggregated to obtain aggregated operational data of the K power generation units, aggregated energy storage status data of the cluster, and aggregated grid load demand data, including:

[0031] The operation data of the K power generation units set within a preset sensing window is extracted and traversed to obtain a set of operation data sequences of the K power generation units; the operation data sequences of the K power generation units are time-series sensing and converged to determine the converged operation data of the K power generation units; the energy storage status and grid load demand of the target new energy field group are time-series sensing and converged within the preset sensing window to obtain the converged data of the field group's energy storage status and the converged data of the grid load demand.

[0032] The operation data of K power generation units in the target new energy power cluster are extracted within a preset sensing window. Each power generation unit will generate a series of operation data within the preset sensing window, including but not limited to: wind power unit data, such as wind speed, output power of wind turbine, and operating status; photovoltaic unit data, such as solar radiation intensity, power output of photovoltaic panels, and temperature; and the time series of operation data, that is, the power output and status data of each power generation unit at each moment during the sensing window.

[0033] The operation data of the K power generation units are extracted and organized to generate the power generation unit operation data sequence of each power generation unit within a preset perception window. The power generation unit operation data sequence refers to a set of continuous data of each power generation unit in the time dimension, which can be multi-dimensional data collected by minute, hour or other time granularity. For all K power generation units, the corresponding set of K power generation unit operation data sequences is obtained.

[0034] Time-aware aggregation is performed on a set of operational data sequences from K power generation units. The purpose of time-aware aggregation is to extract meaningful patterns or features from the raw time-series data and synthesize them into a single aggregation result. This process includes analyzing, statistically analyzing, and summarizing the operational data sequences of each power generation unit. For example, the mean of data from different time periods is calculated, such as the average power output per hour or per minute. Other statistical characteristics, such as standard deviation, maximum value, and minimum value, are calculated to characterize the stability and efficiency of the power generation units. The output change trends of different power generation units are compared to identify potential anomalies or faults. The aggregated data of the K power generation units is a summary result of the operational data sequences from the K power generation units, containing the key operational characteristics of each power generation unit within that time period, such as power output level, stability indicators, and performance trends.

[0035] Using the same method, the energy storage status and grid load demand of the target new energy power plant cluster are time-series sensed and aggregated to obtain cluster energy storage status aggregated data and grid load demand aggregated data. These aggregated data are similar to the power generation unit operation aggregated data, and will not be described in detail here for the sake of brevity.

[0036] Furthermore, time-series sensing aggregation is performed on the set of operational data sequences of the K power generation units to determine the aggregated operational data of the K power generation units, including:

[0037] The system iterates through the set of operating data sequences of the K power generation units to perceive the trend of operating data and determine the set of trend features of operating data of the K power generation units. It then iterates through the set of trend features of operating data of the K power generation units to determine the iterative trend features of operating data of the K power generation units. Finally, it calculates the mean of operating data for each of the set of operating data sequences of the K power generation units to obtain the mean of operating data of the K power generation units. Using the iterative trend features of operating data of the K power generation units, it implicitly converges the corresponding mean of operating data of the K power generation units to determine the converged operating data of the K power generation units.

[0038] Operational data trend perception refers to time-series analysis of the operational data sequences of each power generation unit to identify trend characteristics. Specifically, it involves analyzing the changing patterns, fluctuation amplitudes, and periodicity of the power generation unit data to determine the overall trend of each unit during the perception window. For example, it might identify that the power generation of a wind power unit shows an increasing trend with changes in wind speed, or that the output power of a photovoltaic unit exhibits a linear relationship with changes in solar radiation intensity. During operational data trend perception, corresponding trend characteristics are extracted from the operational data sequences of each power generation unit. These characteristics reflect the typical operating mode of the power generation unit over a period of time. Trend characteristics include average power output, maximum and minimum power values, volatility or standard deviation, and upward or downward trends in operation. After summarizing these trend characteristics, a set of trend characteristics for K power generation unit operational data is obtained, providing the foundational data for subsequent analysis and optimization.

[0039] Interactive iteration refers to iterating multiple times on the set of trend characteristics of the operating data of K power generation units. By analyzing the mutual influence or synergy between different power generation units, the operating trends of the power generation units are further optimized. This method helps to understand the interrelationships between power generation units and improve the coordination of scheduling strategies. The interactive iteration process includes feature combination and feature adjustment. Feature combination refers to pairing or combining the trend characteristics of different power generation units to identify their impact on the overall operation. For example, if the output power of a wind power unit is correlated with the output power of another photovoltaic unit, interactive iteration can identify these correlated characteristics. Feature adjustment refers to adjusting the trend characteristics of each power generation unit interactively to ensure that they match the characteristics of other power generation units. Through continuous iteration, the coordination relationship between power generation units in the system is optimized.

[0040] Through interactive iteration, based on the interaction and synergistic effects between power generation units, the iterative trend characteristics of each power generation unit are obtained, and the iterative trend characteristics of the operating data of K power generation units are determined. These iterative trend characteristics reflect the performance of power generation units in more complex and synergistic operating modes. For example, some power generation units may exhibit higher efficiency when operating in synergy than when operating alone. This synergistic effect can be captured through interactive iteration.

[0041] For each power generation unit's operational data sequence set, the mean of its operational data sequence is calculated. This mean represents the average operational data of the power generation unit within a preset sensing window. The mean calculation indicates the overall performance of each power generation unit within that time period. For example, if the operational data of a wind power unit includes power output at multiple time points, the average power output level within that time period is obtained through mean calculation, reflecting the long-term operational trend of the power generation unit. By calculating the mean, the average operational data of K power generation units is obtained, reflecting the long-term average state of each power generation unit.

[0042] Implicit state convergence refers to the aggregation of iterative trend features and the mean of operating data of K power generation units through a neural network model to obtain a more accurate and comprehensive operating state. In this process, the neural network model is used for pattern recognition. Specifically, the iterative trend features and the corresponding mean of operating data of K power generation units are input into the neural network model. The neural network model is trained and learned through a hierarchical structure to mine deep-level correlations and patterns in the data. In this way, the neural network model can extract the implicit state features of each power generation unit, that is, the deep-level information that cannot be directly identified from the original data.

[0043] The output of the neural network model yields aggregated data of K power generation units. This aggregated data combines the mean and iterative trend characteristics of the power generation units, representing the true performance of each power generation unit in a more complex operating environment.

[0044] Furthermore, the iterative trend characteristics of the operating data of the K power generation units are determined by interactively iterating through the sets of trend characteristics of the operating data of the K power generation units, including:

[0045] Each of the K power generation unit operation data trend feature sets is paired with other features within each set to obtain a set of K power generation unit operation data trend feature combinations. Sub-similarity is calculated on these sets, and the results are normalized. The processed results are then filled into an initially empty matrix to construct K interactive iteration matrices. These K interactive iteration matrices are then convolved with the K power generation unit operation data trend feature combination sets to obtain a set of K power generation unit operation data enhanced trend feature combinations. The mean of these K power generation unit operation data enhanced trend feature combination sets is calculated to obtain the iterative trend features of the K power generation unit operation data.

[0046] Pairwise feature combinations are performed on the set of trend features of the operational data of K power generation units. This means pairing the trend features of each power generation unit's operational data with those of other power generation units to form new combinations. For example, assuming there are three power generation units A, B, and C, each with a trend feature, such as the change trend of power output, the pairwise feature combinations include: feature A with feature B, feature A with feature C, and feature B with feature C. After these pairwise feature combinations, all the combinations constitute the set of trend feature combinations of the K power generation units' operational data. This set contains all possible trend feature pairs of the K power generation units, and the features of these combinations are used for subsequent analysis of the collaborative operation modes or mutual influence relationships between different power generation units.

[0047] Similarity calculation is performed on the trend feature combination of the operating data of each pair of power generation units. The similarity calculation is used to measure the correlation between the two. The similarity measurement methods include Euclidean distance, Pearson correlation coefficient, and cosine similarity. Each pair of trend feature combination of the operating data of power generation units will generate a set of similarity values, which represent the degree of similarity between the two power generation units in all operating features.

[0048] The calculated similarity values ​​are normalized so that all similarity values ​​are within a uniform standard range, usually between 0 and 1. The purpose of normalization is to eliminate scale differences that may exist in the original similarity calculation, so that the similarity of different operating characteristics between different power generation units can be compared fairly.

[0049] The normalized similarity values ​​are filled into an empty matrix, where each element represents the similarity of each operational feature between the two power generation units. The resulting similarity matrix is ​​actually an interactive iteration matrix that describes the interaction or synergistic effect between each pair of power generation units and is used for subsequent iterative calculations. Finally, for a set of K combinations of operational data trend features from power generation units, K interactive iteration matrices are obtained.

[0050] Convolution is an operation that transforms input data using a specific kernel (or filter). Convolution helps extract more complex feature patterns, thereby enhancing data features and enabling the utilization of richer and more representative trend information when optimizing collaborative control strategies. In this step, the convolution operation is used to apply an interactive iteration matrix to the combination of trend features in the power generation unit operation data to enhance the representation of these features. Specifically, the convolution operation performs a weighted multiplication of each element of the interactive iteration matrix with the corresponding feature in the combination of trend features in the power generation unit operation data, and then aggregates the results. This convolution result enhances power generation unit combinations that better conform to collaborative features, while reducing irrelevant or insignificant features. The convolution result produces a set of K enhanced trend feature combinations for power generation unit operation data. This set of enhanced feature combinations can better capture the complex interrelationships and collaborative effects between power generation units.

[0051] The mean of the enhanced trend feature combination set of K power generation unit operation data is calculated. The mean calculation process is to average all features in the enhanced trend feature combination set of power generation unit operation data to obtain an overall trend feature for each power generation unit. Through mean calculation, the iterative trend features of the K power generation unit operation data are obtained. These features can comprehensively reflect the operation performance of the power generation unit in multiple time periods and the synergistic effect with other power generation units.

[0052] Furthermore, the operating condition prototype is identified using the aggregated data of the K power generation units' operation, the aggregated data of the power storage status, and the aggregated data of the grid load demand, based on the identification results. The initial power generation coordinated control strategy is then determined, including:

[0053] A historical operating condition prototype library of the digital twin platform for the power plant cluster is obtained. Each historical operating condition prototype includes aggregated operational data of K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power grids. Using the aggregated operational data of the K power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power plant clusters as indexes, an approximate analysis is performed on each historical operating condition prototype in the historical operating condition prototype library, including the aggregated operational data of the K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power plant clusters. The historical operating condition prototype with the highest approximation is taken as the identification result. The prototype power generation coordinated control strategy corresponding to the identification result is taken as the initial power generation coordinated control strategy.

[0054] The digital twin platform for power plant clusters provides a historical operating condition prototype library, which contains operational data under different historical operating conditions. Each historical operating condition represents the typical operating state of the power plant cluster within a specific time period. Specifically, the aggregated operational data of K prototype power generation units describes the operating status of the power generation units under the corresponding historical operating conditions, including information such as power generation, equipment status, and operating efficiency. The aggregated energy storage status data of the prototype power plant cluster refers to the operational data of the energy storage devices within the cluster under the corresponding historical operating conditions, including the charging and discharging status of the energy storage devices and remaining power. The aggregated grid load demand data of the prototype reflects the grid load demand under the corresponding historical operating conditions, including peak load, trough load, and load fluctuation trends. These historical operating condition prototypes provide a reference for subsequent real-time operation, enabling rapid response to current operating conditions and optimization of control strategies.

[0055] The system uses aggregated data of the current operation of K generating units, aggregated data of the energy storage cluster status, and aggregated data of grid load demand as indexes to find the most similar historical operating condition in the historical operating condition prototype library. Approximate analysis involves calculating the similarity between the current system state and historical operating condition prototypes to find the closest operating condition. Similarity calculation methods include Euclidean distance, Pearson correlation coefficient, and cosine similarity. Through similarity calculation, it can be determined which historical operating condition prototype best matches the current operating state. Specifically, for each historical operating condition prototype, similarity calculations are performed between the aggregated data of the K generating units and the aggregated data of the K prototype generating units, the aggregated data of the energy storage cluster status and the prototype aggregated data of the energy storage cluster status, and the aggregated data of the grid load demand and the prototype grid load demand. The similarity values ​​of each pair are summed to obtain the total similarity between the historical operating condition prototype and the current system state. For all historical operating condition prototypes in the historical operating condition prototype library, the historical operating condition prototype with the highest total similarity is extracted as the identification result, representing the reference template for the current system state.

[0056] Each historical operating condition prototype includes not only operational data of the generation units, energy storage status, and grid load data, but also a corresponding prototype generation coordination and control strategy. This strategy reflects how the system coordinated the demands of various generation units, energy storage devices, and grid loads at that time to achieve optimal power supply and demand balance. The prototype generation coordination and control strategy under this historical operating condition prototype is used as the initial generation coordination and control strategy for the current system. This initial strategy may not be entirely applicable to the current environment, but it provides a starting point for subsequent regulation and control.

[0057] Furthermore, the historical operating condition prototype library of the aforementioned power plant cluster digital twin platform is obtained, wherein each historical operating condition prototype includes aggregated operational data of K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power grids, including:

[0058] The process involves acquiring a set of historical operating condition data from the digital twin platform for the power generation cluster. Each set of historical operating condition data includes aggregated data on the operation of K historical power generation units, aggregated data on the energy storage status of the power generation cluster, and aggregated data on the grid load demand. Using a preset similarity score as a constraint, the historical operating condition data set is aggregated to determine multiple aggregated historical operating condition data sets. The average of multiple aggregated historical power generation collaborative control strategies from these multiple aggregated historical operating condition data sets is calculated to obtain multiple prototype historical power generation collaborative control strategies. The average of these multiple aggregated historical operating condition data sets is then calculated to determine multiple historical operating condition prototypes. Finally, the multiple prototype historical power generation collaborative control strategies are associated with the multiple historical operating condition prototypes to construct a historical operating condition prototype library.

[0059] Historical operating condition data sets are obtained from the digital twin platform of the power plant cluster. These historical operating condition data describe the typical states of the power plant cluster in its past operation. Among them, the aggregated data of K historical power generation units describes the operating status of the power generation units under the corresponding historical operating conditions, including information such as power generation, equipment status, and operating efficiency; the aggregated data of historical power plant cluster energy storage status refers to the operating data of energy storage devices within the power plant cluster under the corresponding historical operating conditions, including the charging and discharging status of energy storage devices and remaining power; the aggregated data of historical grid load demand reflects the load demand of the grid under the corresponding historical operating conditions, including peak load, trough load, and load fluctuation trends. Through continuous monitoring and recording, these historical operating condition data sets form the typical operating modes of the power plant cluster under different times and conditions, providing rich historical experience data for the power plant cluster digital twin platform.

[0060] The preset operating condition data similarity standard is used for cluster analysis of historical operating condition data sets. The goal of similarity measurement is to assess the degree of similarity between two historical operating condition data sets in terms of generation unit, energy storage status, and grid load. Category aggregation classifies and aggregates data in the historical operating condition data set based on similarity. In this process, historical operating condition data with similar characteristics are identified and grouped into the same category. For example, if a certain category of historical operating conditions shows high similarity in generation unit output, energy storage charging and discharging patterns, and grid load demand fluctuations, then they will be grouped into the same category. This aggregation operation helps reduce data complexity and helps the system extract useful patterns and regularities from large-scale datasets. After similarity calculation and category aggregation, the historical operating condition data set is divided into multiple aggregated historical operating condition data sets. The data in each aggregated historical operating condition data set have similar characteristics and can represent a typical operating condition pattern.

[0061] For each aggregated historical operating condition data set, there is a corresponding aggregated historical power generation coordinated control strategy, that is, the corresponding coordinated control scheme at that time. The aggregated historical power generation coordinated control strategy is a control scheme formulated based on factors such as the grid load demand, power generation unit output, and energy storage status at that time, in order to optimize the coordination of power generation, energy storage and load.

[0062] For each aggregated historical operating condition data set corresponding to the aggregated historical power generation coordinated control strategy, an average value is calculated. The purpose of the average value calculation is to derive a comprehensive control strategy, i.e., the typical control strategy under that aggregated historical operating condition, by summarizing the control strategies under multiple similar historical operating conditions. Each aggregated historical operating condition data set yields a prototype historical power generation coordinated control strategy, which represents the most typical control strategy under that historical operating condition. These prototype strategies serve as templates and can be used to reference and optimize the control strategy under the current operating condition.

[0063] By traversing each aggregated historical operating condition data set, the mean of all historical operating condition data in it is calculated. Through the mean calculation, a typical representative of each operating condition set can be obtained, namely the historical operating condition prototype. These operating condition prototypes reflect the overall trend of the aggregated data set. The mean of each aggregated historical operating condition data set represents a historical operating condition prototype, and these historical operating condition prototypes can show different types of historical operating modes.

[0064] Each prototype's historical power generation coordinated control strategy is associated with its corresponding historical operating condition prototype. This association is achieved by matching the operating mode of each historical operating condition prototype with its corresponding power generation coordinated control strategy, ensuring that each historical operating condition prototype has a corresponding control strategy to support it. Through matching, a historical operating condition prototype library is constructed. This library contains multiple historical operating condition prototypes, each associated with a corresponding prototype's historical power generation coordinated control strategy. The historical operating condition prototype library provides the system with a decision support tool based on historical experience, enabling it to quickly find the most suitable control strategy under the current operating condition.

[0065] Furthermore, the initial power generation coordinated control strategy, the aggregated operational data of K power generation units, the aggregated data of the energy storage status of the power grid cluster, and the aggregated data of grid load demand are transmitted to the digital twin platform of the power grid cluster for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy, including:

[0066] The difference between the twin simulation results and the expected control target is determined; based on the magnitude of the difference, a correction scale is determined, and the initial power generation coordinated control strategy is corrected multiple times according to a preset correction method according to the preset correction scale to obtain a set of corrected power generation coordinated control strategies; the set of corrected power generation coordinated control strategies, together with the aggregated operation data of K power generation units, the aggregated data of the energy storage status of the power grid, and the aggregated data of the grid load demand, are transmitted to the power grid digital twin platform for twin simulation, and the set of correction difference is determined in conjunction with the expected control target; it is determined whether there is a correction difference less than or equal to the difference in the set of correction difference; if so, the corresponding corrected power generation coordinated control strategy is taken as the target power generation coordinated control strategy.

[0067] The expected control target is a pre-defined system performance objective. This can be a static objective, such as a fixed value for power output, or a dynamic objective, such as the changing trend of load demand. After the twin simulation is completed, the differences between the twin simulation results and the expected control target are compared. For example, the mean square error is used to calculate the average of the sum of squared errors between the twin simulation results and the expected control target; the smaller the error, the better the system performance. The absolute error is used to directly calculate the difference between the twin simulation results and the expected control target, reflecting the direct deviation between the two. A weighted difference degree is used to assign weights to multiple objectives and calculate the weighted difference degree. If the difference degree is large, it means that the strategy has not yet effectively met the expected control target and further adjustments are needed.

[0068] The correction scale refers to the magnitude of the adjustment strategy determined by the degree of difference. Specifically, the greater the difference, the larger the correction scale; the smaller the difference, the smaller the correction scale. The preset correction method is to randomly increase or decrease the output of different power generation units in the initial power generation coordinated control strategy according to the correction scale. The correction process is not completed in one step, but through multiple iterations for optimization. After each correction, a twin simulation is performed again to check whether the corrected strategy is closer to the expected control target. Each correction is based on the simulation results and correction scale of the previous step, gradually approaching the expected control target. Through this iterative process, the effectiveness of the control strategy can be continuously improved until the target requirements are met, resulting in a set of corrected power generation coordinated control strategies. These strategies represent strategies at different correction stages, and the effectiveness of the strategies will gradually improve as the correction progresses.

[0069] Each control strategy in the modified power generation coordinated control strategy set, along with the aggregated operation data of K power generation units, the aggregated energy storage status data of the power grid, and the aggregated grid load demand data, are transmitted to the power grid digital twin platform to simulate the operation of the power generation system under the current state. The goal of the simulation is to verify whether each modified power generation coordinated control strategy can meet the expected control target. After each simulation, the difference between the twin simulation results of each modified power generation coordinated control strategy and the expected control target is calculated. These differences constitute a new set as the modified difference set.

[0070] Check if any of the modified difference degrees in the set of modified difference degrees is less than or equal to the initial difference degree. If such a modified difference degree exists, it means that the modified power generation coordinated control strategy is closer to the expected control target and can be used as the final control strategy. Using it as the target power generation coordinated control strategy means that this strategy is the strategy that is finally applied to actual operation, which can meet the grid load demand, optimize the use of energy storage, and coordinate the output of each power generation unit.

[0071] Furthermore, the preset correction method is to randomly increase or decrease the output level of different power generation units in the initial power generation coordinated control strategy according to the correction scale.

[0072] The output level of different power generation units refers to the power output level of each power generation unit under the initial power generation coordinated control strategy. The output level of each power generation unit is affected by many factors, such as environmental conditions, equipment status, and grid load demand. The correction scale is obtained by calculating the difference degree, reflecting the magnitude of the difference degree and the required adjustment range. The larger the correction scale, the greater the adjustment required to the initial power generation coordinated control strategy in order to better match the expected control target.

[0073] To optimize the initial power generation coordination and control strategy, the output levels of different power generation units are randomly increased or decreased according to a correction scale. The purpose of this random adjustment is to explore different adjustment ranges to find the output configuration that best meets the objectives. Random increase refers to randomly increasing the output of certain power generation units to meet grid load demands or improve power generation efficiency; random decrease refers to randomly decreasing the output of certain power generation units to avoid over-generation or reduce equipment load. These adjustments are made within a preset correction scale range to ensure system stability and avoid excessive or insufficient adjustments.

[0074] Furthermore, if there is no modified difference degree less than or equal to the difference degree in the set of modified difference degrees, then the initial power generation coordinated control strategy will be used as the target power generation coordinated control strategy.

[0075] Check if any of the modified difference values ​​in the set of modified difference values ​​is less than or equal to the initial difference value. If no such modified difference value exists, it means that the current modification strategy has not significantly improved the degree of matching with the target. In this case, revert to the initial power generation coordinated control strategy, which is considered to be the most suitable target strategy. Although it may not fully meet the expected target, it is still used as the best strategy at present and is adopted as the target power generation coordinated control strategy.

[0076] In summary, the power generation coordination and control method for new energy power plant clusters provided in this application has the following technical effects:

[0077] By acquiring the set of K power generation units from K power plants within the target renewable energy cluster, a comprehensive understanding of the cluster's actual operation can be achieved, ensuring integrated monitoring of various power generation units and providing accurate input data for subsequent control strategies. Through time-series sensing aggregation, the operational data of the K power generation unit sets within a preset sensing window, along with the energy storage status and grid load demand of the target renewable energy cluster, can be effectively integrated to generate precise aggregated data. This not only provides a comprehensive view of the status of each power generation unit and the system, but also, through time-series data analysis within a preset sensing window, captures dynamic changes in power generation, energy storage, and grid load, improving the system's timeliness and responsiveness, and providing a data foundation for subsequent operating condition analysis and optimization of coordinated control strategies. Utilizing digital twin technology, the historical operating conditions of the target renewable energy cluster can be virtually reconstructed for efficient operating condition prototype identification. This is achieved by aggregating the operational data of the K power generation units and the energy storage status of the cluster. By aggregating data on aggregated data and grid load demand for prototype identification of operating conditions, the similarity between current and historical operating conditions can be identified. This allows for the rapid identification of the closest historical operating condition pattern when facing complex operating states, and the formulation of initial generation coordination and control strategies based on this pattern. This not only improves the rationality of the strategy but also accelerates the strategy generation process. In the digital twin platform of the power grid cluster, the initial generation coordination and control strategy is simulated based on the current system state. This simulation can verify the impact of the initial generation coordination and control strategy on grid load, generation unit operation, and energy storage status. The initial generation coordination and control strategy is iteratively revised based on the twin simulation results, and the target generation coordination and control strategy is finally determined. This process ensures that the target generation coordination and control strategy can adapt to the current grid demand and be flexibly adjusted according to real-time data, improving the automation and intelligence level of the control strategy and reducing risks and unnecessary costs during operation.

[0078] Example 2, based on the same inventive concept as the power generation coordination and control method for new energy power plant clusters in the foregoing examples, such as... Figure 2 As shown in the figure, this application embodiment provides a power generation coordinated control platform for new energy power plant clusters, the platform comprising:

[0079] The power generation unit acquisition module 10 is used to acquire the set of K power generation units of K stations in the target new energy power cluster.

[0080] The timing perception aggregation module 20 is used to traverse the operating data of the K power generation unit set, as well as the energy storage status and grid load demand of the target new energy field group within a preset perception window, to obtain the K power generation unit operation aggregation data, field group energy storage status aggregation data, and grid load demand aggregation data.

[0081] The operating condition prototype identification module 30 is used to identify the operating condition prototype of the K power generation units' operation data, the power storage status data, and the grid load demand data using the digital twin platform of the power generation cluster, and to determine the initial power generation coordinated control strategy based on the identification results.

[0082] The regulation strategy acquisition module 40 is used to transmit the initial power generation coordinated regulation strategy, the operation aggregation data of K power generation units, the aggregation data of the energy storage status of the power grid, and the aggregation data of the grid load demand to the digital twin platform of the power grid for twin simulation, and to modify the initial power generation coordinated regulation strategy according to the twin simulation results to determine the target power generation coordinated regulation strategy.

[0083] Furthermore, the time-aware convergence module 20 is used to perform the following operation steps:

[0084] The operation data of the K power generation units set within a preset sensing window is extracted and traversed to obtain a set of operation data sequences of the K power generation units; the operation data sequences of the K power generation units are time-series sensing and converged to determine the converged operation data of the K power generation units; the energy storage status and grid load demand of the target new energy field group are time-series sensing and converged within the preset sensing window to obtain the converged data of the field group's energy storage status and the converged data of the grid load demand.

[0085] Furthermore, the time-aware convergence module 20 is used to perform the following operation steps:

[0086] The system iterates through the set of operating data sequences of the K power generation units to perceive the trend of operating data and determine the set of trend features of operating data of the K power generation units. It then iterates through the set of trend features of operating data of the K power generation units to determine the iterative trend features of operating data of the K power generation units. Finally, it calculates the mean of operating data for each of the set of operating data sequences of the K power generation units to obtain the mean of operating data of the K power generation units. Using the iterative trend features of operating data of the K power generation units, it implicitly converges the corresponding mean of operating data of the K power generation units to determine the converged operating data of the K power generation units.

[0087] Furthermore, the time-aware convergence module 20 is used to perform the following operation steps:

[0088] Each of the K power generation unit operation data trend feature sets is paired with other features within each set to obtain a set of K power generation unit operation data trend feature combinations. Sub-similarity is calculated on these sets, and the results are normalized. The processed results are then filled into an initially empty matrix to construct K interactive iteration matrices. These K interactive iteration matrices are then convolved with the K power generation unit operation data trend feature combination sets to obtain a set of K power generation unit operation data enhanced trend feature combinations. The mean of these K power generation unit operation data enhanced trend feature combination sets is calculated to obtain the iterative trend features of the K power generation unit operation data.

[0089] Furthermore, the working condition prototype identification module 30 is used to perform the following operation steps:

[0090] A historical operating condition prototype library of the digital twin platform for the power plant cluster is obtained. Each historical operating condition prototype includes aggregated operational data of K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power grids. Using the aggregated operational data of the K power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power plant clusters as indexes, an approximate analysis is performed on each historical operating condition prototype in the historical operating condition prototype library, including the aggregated operational data of the K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power plant clusters. The historical operating condition prototype with the highest approximation is taken as the identification result. The prototype power generation coordinated control strategy corresponding to the identification result is taken as the initial power generation coordinated control strategy.

[0091] Furthermore, the working condition prototype identification module 30 is used to perform the following operation steps:

[0092] The process involves acquiring a set of historical operating condition data from the digital twin platform for the power generation cluster. Each set of historical operating condition data includes aggregated data on the operation of K historical power generation units, aggregated data on the energy storage status of the power generation cluster, and aggregated data on the grid load demand. Using a preset similarity score as a constraint, the historical operating condition data set is aggregated to determine multiple aggregated historical operating condition data sets. The average of multiple aggregated historical power generation collaborative control strategies from these multiple aggregated historical operating condition data sets is calculated to obtain multiple prototype historical power generation collaborative control strategies. The average of these multiple aggregated historical operating condition data sets is then calculated to determine multiple historical operating condition prototypes. Finally, the multiple prototype historical power generation collaborative control strategies are associated with the multiple historical operating condition prototypes to construct a historical operating condition prototype library.

[0093] Furthermore, the regulation strategy acquisition module 40 is used to perform the following operation steps:

[0094] The difference between the twin simulation results and the expected control target is determined; based on the magnitude of the difference, a correction scale is determined, and the initial power generation coordinated control strategy is corrected multiple times according to a preset correction method according to the preset correction scale to obtain a set of corrected power generation coordinated control strategies; the set of corrected power generation coordinated control strategies, together with the aggregated operation data of K power generation units, the aggregated data of the energy storage status of the power grid, and the aggregated data of the grid load demand, are transmitted to the power grid digital twin platform for twin simulation, and the set of correction difference is determined in conjunction with the expected control target; it is determined whether there is a correction difference less than or equal to the difference in the set of correction difference; if so, the corresponding corrected power generation coordinated control strategy is taken as the target power generation coordinated control strategy.

[0095] Furthermore, the preset correction method is to randomly increase or decrease the output level of different power generation units in the initial power generation coordinated control strategy according to the correction scale.

[0096] Furthermore, if there is no modified difference degree less than or equal to the difference degree in the set of modified difference degrees, then the initial power generation coordinated control strategy will be used as the target power generation coordinated control strategy.

[0097] Through the foregoing detailed description of a power generation coordination and control method for a new energy power plant cluster, those skilled in the art can clearly understand the power generation coordination and control platform for a new energy power plant cluster in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to in the method section.

[0098] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for coordinated power generation control in new energy power plant clusters, characterized in that, The method includes: Obtain the set of K power generation units from K power plants in the target renewable energy cluster; Within a preset sensing window, the operation data of the K power generation unit set, as well as the energy storage status and grid load demand of the target new energy field group, are traversed and time-series sensing data are aggregated to obtain the aggregated operation data of the K power generation units, the aggregated energy storage status data of the field group, and the aggregated grid load demand data. Using the digital twin platform of the power generation cluster, the operating condition prototypes of the K power generation units, the power storage status, and the grid load demand are identified, and the initial power generation coordinated control strategy is determined based on the identification results. The initial power generation coordinated control strategy, the aggregated operation data of K power generation units, the aggregated data of the energy storage status of the power grid cluster, and the aggregated data of the grid load demand are transmitted to the digital twin platform of the power grid cluster for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy. Specifically, the operating condition prototype is identified using the aggregated data of the K power generation units' operation, the aggregated data of the power storage status, and the aggregated data of the grid load demand, based on the identification results. The initial power generation coordinated control strategy is then determined, including: Obtain the historical operating condition prototype library of the digital twin platform for the power plant cluster, wherein each historical operating condition prototype includes aggregated data of K prototype power generation units, aggregated data of prototype power plant cluster energy storage status, and aggregated data of prototype grid load demand. Using the aggregated data of the operation of the K power generation units, the aggregated data of the energy storage status of the power plant cluster, and the aggregated data of the power grid load demand as indexes, an approximate analysis is performed on each historical operating condition prototype in the historical operating condition prototype library, including the aggregated data of the operation of the K prototype power generation units, the aggregated data of the prototype energy storage status of the prototype power plant cluster, and the aggregated data of the prototype power grid load demand. The historical operating condition prototype with the highest approximation is taken as the identification result. The prototype power generation coordinated control strategy corresponding to the identification results will be used as the initial power generation coordinated control strategy.

2. The method for coordinated power generation control of new energy power plants as described in claim 1, characterized in that, Within a preset sensing window, the operational data of the K power generation units are traversed, along with the energy storage status and grid load demand of the target new energy cluster, for time-series sensing aggregation. This yields aggregated operational data of the K power generation units, aggregated energy storage status data of the cluster, and aggregated grid load demand data, including: The operating data of the K power generation units set within a preset sensing window is extracted by traversing the data to obtain a set of K power generation unit operating data sequences; Time-series sensing aggregation is performed on the set of operational data sequences of the K power generation units to determine the aggregated operational data of the K power generation units; The energy storage status and grid load demand of the target new energy power plant cluster are time-series sensing and aggregation within a preset sensing window to obtain aggregated data on the energy storage status of the power plant cluster and aggregated data on the grid load demand.

3. The method for coordinated power generation control of new energy power plants as described in claim 2, characterized in that, Time-series sensing aggregation is performed on the set of operational data sequences of the K power generation units to determine the aggregated operational data of the K power generation units, including: The set of operating data sequences of the K power generation units is traversed to perceive the operating data trends and determine the set of operating data trend features of the K power generation units. The trend feature sets of the operating data of the K power generation units are interactively iterated to determine the iterative trend features of the operating data of the K power generation units; Calculate the mean value of the operating data of each of the K power generation unit operating data sequence sets to obtain the mean value of the operating data of the K power generation units; By utilizing the iterative trend characteristics of the operating data of the K power generation units, implicit state aggregation is performed on the mean of the corresponding operating data of the K power generation units to determine the aggregated operating data of the K power generation units.

4. The power generation coordinated control method for new energy power plant clusters as described in claim 3, characterized in that, The operational data trend feature sets of the K power generation units are iteratively evaluated to determine the iterative trend features of the operational data of the K power generation units, including: Each of the K power generation unit operation data trend feature sets is combined pairwise within the set to obtain a set of K power generation unit operation data trend feature combinations. Sub-similarity calculation is performed on the combination set of trend features of the operating data of the K power generation units, and the calculation results are normalized. The processed results are then filled into the initially empty matrix to construct K interactive iteration matrices. The K interactive iteration matrices are used to convolve the set of trend features of the K power generation unit operation data to obtain the set of enhanced trend features of the K power generation unit operation data. Calculate the mean of the set of enhanced trend features of the K power generation unit operation data to obtain the iterative trend features of the K power generation unit operation data.

5. The method for coordinated power generation control of new energy power plants as described in claim 1, characterized in that, Obtain the historical operating condition prototype library of the aforementioned power plant cluster digital twin platform. Each historical operating condition prototype includes aggregated operational data of K prototype power generation units, aggregated energy storage status data of prototype power plant clusters, and aggregated grid load demand data of prototype power grids, including: Acquire the historical operating condition data set of the digital twin platform of the power generation cluster, wherein each historical operating condition data set includes aggregated data of K historical power generation units, aggregated data of historical power generation cluster energy storage status, and aggregated data of historical grid load demand. Using a preset working condition data similarity as a constraint, the historical working condition data set is aggregated into similar categories to determine multiple aggregated historical working condition data sets. The average value of multiple aggregated historical power generation coordinated control strategies obtained from the multiple aggregated historical operating condition data sets is calculated to obtain multiple prototype historical power generation coordinated control strategies. The average value is calculated by traversing the multiple aggregated historical working condition data sets to determine multiple historical working condition prototypes; The multiple prototype historical power generation coordinated control strategies are associated with the multiple historical operating condition prototypes to construct a historical operating condition prototype library.

6. The method for coordinated power generation control of new energy power plants as described in claim 1, characterized in that, The initial power generation coordinated control strategy, the aggregated operational data of K power generation units, the aggregated data of the energy storage status of the power grid cluster, and the aggregated data of grid load demand are transmitted to the digital twin platform of the power grid cluster for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy, including: Determine the degree of difference between the twin simulation results and the expected regulatory target; Based on the magnitude of the difference, a correction scale is determined, and the initial power generation coordinated control strategy is corrected multiple times according to a preset correction method according to the preset correction scale to obtain a set of corrected power generation coordinated control strategies. The modified set of power generation coordinated control strategies, the aggregated operation data of K power generation units, the aggregated data of the energy storage status of the power grid cluster, and the aggregated data of the grid load demand are respectively transmitted to the digital twin platform of the power grid cluster for twin simulation, and the modified difference set is determined in combination with the expected control target. Determine whether there exists a correction difference degree less than or equal to the correction difference degree in the set of correction difference degrees. If so, then take the corresponding correction power generation coordinated control strategy as the target power generation coordinated control strategy.

7. The method for coordinated power generation control of new energy power plants as described in claim 6, characterized in that, The preset correction method is to randomly increase or decrease the output level of different power generation units in the initial power generation coordinated control strategy according to the correction scale.

8. The method for coordinated power generation control of new energy power plants as described in claim 6, characterized in that, If there is no modified difference degree less than or equal to the difference degree in the set of modified difference degrees, then the initial power generation coordinated control strategy will be used as the target power generation coordinated control strategy.

9. A power generation coordination and control platform for new energy power plant clusters, characterized in that, The platform is used to implement the power generation coordination and control method for a new energy power plant cluster as described in any one of claims 1-8, the platform comprising: The power generation unit acquisition module is used to acquire the set of K power generation units from K stations in the target new energy power cluster; The time-series perception aggregation module is used to traverse the operation data of the K power generation unit set, as well as the energy storage status and grid load demand of the target new energy field group within a preset perception window, and to obtain the operation aggregation data of the K power generation units, the energy storage status aggregation data of the field group, and the grid load demand aggregation data. The operating condition prototype identification module is used to identify the operating condition prototype of the K power generation units, the energy storage status of the power generation group, and the grid load demand data using the digital twin platform of the power generation group, and to determine the initial power generation coordinated control strategy based on the identification results. The control strategy acquisition module is used to transmit the initial power generation coordinated control strategy, the aggregated operation data of K power generation units, the aggregated energy storage status data of the power grid, and the aggregated grid load demand data to the power grid digital twin platform for twin simulation. Based on the twin simulation results, the initial power generation coordinated control strategy is modified to determine the target power generation coordinated control strategy.

Images