Energy Storage Grid-Forming Control Method, Device, Equipment, and Storage Medium

Abstract

The present invention relates to an energy storage grid-forming control method, device, equipment, and storage medium, including the following steps: real-time data collection of the target power grid is performed through an intelligent sensor to obtain power grid collection data; the power grid collection data is predicted through a preset big data analysis technique to obtain the power grid operation trend; the energy state of the target energy storage system is evaluated to obtain the available capacity of the target energy storage system; a scheduling control plan is formulated based on the available capacity and the power grid operation trend; wherein, the scheduling control plan is used to schedule and control the target power grid or the target energy storage system to maintain the stability of the target power grid frequency, solving the technical problem of how to effectively integrate the energy storage system into the existing power network and ensure that it can stably support the power grid frequency and other key performance indicators.

Description

Technical Field

[0001] The present invention relates to the technical field of power grids, and particularly relates to a control method, device, equipment, and storage medium for an energy storage grid-forming system. Background Art

[0002] With the continuous growth of the global demand for clean energy and the gradual depletion of traditional fossil fuel resources, the power system is undergoing a transformation from centralized power generation to distributed energy resources (DERs). In this process, energy storage systems, as key components for balancing power supply and demand and improving the flexibility of the power grid, are becoming increasingly important. However, the current operation mode of the power system is mainly based on relatively stable traditional power sources, and has limited adaptability to rapidly changing renewable energy inputs and volatile load demands. Therefore, how to effectively integrate energy storage systems into the existing power network and ensure their stable support for grid frequency and other key performance indicators has become an urgent technical problem to be solved.

[0003] One of the core issues in the research background lies in real-time performance and prediction accuracy. To achieve efficient energy storage scheduling, precise data acquisition and analysis technologies must be relied on to capture changes in the grid state. Traditional data processing methods are difficult to meet the requirements of modern smart grids for real-time decision-making. Although emerging big data analysis and artificial intelligence algorithms provide new solutions, their applications still face challenges such as high computational complexity and insufficient model training data. In addition, the assessment of the energy state of energy storage systems is also a difficult point, which not only affects the safety and lifespan of energy storage devices themselves, but also directly relates to the formulation of reasonable scheduling plans. There is still much room for improvement in the existing technology for accurately assessing the available capacity of energy storage systems, especially when facing multiple different types of energy storage devices.

[0004] Another important consideration is the effectiveness and adaptability of the energy storage grid-forming control strategy. As more and more energy storage systems are connected to the power grid, how to design a control method that can ensure grid stability while maximizing the benefits of energy storage systems has become the focus of research. The current control strategies are mostly based on fixed rules or simple feedback mechanisms, and are unable to cope with complex grid operation trends. For example, in the case of a sudden increase or decrease in renewable energy generation, these strategies may not be able to respond in a timely manner, resulting in the grid frequency deviation exceeding the allowable range. Therefore, it is necessary to develop a control method for energy storage grid-forming that can dynamically adjust and has strong adaptability to cope with the more flexible and complex future power market environment and grid operation conditions. This control method should not only consider short-term power balance, but also take into account long-term economic and environmental friendliness, providing technical support for building a sustainable smart grid. Summary of the Invention

[0005] The main objective of the present invention is to provide an energy storage grid-forming control method, device, equipment, and storage medium, which solve the technical problem of how to effectively integrate an energy storage system into an existing power grid and ensure its stable support for grid frequency and other key performance indicators.

[0006] To achieve the above objective, the present invention provides an energy storage grid-forming control method, including the following steps:

[0007] Real-time data of the target power grid is collected through intelligent sensors to obtain grid acquisition data;

[0008] The grid acquisition data is predicted through a preset big data analysis technology to obtain the grid operation trend;

[0009] The energy state of the target energy storage system is evaluated to obtain the available capacity of the target energy storage system;

[0010] A dispatching control plan is formulated based on the available capacity and the grid operation trend; wherein, the dispatching control plan is used to dispatch and control the target power grid or the target energy storage system to maintain the stability of the target power grid frequency.

[0011] Further, the intelligent sensor is a distributed optical fiber sensing array, and the real-time data of the target power grid is collected through the intelligent sensor to obtain grid acquisition data, including:

[0012] Key nodes of the target power grid are sampled through the distributed optical fiber sensing array to obtain original electrical parameter data;

[0013] Wavelet packet transform analysis is performed on the original electrical parameter data to obtain time-frequency electrical feature data;

[0014] High-order singular value decomposition is performed on the time-frequency electrical feature data by using a preset tensor decomposition algorithm to obtain a key feature tensor; wherein, the key feature tensor includes time-frequency features of power consumption fluctuations, time-frequency distribution of electrical loads, and time-frequency features of power spectra;

[0015] The key feature tensor is dimensionally reduced to obtain grid acquisition data.

[0016] Further, the grid acquisition data is predicted through a preset big data analysis technology to obtain the grid operation trend, including:

[0017] Data preprocessing is performed on the grid acquisition data to obtain preprocessed electrical data;

[0018] Feature extraction is performed on the preprocessed electrical data to obtain electrical extraction features; wherein, the electrical extraction features;

[0019] Perform time-series data decomposition on the electrical extraction features to obtain multi-scale time-series electrical features; wherein, the multi-scale time-series electrical features include load change features, voltage time-domain features, current time-domain features, power quality features, and power factor dynamic characteristics;

[0020] Perform causal relationship analysis on the multi-scale time-series electrical features through the transfer entropy algorithm to obtain a feature correlation network; wherein, the feature correlation network is used to reveal the dynamic operation correlations between the multi-scale time-series electrical features in the target power grid;

[0021] Use the adaptive kernel density estimation method to estimate the state transition probability of the target power grid based on the feature correlation network to obtain a power grid state transition probability table; wherein, the power grid state transition probability table is used to quantify the conversion probability between different operating conditions of the target power grid, so as to perform quantitative dynamic probability prediction on the future operating state of the power grid;

[0022] Perform Markov chain prediction on the target power grid based on the power grid state transition probability table to obtain a power grid state prediction sequence;

[0023] Adopt the self-organized criticality analysis method to perform non-linear dynamic analysis on the power grid state prediction sequence to obtain the power grid operation trend; wherein, the power grid operation trend includes voltage stability, frequency stability, load change trend, and change trend of power quality.

[0024] Further, the energy state assessment of the target energy storage system to obtain the available capacity of the target energy storage system includes:

[0025] Perform real-time current monitoring on the target energy storage system through an intelligent sensor to obtain real-time current data;

[0026] Use the segmented Coulomb meter method to calculate the cumulative charge amount of the target energy storage system based on the real-time current data to obtain a cumulative charge amount sequence;

[0027] Perform multi-scale decomposition on the cumulative charge amount sequence through a multi-resolution analysis method to obtain multi-scale cumulative charge amount features;

[0028] Use the adaptive filtering algorithm to dynamically adjust the multi-scale cumulative charge amount features to obtain an estimated value of the cumulative charge amount;

[0029] Based on the estimated value of the cumulative charge amount, analyze the relationship between the storage voltage and the stored charge of the target energy storage system to obtain a storage voltage-stored charge characteristic curve;

[0030] Obtain the rated capacity of the target energy storage system and the performance record of the target energy storage system from the database, and perform dynamic evaluation and back-calculation of the available capacity of the target energy storage system through the storage voltage-storage charge characteristic curve, rated capacity, and performance record to obtain the available capacity of the target energy storage system.

[0031] Further, based on performing dynamic evaluation and back-calculation of the available capacity of the target energy storage system through the storage voltage-storage charge characteristic curve, rated capacity, and performance record to obtain the available capacity of the target energy storage system, it includes:

[0032] Use the piecewise exponential fitting algorithm to perform piecewise fitting processing on the storage voltage-storage charge characteristic curve to obtain the charge and discharge capacity loss characteristic parameters;

[0033] Based on the charge and discharge capacity loss characteristic parameters, perform weighted statistics of the cycle times of the target energy storage system to obtain the capacity attenuation coefficient;

[0034] Based on the capacity attenuation coefficient and the performance record, perform collaborative analysis on the target energy storage system to obtain the health status of the target energy storage system;

[0035] Based on the health status and the rated capacity, perform dynamic capacity evaluation on the current energy storage system to obtain the theoretical available capacity value;

[0036] Obtain the current temperature parameter through a preset temperature sensor, and perform capacity value compensation on the theoretical available capacity value based on the current temperature parameter to obtain the temperature-corrected capacity value;

[0037] Based on the temperature-corrected capacity value, the charge and discharge capacity loss characteristic parameters, the capacity attenuation coefficient, and the health status, perform back-calculation of the available capacity of the target energy storage system to obtain the available capacity of the target energy storage system.

[0038] Further, based on the charge and discharge capacity loss characteristic parameters, perform weighted statistics of the cycle times of the target energy storage system to obtain the capacity attenuation coefficient, including:

[0039] Perform time series analysis on the charge and discharge capacity loss characteristic parameters to obtain the charge and discharge cycle sequence;

[0040] Based on the charge and discharge cycle sequence, count the number of charging and discharging times of each cycle of the target energy storage system to obtain the cycle times dataset;

[0041] Use the weighted moving average method to perform weighted average calculation on the cycle times dataset to obtain the weighted cycle times;

[0042] Based on the weighted cycle times, perform a fitting analysis on the capacity loss of the energy storage system to obtain a capacity decay curve;

[0043] Perform a slope analysis on the capacity decay curve to obtain a capacity decay rate;

[0044] Based on the capacity decay rate and the weighted cycle times, calculate a capacity decay coefficient for the target energy storage system to obtain a capacity decay coefficient.

[0045] Further, the formulating a scheduling control plan based on the available capacity and the grid operation trend includes:

[0046] Perform a state correlation modeling on the available capacity and the grid operation trend through a dynamic programming algorithm to obtain available capacity-grid operation trend correlation data; wherein, the available capacity-grid operation trend correlation data includes the energy storage capacity state and the load power deviation of the target grid;

[0047] Use a preset fuzzy neural network to perform a quantitative evaluation on the available capacity-grid operation trend correlation data to obtain a state evaluation result; wherein, the state evaluation result includes a capacity adequacy parameter and a grid stability parameter;

[0048] Monitor the state evaluation result. If any of the capacity adequacy parameter, grid stability parameter, and load balance parameter in the state evaluation result is not within a preset range, perform a multi-objective weight allocation on the state evaluation result through a hybrid particle swarm optimization algorithm to obtain a scheduling priority sequence;

[0049] Convert the scheduling priority sequence into a preliminary scheduling instruction sequence;

[0050] Perform a timing coordination on the preliminary scheduling instruction sequence through a distributed cooperative optimization algorithm to obtain a scheduling control plan.

[0051] The present invention also provides an energy storage power grid forming control device, including:

[0052] An acquisition module, configured to perform real-time data acquisition on a target grid through an intelligent sensor to obtain grid acquisition data;

[0053] A prediction module, configured to predict the grid acquisition data through a preset big data analysis technique to obtain a grid operation trend;

[0054] An evaluation module, configured to perform an energy state evaluation on a target energy storage system to obtain the available capacity of the target energy storage system;

[0055] A control module formulates a dispatching control plan based on the available capacity and the power grid operation trend; wherein, the dispatching control plan is used to dispatch and control the target power grid or the target energy storage system to maintain the frequency stability of the target power grid.

[0056] The present invention also provides a computer device, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0057] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in any one of the above are implemented.

[0058] The energy storage grid-forming control method provided by the present invention includes the following steps: real-time data collection of the target power grid is performed through an intelligent sensor to obtain power grid collection data; the power grid collection data is predicted through a preset big data analysis technology to obtain the power grid operation trend; the energy state of the target energy storage system is evaluated to obtain the available capacity of the target energy storage system; a dispatching control plan is formulated based on the available capacity and the power grid operation trend; wherein, the dispatching control plan is used to dispatch and control the target power grid or the target energy storage system to maintain the frequency stability of the target power grid. Through the above technical means, the technical problem of how to effectively integrate the energy storage system into the existing power network and ensure that it can stably support the power grid frequency and other key performance indicators is solved. The implementation realizes that the preset big data analysis technology is used to process the power grid collection data, and can provide a more accurate and reliable prediction of the power grid operation trend. Compared with the traditional method, this method can significantly improve the prediction accuracy and provide a more solid data basis for decision-making. At the same time, the real-time data collection and fast data analysis capabilities also enhance the response speed of the system to emergencies, enabling the energy storage control system to make the best response in the shortest time. Description of the Drawings

[0059] Figure 1 is a schematic diagram of the steps of the energy storage grid-forming control method in an embodiment of the present invention;

[0060] Figure 2 is a structural block diagram of the energy storage grid-forming control device in an embodiment of the present invention;

[0061] Figure 3 is a schematic structural block diagram of the computer device in an embodiment of the present invention.

[0062] The realization, functional characteristics and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiments

[0063] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] As Figure 1 shown, Figure 1 is a schematic diagram of the steps of a grid-forming control method for energy storage in an embodiment of the present invention;

[0065] An embodiment of the present invention provides a grid-forming control method for energy storage, including the following steps:

[0066] Step S1, real-time data of the target power grid is collected through an intelligent sensor to obtain power grid acquisition data.

[0067] Specifically, the step of collecting real-time data from the target power grid through intelligent sensors to obtain power grid acquisition data is the basis of the entire energy storage grid-forming control method. In this process, intelligent sensors, as key tools for data collection, are deployed at various key nodes of the power system, which can include substations, transmission lines, distribution facilities, and user access points, etc. Intelligent sensors can monitor and record various operating parameters of the power grid in real time, such as voltage, current, power factor, frequency, etc., ensuring that the obtained data has high precision and high reliability. For example, in an urban power grid containing a large number of distributed energy resources (DERs), intelligent sensors can help capture fluctuations in the power generation of solar panels caused by weather changes, or sudden changes in the output power of wind turbines. Intelligent sensors are not limited to the measurement of physical parameters; they can also be integrated with other devices and technologies to achieve more complex functions. For instance, intelligent sensors can be combined with communication modules to instantly transmit the collected data to the data analysis center or cloud platform. This instantaneity is crucial for timely adjusting the working state of the energy storage system because any delay may result in missing the best response opportunity. In addition, to ensure the security and integrity of the data, intelligent sensors usually come with encryption technology and error detection mechanisms to prevent the data from being tampered with or lost during transmission. When the intelligent sensors complete the data collection of the target power grid, these data will form the so-called power grid acquisition data, which is the basis for all subsequent analysis and decision-making. In modern power systems, with more and more renewable energy integrated into the grid, the operating conditions of the power grid have become increasingly complex, and traditional static data can no longer meet the requirements. Therefore, the dynamic and continuous data stream provided by intelligent sensors has become an indispensable part. By deeply mining and analyzing these data, the future operating trends of the power grid can be predicted, the energy state of the energy storage system can be evaluated, and finally an effective scheduling control plan can be formulated. For example, in a regional power grid with a high proportion of wind power generation, the accurate wind speed and direction data provided by intelligent sensors can help predict the wind power output in the next period of time, so as to plan the charge and discharge strategies of the energy storage system in advance and ensure the stable power supply of the power grid even in the case of insufficient wind power. In short, collecting real-time data from the target power grid through intelligent sensors not only provides the necessary information support for the energy storage grid-forming control method, but also largely determines the effect and efficiency of the entire control system. The application of intelligent sensors enables us to still maintain precise control over the power grid state in a more complex and changeable power environment, and thus achieve more intelligent and efficient power management.

[0068] Step S2, predicting the power grid operation trend by preset big data analysis technology for the power grid acquisition data.

[0069] Specifically, the step of predicting the power grid collected data by using the preset big data analysis technology to obtain the power grid operation trend is a key link in the energy storage network control method. This process is immediately after the smart sensor completes the real-time data collection of the target power grid, and aims to use advanced data analysis methods to process and analyze the massive power grid collected data provided by the smart sensor. With the increasing complexity of the power system, especially when a large number of distributed energy resources (DERs) such as wind power and solar energy are connected to the power grid, traditional analysis methods have been unable to cope with such a large and dynamically changing data set. Therefore, the introduction of preset big data analysis technology has become an inevitable choice. The preset big data analysis technology covers a variety of algorithms and models, which are pre-selected and configured to efficiently process the data stream from the smart sensor. For example, in an urban power grid containing multiple wind farms, big data analysis technology can integrate time series analysis, machine learning algorithms, and deep learning models to mine potential patterns and correlations in the data. Specifically, time series analysis can help identify the trend of wind power output changing over time and seasons; machine learning algorithms can train models based on historical data to predict wind power output in a specific time period in the future; and deep learning models can more deeply understand complex nonlinear relationships and improve prediction accuracy. Through such comprehensive applications, the original power grid collection data can be effectively converted into forward-looking power grid operation trend predictions. In addition, the preset big data analysis technology is not limited to processing a single type of input data, it can also integrate multi-source information, including weather forecasts, user electricity behavior, market electricity price fluctuations and other external factors. For example, when predicting the load demand of a city's power grid, in addition to considering the production capacity of local power generation facilities, weather forecast data can also be combined to evaluate the changes in the frequency of air conditioning use, or refer to the impact of major historical events on electricity consumption. This multi-dimensional data fusion makes the prediction results more comprehensive and accurate, and provides a solid foundation for the subsequent formulation of reasonable dispatching and control plans. In order to ensure the reliability of the prediction results, the preset big data analysis technology will continue to optimize its own model parameters and self-correct through a feedback mechanism. This means that with the influx of more new data and the advancement of technology, the prediction model will gradually become more accurate. For example, in an actual application scenario, if a regional power grid experiences several unforeseen peak load conditions, the big data analysis system will automatically adjust its internal algorithm to learn the laws behind these anomalies, thereby making more accurate predictions in the future. Such adaptive capabilities are essential to maintaining grid stability, especially when facing rapidly changing renewable energy inputs, so that it can prepare in advance and avoid frequency fluctuations caused by imbalances in supply and demand. In summary, predicting the data collected by the power grid through the preset big data analysis technology not only improves the understanding and grasp of the power grid operation trend, but also provides a scientific basis for the effective management and scheduling of energy storage systems.The application of this method enables the power system to maintain a high degree of flexibility and response speed in a complex and changing environment, ultimately achieving more intelligent and efficient energy management.

[0070] Step S3: Evaluate the energy state of the target energy storage system to obtain the available capacity of the target energy storage system.

[0071] Specifically, the process of evaluating the energy state of the target energy storage system to obtain the available capacity of the target energy storage system is a key step in ensuring the effective implementation of the energy storage grid-forming control method. This process follows immediately after collecting grid data through intelligent sensors and predicting the grid operation trend using preset big data analysis techniques, aiming to accurately understand the current charge and discharge state of the energy storage system and its actual energy support capacity. To achieve this, a series of advanced monitoring and evaluation techniques are required, which work together to provide a comprehensive and accurate energy state report. In specific operations, the energy state evaluation first involves monitoring the internal parameters of the energy storage system, including but not limited to battery voltage, current, temperature, etc. These parameters not only reflect the immediate working conditions of the energy storage device but also provide important clues about its health status. For example, in an urban power grid containing multiple lithium battery energy storage devices, by real-time monitoring the voltage levels of each battery cell, it can be determined which batteries are close to full charge and which may be in a low-power or even faulty state. In addition, temperature information is equally important for evaluating battery life and safety. Excessive temperature may lead to performance degradation or safety hazards, so it must be taken into consideration. Further, to understand the overall energy state of the energy storage system from a more macroscopic perspective, it is also necessary to conduct a comprehensive analysis by combining historical operation data and current grid demands. This means not only considering the technical characteristics of the energy storage system itself but also examining it in the context of the entire power system. For example, in a regional power grid with a high proportion of wind power generation, when it is predicted that the upcoming weakening of the wind speed will lead to a reduction in wind power output, the energy storage system plays a role in supplementing energy. At this time, the energy state evaluation will consider the charge and discharge records of the energy storage system in the past few hours, combined with the expected load demand in the next few hours, calculate the total available energy for scheduling, and determine the optimal charge and discharge strategy to maintain grid frequency stability. In addition to direct physical parameters and technical indicators, economic factors can also be introduced into the energy state evaluation as an auxiliary decision-making basis. For example, considering the price fluctuations in the electricity market during different time periods, the energy storage system can preferentially charge when the electricity price is low and release the stored energy to earn price differences when the electricity price is high. Such a strategy can not only improve the economic benefits of the energy storage system but also enhance its support for the power grid. At the same time, this flexibility also enables the energy storage system to respond quickly to emergencies, such as sudden load peaks or insufficient renewable energy generation. Finally, through a comprehensive energy state evaluation of the target energy storage system, we can obtain a result that accurately reflects its available capacity. This is not only the basis for formulating a reasonable scheduling control plan but also provides a scientific basis for optimizing the use efficiency of energy storage resources. For example, in the aforementioned application scenario, if the evaluation result shows that the available capacity of a certain energy storage facility is insufficient to cope with the upcoming electricity peak, then other energy storage systems or traditional power generation sources can be arranged in advance to ensure the continuity and stability of power supply.In summary, this evaluation process is crucial for achieving the efficient management of energy storage systems and the stable operation of the power grid.

[0072] Step S4: Develop a scheduling control plan based on the available capacity and the power grid operation trend; wherein, the scheduling control plan is used to schedule and control the target power grid or the target energy storage system to maintain the stability of the target power grid frequency.

[0073] Specifically, the process of formulating a dispatching control plan based on the available capacity and the power grid operation trend is a core component of the energy storage grid-forming control method. This process is based on the previous energy state assessment of the target energy storage system and the prediction of the power grid operation trend through preset big data analysis techniques. The aim is to formulate a set of dispatching control plans that can effectively maintain the frequency stability of the target power grid by comprehensively considering the current capabilities of the energy storage system and the future demand changes of the power grid. First of all, in order to formulate such a dispatching control plan, it is necessary to combine the available capacity information obtained from the energy state assessment of the energy storage system with the power grid operation trend predicted by big data analysis techniques. This means deeply analyzing the maximum energy output or absorption capacity that the energy storage system can provide in different time periods and matching it with the expected power supply and demand situation. For example, in an urban power grid with multiple wind farms, when it is predicted that the wind power output will decrease due to a drop in wind speed in the next few hours, the dispatching control system needs to plan in advance how to use the energy storage system to make up for the possible energy gap. At this time, the available capacity data of the energy storage system provides key information input to help determine which energy storage devices can be put into use at what time to ensure that the power grid frequency stability is not affected. Further, formulating a dispatching control plan is not just a simple arrangement of charge and discharge instructions. It also involves the application of complex decision-making algorithms and optimization models. These tools can simulate different dispatching scenarios, evaluate the effects of various solutions, and select the optimal solution. For example, the dispatching control system may calculate the best strategy that can meet the immediate power demand and maximize economic benefits by combining factors such as real-time electricity prices, user load demands, and weather forecasts. At the same time, considering the safety and reliability requirements of the power grid, the plan will also set a series of constraints, such as avoiding overuse of energy storage devices and maintaining a certain reserve capacity, to prevent the risk of power supply interruption due to unexpected situations. In addition, to ensure the dynamic adaptability of the dispatching control plan, the system also needs to have the ability to adjust in real time. This is because the operating environment of the power system is highly dynamic and not completely predictable, and even the most accurate prediction model may have errors. Therefore, once the actual operating situation deviates from the expectation, such as a sudden unexpected peak load appears in a certain area or a certain energy storage facility fails, the dispatching control system should be able to respond quickly and re-optimize the dispatching plan. For example, in the application scenario of the aforementioned urban power grid, if the actual electricity consumption in a certain period far exceeds the expectation, the dispatching control system will immediately activate additional energy storage resources and even coordinate power support from neighboring regions to quickly restore the power grid frequency to the normal range. Finally, through the dispatching control plan formulated based on the available capacity and the power grid operation trend, not only can the efficient dispatching and precise control of the target power grid or energy storage system be achieved, but also a strong guarantee for the stable operation of the entire power system can be provided.The implementation of such a plan enables the energy storage system to function when it is most needed. Whether it is to address the intermittency of renewable energy generation or handle sudden load fluctuations, it can effectively maintain the stability of the grid frequency. At the same time, this also promotes the flexibility and economy of the power market, laying a solid foundation for building a more intelligent and sustainable modern power grid system.

[0074] In a specific embodiment, the intelligent sensor is a distributed optical fiber sensing array. The real-time data collection of the target power grid through the intelligent sensor to obtain power grid collection data includes:

[0075] Sampling key nodes of the target power grid through the distributed optical fiber sensing array to obtain original electrical parameter data;

[0076] Performing wavelet packet transform analysis on the original electrical parameter data to obtain time-frequency electrical feature data;

[0077] Using a preset tensor decomposition algorithm to perform high-order singular value decomposition on the time-frequency electrical feature data to obtain a key feature tensor; wherein, the key feature tensor includes the time-frequency feature of electricity consumption fluctuation, the time-frequency distribution of electricity load, and the time-frequency feature of power spectrum.

[0078] Performing dimensionality reduction processing on the key feature tensor to obtain power grid collection data.

[0079] Specifically, the intelligent sensor is a distributed optical fiber sensing array. By using this specific type of intelligent sensor to collect real-time data from the target power grid, the process of obtaining power grid acquisition data is the crucial first step in the entire energy storage grid-forming control method. This process not only involves the use of advanced sensing technologies to acquire raw electrical parameter data but also includes complex data processing steps such as wavelet packet transform analysis, the application of tensor decomposition algorithms, and dimensionality reduction processing to ultimately obtain the power grid acquisition data for subsequent analysis and decision-making. First, as a highly sensitive and widely distributed monitoring tool, the distributed optical fiber sensing array is deployed at key nodes of the target power grid. These key nodes can include substations, transmission lines, distribution facilities, and user access points, etc. The distributed optical fiber sensing array can reflect the changes in the surrounding environment by accurately measuring the propagation characteristics of optical signals, thereby realizing the monitoring of the internal state of the power system. For example, in an urban power grid containing multiple wind farms, the distributed optical fiber sensing array can help capture the fluctuations in the power generation of solar panels caused by weather changes or sudden changes in the output power of wind turbines. Through such a sampling method, high-resolution raw electrical parameter data covering a wide area can be obtained, providing a solid foundation for subsequent in-depth analysis. Next, in order to extract valuable information from a large amount of raw electrical parameter data, wavelet packet transform analysis needs to be performed on it. Wavelet packet transform is a multi-resolution analysis tool that can decompose the signal in the time domain into components in different frequency bands, thereby revealing the time-frequency characteristics of the signal. For the power grid, this means that it can better understand the fluctuations in electricity consumption, load distribution, and the variation laws of power spectra over time and frequency. For example, in the aforementioned application scenario of the urban power grid, when the wind speed suddenly increases or decreases, wavelet packet transform analysis can help identify which time periods have significant changes in wind power output and how these changes affect the overall operation state of the power grid. In addition, it can also detect changes in the load patterns on the user side, such as the differences between the peak hours during the day and the low valley hours at night in the commercial area, or the different electricity consumption behaviors on weekends and weekdays in the residential area. Then, the preset tensor decomposition algorithm is used to perform high-order singular value decomposition on the time-frequency electrical feature data obtained after wavelet packet transform to obtain the key feature tensor. The key feature tensor here contains information in three main aspects: the time-frequency characteristics of electricity consumption fluctuations, the time-frequency distribution of electrical loads, and the time-frequency characteristics of power spectra. The tensor decomposition algorithm can effectively extract the core structure from multi-dimensional data sets, making the originally complex data more intuitive and understandable.For example, when analyzing the long-term operation trend of a regional power grid, the future power demand can be predicted by observing the time-frequency characteristics of the electricity consumption fluctuations; the dispatching strategy can be optimized through the time-frequency distribution of the electrical load to ensure sufficient reserve capacity during peak periods; and the time-frequency characteristics of the power spectrum help evaluate the quality of the power grid. Especially when facing the intermittent input of renewable energy, timely adjustments can be made to maintain a stable power supply frequency. Finally, in order to make the obtained key feature tensors applicable to fast processing and efficient storage in practical applications, dimensionality reduction processing is also required. The purpose of dimensionality reduction processing is to remove redundant information, retain the core features that best represent the power grid state, and minimize the consumption of computing resources as much as possible. In this process, principal component analysis (PCA), linear discriminant analysis (LDA), or other suitable dimensionality reduction techniques can be used. For example, in the aforementioned example of the urban power grid, the power grid acquisition data after dimensionality reduction processing can be used to train a machine learning model to predict the power grid operation trend in the future for a period of time and formulate a reasonable energy storage dispatching plan. This can not only improve the prediction accuracy but also speed up the response speed, ensuring that even in a complex power environment, the optimal response can be made quickly to maintain the power grid frequency stability. To sum up, sampling key nodes of the target power grid through a distributed optical fiber sensing array to obtain the original electrical parameter data, and further performing wavelet packet transform analysis on these data and using the tensor decomposition algorithm for high-order singular value decomposition, and finally completing the dimensionality reduction processing. This series of steps constitutes a complete process from the original data to the power grid acquisition data that can be used for decision support. This method not only improves the accuracy and breadth of data acquisition but also excavates the deep-level information of the power grid operation through advanced data analysis means, providing strong technical support for the effective management of the energy storage system and the stable operation of the power grid. In the context of the increasing complexity of modern power systems, such data processing and analysis capabilities are particularly important, enabling us to maintain precise control over the power grid state in a more dynamic and uncertain environment and thus achieve more intelligent and efficient energy management.

[0080] In a specific embodiment, the prediction of the power grid acquisition data through a preset big data analysis technology to obtain the power grid operation trend includes:

[0081] Performing data preprocessing on the power grid acquisition data to obtain preprocessed electrical data;

[0082] Performing feature extraction on the preprocessed electrical data to obtain electrical extraction features; wherein, the electrical extraction features;

[0083] Performing time series decomposition on the electrical extraction features to obtain multi-scale time series electrical features; wherein, the multi-scale time series electrical features include load change features, voltage time domain features, current time domain features, power quality features, and power factor dynamic characteristics;

[0084] Perform a causality analysis on the multi-scale time series electrical characteristics through the transfer entropy algorithm to obtain a feature correlation network; wherein, the feature correlation network is used to reveal the dynamic operation correlations between the multi-scale time series electrical characteristics in the target power grid.

[0085] Use the adaptive kernel density estimation method to estimate the state transition probability of the target power grid based on the feature correlation network to obtain a power grid state transition probability table; wherein, the power grid state transition probability table is used to quantify the conversion probability of the target power grid between different operating conditions, so as to perform a quantitative dynamic probability prediction on the future power grid operating state.

[0086] Perform a Markov chain prediction on the target power grid based on the power grid state transition probability table to obtain a power grid state prediction sequence.

[0087] Adopt the self-organized criticality analysis method to perform a non-linear dynamic analysis on the power grid state prediction sequence to obtain the power grid operation trend; wherein, the power grid operation trend includes voltage stability, frequency stability, load change trend, and change trend of power quality.

[0088] Specifically, the process of predicting the grid operation trend by preset big data analysis technology for the grid acquisition data is one of the core links in the energy storage grid-forming control method. This process not only involves the preliminary processing and feature extraction of the grid acquisition data, but also deeply applies various advanced technical means such as time series data analysis, causal relationship analysis, state transition probability estimation, and non-linear dynamic analysis, and finally realizes the quantitative prediction of the future grid operation state. The following is a specific explanation of this complex process. First, for the original grid acquisition data obtained from the distributed optical fiber sensing array, data preprocessing must be carried out to obtain preprocessed electrical data. The data preprocessing steps aim to clean and standardize these original data to ensure their quality and consistency. For example, in an urban power grid containing multiple wind farms, due to the possible influence of environmental noise or equipment failures on the sensors, the collected data may have missing values or outliers. To ensure the effectiveness of subsequent analysis, interpolation methods are needed to fill in the missing data, and statistical tests are used to identify and remove outliers. In addition, considering the possible differences in measurement units and timestamps between different sensors, unified conversion and synchronization processing are also required. Such preprocessing work lays a solid foundation for subsequent feature extraction. Next, feature extraction is performed on the preprocessed electrical data to obtain electrical extraction features. The goal of this step is to extract the most representative and discriminatory features from a large amount of preprocessed electrical data to better describe the state of the power grid. Feature extraction can be based on traditional statistical methods such as mean, variance, etc., or combined with modern machine learning algorithms such as principal component analysis (PCA), independent component analysis (ICA), etc. For example, in the application scenario of the aforementioned urban power grid, feature extraction can help identify which time periods have the most drastic fluctuations in electricity consumption, or which areas have the most unstable voltage levels. In this way, the internal dynamic change patterns of the power grid can be understood more clearly, providing support for further in-depth analysis. Subsequently, time series data decomposition is performed on the electrical extraction features to obtain multi-scale time series electrical features. Here, the multi-scale time series electrical features include multiple aspects such as load change features, voltage time domain features, current time domain features, power quality features, and power factor dynamic characteristics. Time series data decomposition techniques such as empirical mode decomposition (EMD), wavelet transform, etc., can decompose complex time series data into different frequency components, revealing the change laws at each level. For example, when analyzing the output power of a certain wind farm, long-term trend changes caused by weather conditions and short-term rapid changes caused by instantaneous wind speed fluctuations can be identified through time series data decomposition. This multi-scale perspective helps to comprehensively understand the grid operation state and provides a rich information source for subsequent causal relationship analysis. Then, causal relationship analysis is performed on the multi-scale time series electrical features through the transfer entropy algorithm to obtain a feature association network.Transfer entropy is a method for measuring the causal relationship between variables, which can reveal the mutual influence between different electrical characteristics and their transmission paths. The feature correlation network is used to visually display these causal relationships, helping us understand the dynamic operational associations between multi-scale time series electrical characteristics in the target power grid. For example, in the aforementioned example of the urban power grid, by constructing a feature correlation network, we can discover how certain specific load changes trigger voltage level changes, or how the adjustment of power factor affects the overall power quality. Such analysis not only deepens our understanding of the internal mechanisms of the power grid but also provides a theoretical basis for formulating effective control strategies. Further, using the adaptive kernel density estimation method, the state transition probability of the target power grid is estimated based on the feature correlation network, obtaining a power grid state transition probability table. Adaptive kernel density estimation is a non-parametric probability density estimation method that can flexibly adapt to different types of data distributions, thus accurately quantifying the conversion probability of the target power grid between different operating conditions. For example, in a practical application scenario, if the wind power output increases significantly during a certain period, the power grid state transition probability table can tell us the probability of corresponding changes in other key indicators (such as voltage stability, frequency stability, etc.) under such circumstances. Such probability estimation provides a quantitative dynamic prediction for the future operating state of the power grid, enabling us to make preparations in advance and take necessary preventive measures. Finally, based on the power grid state transition probability table, a Markov chain prediction is performed on the target power grid to obtain a power grid state prediction sequence. The Markov chain prediction model assumes that the future state of the system depends only on the current state and is not affected by past states. Through this method, a series of possible power grid state evolution paths, that is, the power grid state prediction sequence, can be generated. For example, in the aforementioned application scenario of the urban power grid, Markov chain prediction can help us simulate different state combinations that the power grid may experience in the next few hours, including possible peak load periods, voltage drop risks, etc. Such prediction results are crucial for optimizing the energy storage scheduling plan because it provides clear guidance on when and where to take actions. On this basis, using the self-organized criticality analysis method, a non-linear dynamic analysis is performed on the power grid state prediction sequence to obtain the power grid operation trend. Self-organized criticality analysis focuses on the behavior of the system when approaching the critical point and is particularly suitable for describing networks with complex dynamic characteristics such as power systems. Through this method, the ability of the power grid to maintain stability under different load conditions can be evaluated, and the change trends of voltage stability, frequency stability, load change, and power quality can be predicted. For example, in the face of a sudden high load demand, self-organized criticality analysis can help determine whether the power grid is within a safe range or whether there are potential instability risks. Such analysis results not only enhance our understanding of the power grid behavior but also strengthen the ability to respond to emergencies, ensuring the safe and reliable power supply.In summary, the process of predicting the power grid acquisition data through the preset big data analysis technology to obtain the power grid operation trend not only demonstrates the complete chain from data to knowledge to decision support, but also reflects the advanced concepts and technical means in modern power system management and optimization. In this process, each step is closely linked, jointly constituting a powerful prediction framework, enabling us to more accurately foresee the future operation state of the power grid, make reasonable regulation decisions in a timely manner to maintain the stability of the power grid frequency and the excellent quality of electric energy. The application of this method is of great significance for promoting the efficient utilization of clean energy, improving the flexibility and reliability of the power system.

[0089] In a specific embodiment, the energy state assessment of the target energy storage system to obtain the available capacity of the target energy storage system includes:

[0090] Real-time current monitoring of the target energy storage system is performed through an intelligent sensor to obtain real-time current data;

[0091] Using the segmented coulomb meter method, the cumulative charge amount of the target energy storage system is calculated based on the real-time current data to obtain a cumulative charge amount sequence;

[0092] Multi-scale decomposition of the cumulative charge amount sequence is performed through a multi-resolution analysis method to obtain multi-scale cumulative charge amount characteristics;

[0093] The multi-scale cumulative charge amount characteristics are dynamically adjusted using an adaptive filtering algorithm to obtain an estimated value of the cumulative charge amount;

[0094] Based on the estimated value of the cumulative charge amount, an analysis of the relationship between the storage voltage and the stored charge of the target energy storage system is performed to obtain a storage voltage-stored charge characteristic curve;

[0095] The rated capacity of the target energy storage system and the performance record of the target energy storage system are obtained from the database, and the available capacity of the target energy storage system is dynamically evaluated and inversely calculated through the storage voltage-stored charge characteristic curve, the rated capacity, and the performance record to obtain the available capacity of the target energy storage system.

[0096] Specifically, the process of evaluating the energy state of the target energy storage system to obtain the available capacity of the target energy storage system is one of the key steps to ensure the effective implementation of the energy storage grid-forming control method. This process not only involves obtaining real-time current data through intelligent sensors, but also covers multiple complex technical links, from calculating the cumulative charge amount, multi-scale decomposition, dynamic adjustment to the final evaluation of the available capacity. The following is a detailed explanation of this specific solution. First, the target energy storage system is monitored in real-time for current through intelligent sensors to obtain real-time current data. The intelligent sensors are deployed at various key positions of the energy storage system and can continuously capture the current changes flowing through the energy storage device. For example, in an urban power grid containing multiple lithium battery energy storage units, these sensors can be installed on the input and output ports of each battery module to ensure high-precision and high-frequency current measurement values. The real-time current data is not only the basis for understanding the current working state of the energy storage system, but also provides important original information for subsequent energy state evaluation. Next, using the segmented Coulomb meter method, the cumulative charge amount of the target energy storage system is calculated based on the real-time current data to obtain a cumulative charge amount sequence. The segmented Coulomb meter method is a method for accurately calculating the cumulative charge amount during the charge and discharge process. It divides the entire charge and discharge cycle into several small time periods and calculates the net charge inflow or outflow in each time period separately. This method is particularly suitable for dealing with non-linear charging curves and complex current waveforms. For example, in the application scenario of the aforementioned urban power grid, when a certain energy storage unit experiences a period of rapid charging and then enters a slow discharge stage, the segmented Coulomb meter method can accurately track the change of the cumulative charge through detailed time segmentation and generate a detailed cumulative charge amount sequence. This provides a reliable data basis for subsequent analysis. Then, through multi-resolution analysis methods, the multi-scale decomposition of the cumulative charge amount sequence is carried out to obtain multi-scale cumulative charge amount characteristics. Multi-resolution analysis methods such as wavelet transform can decompose the cumulative charge amount sequence into components in different frequency ranges, thereby revealing the charge accumulation characteristics at different time scales. For example, for a long-term operating energy storage system, there may be daily periodic load fluctuations and seasonal power generation pattern changes, while on a short time scale, there are the effects of instantaneous power mutations. By decomposing the cumulative charge amount sequence into multiple scales, we can identify the change laws at these different levels, and thus better understand the usage history of the energy storage system and its impact on the current state. Further, an adaptive filtering algorithm is used to dynamically adjust the multi-scale cumulative charge amount characteristics to obtain an estimated value of the cumulative charge amount. The adaptive filtering algorithm can automatically adjust parameters according to the changes of the multi-scale cumulative charge amount characteristics to eliminate noise interference and improve the estimation accuracy. For example, in practical applications, measurement errors caused by environmental temperature changes or other factors may cause abnormal fluctuations in the cumulative charge amount sequence.At this time, the adaptive filtering algorithm can correct these errors in real time according to the historical trend and current change of the cumulative charge quantity characteristics, and provide a more accurate estimated value of the cumulative charge quantity. Such dynamic adjustment helps to maintain the consistency and reliability of the evaluation results of the energy storage system. On this basis, based on the estimated value of the cumulative charge quantity, the relationship between the storage voltage and the stored charge of the target energy storage system is analyzed to obtain a storage voltage-stored charge characteristic curve. This characteristic curve describes the relationship between the internal charge storage and the external voltage of the energy storage system, reflecting its physical characteristics and working state. For example, in a lithium battery energy storage system, as the charge and discharge level changes, the terminal voltage of the battery will change accordingly, forming a specific voltage-charge curve. By comparing the estimated value of the cumulative charge quantity with the actually measured storage voltage, a characteristic curve that accurately reflects the current health status of the energy storage system can be drawn, which is crucial for evaluating its remaining charge and predicting future performance. Finally, the rated capacity of the target energy storage system and the performance records of the target energy storage system are obtained from the database, and the available capacity of the target energy storage system is dynamically evaluated and back-calculated through the storage voltage-stored charge characteristic curve, the rated capacity and the performance records, to obtain the available capacity of the target energy storage system. The database stores detailed information about the energy storage system, including the rated capacity at the time of factory, previous maintenance records, and historical operation data, etc. Combining the above characteristic curve, the current available capacity of the energy storage system can be dynamically evaluated. For example, in the aforementioned example of the urban power grid, if a certain energy storage unit has been frequently charged and discharged for a period of time, its actual available capacity may have been lower than the initial rated value. By comprehensively analyzing the characteristic curve, the rated capacity and the performance records, the current true available capacity of the energy storage unit can be accurately calculated to reasonably arrange its tasks of participating in power grid dispatching. In summary, the process of obtaining the available capacity of the target energy storage system through the energy state evaluation of the target energy storage system not only demonstrates the complete chain from real-time data collection to advanced data analysis, but also reflects the advanced concepts and technical means in the modern energy storage management system. In this process, each technical link is closely connected, jointly constituting a powerful evaluation framework, enabling us to more accurately grasp the actual state of the energy storage system, make reasonable dispatching decisions in a timely manner, and maintain the stability of the power grid frequency and the excellent quality of electric energy. The application of this method is of great significance for optimizing the use efficiency of energy storage resources and improving the flexibility and reliability of the power system.

[0097] In a specific embodiment, based on the dynamic evaluation and back-calculation of the available capacity of the target energy storage system through the storage voltage-stored charge characteristic curve, the rated capacity and the performance records, to obtain the available capacity of the target energy storage system, including:

[0098] The segmented exponential fitting algorithm is used to perform segmented fitting on the storage voltage-storage charge characteristic curve to obtain the charge and discharge capacity loss characteristic parameters;

[0099] Based on the charge and discharge capacity loss characteristic parameters, weighted statistics of the cycle times of the target energy storage system are performed to obtain the capacity attenuation coefficient;

[0100] Based on the capacity attenuation coefficient and the performance record, a collaborative analysis of the target energy storage system is carried out to obtain the health status of the target energy storage system;

[0101] Based on the health status and the rated capacity, a dynamic capacity assessment of the current energy storage system is carried out to obtain the theoretical available capacity value;

[0102] The current temperature parameter is obtained through a preset temperature sensor, and the capacity value compensation is performed on the theoretical available capacity value based on the current temperature parameter to obtain the temperature-corrected capacity value;

[0103] Based on the temperature-corrected capacity value, the charge and discharge capacity loss characteristic parameters, the capacity attenuation coefficient and the health status, the available capacity of the target energy storage system is calculated by backtracking to obtain the available capacity of the target energy storage system.

[0104] Specifically, the process of dynamically evaluating and back-calculating the available capacity of the target energy storage system based on the storage voltage-storage charge characteristic curve, rated capacity, and performance records is a key link in ensuring the efficient management and optimal scheduling of the energy storage system. This process not only involves the fine processing of the storage voltage-storage charge characteristic curve but also includes multiple technical steps such as the extraction of charge-discharge capacity loss characteristic parameters, the statistics of capacity attenuation coefficients, the collaborative analysis of health status, the dynamic evaluation of theoretical available capacity values, and the compensation of temperature-corrected capacity values. The following is a specific explanation of this complex solution. First, the piecewise exponential fitting algorithm is used to perform piecewise fitting on the storage voltage-storage charge characteristic curve to obtain the charge-discharge capacity loss characteristic parameters. Piecewise exponential fitting is a mathematical modeling method that divides the characteristic curve into several linear or non-linear parts and fits an exponential function to each part. This method is particularly suitable for describing the non-linear behavior of the energy storage system under different charge states. For example, in an urban power grid with multiple lithium battery energy storage units, as the number of charge-discharge cycles of the battery increases, internal chemical reactions may cause the capacity to gradually decrease. Through piecewise exponential fitting, we can identify these changing trends and quantify the capacity loss at each stage, thereby obtaining the charge-discharge capacity loss characteristic parameters. This not only helps to understand the current state of the energy storage system but also provides a scientific basis for subsequent capacity evaluation. Next, based on the charge-discharge capacity loss characteristic parameters, weighted statistics of the number of cycles are performed on the target energy storage system to obtain the capacity attenuation coefficient. Here, the weighted statistics of the number of cycles refer to assigning different weights according to the historical operation data of the energy storage system, especially the number of charge-discharge cycles, to reflect the impact of each cycle on the total capacity. For example, in the application scenario of the aforementioned urban power grid, if a certain energy storage unit has experienced a large number of rapid charge-discharge operations in the past year, then these high-frequency short-cycle operations may have a greater impact on the total capacity than low-frequency long-cycle operations. By performing weighted statistics on these data, the capacity attenuation coefficient can be accurately calculated, which directly reflects the capacity loss of the energy storage system due to long-term use. Further, based on the capacity attenuation coefficient and the performance records, a collaborative analysis of the target energy storage system is carried out to obtain the health status of the target energy storage system. The so-called collaborative analysis here is not just a simple data summary but comprehensively considers information such as the capacity attenuation coefficient and the maintenance records and fault reports of the energy storage system in previous times. For example, in practical applications, in addition to the number of charge-discharge cycles, factors such as environmental conditions (such as temperature and humidity) and external load requirements will also affect the life of the energy storage system. Through collaborative analysis, the health status of the energy storage system can be comprehensively evaluated to determine whether it is within the normal working range or whether there are potential risks that need to be dealt with in a timely manner.Such analysis results are of great significance for formulating a reasonable maintenance plan and extending the service life of the energy storage system. On this basis, based on the health status and the rated capacity, a dynamic capacity assessment of the current energy storage system is carried out to obtain a theoretical available capacity value. The dynamic capacity assessment is a continuously updated process that combines the latest health status assessment results and the initial rated capacity of the energy storage system, providing a value that reflects the current energy storage capacity of the energy storage system in real time. For example, in the aforementioned example of the urban power grid, if the health status of a certain energy storage unit has declined after a period of frequent use, then even if its rated capacity has not changed, its theoretical available capacity value should be adjusted accordingly. This dynamic assessment can help us more accurately grasp the actual state of the energy storage system, so as to reasonably arrange its tasks of participating in power grid dispatching. In addition, in order to improve the accuracy of the assessment results, it is also necessary to obtain the current temperature parameter through a preset temperature sensor and compensate the theoretical available capacity value based on the current temperature parameter to obtain a temperature-corrected capacity value. Temperature is one of the important factors affecting the performance of the energy storage system. Especially under extreme conditions, temperature changes may cause changes in the internal chemical reaction rate of the battery, thereby affecting its charge and discharge efficiency and capacity. For example, in cold winter weather, the charge and discharge performance of lithium batteries is usually inhibited. Therefore, it is necessary to appropriately adjust the theoretical available capacity value according to the actual measured temperature parameter to obtain a more realistic temperature-corrected capacity value. Such compensation measures ensure the consistency and reliability of the assessment results under different environmental conditions. Finally, based on the temperature-corrected capacity value, the charge and discharge capacity loss characteristic parameters, the capacity attenuation coefficient and the health status, a reverse calculation of the available capacity of the target energy storage system is carried out to obtain the available capacity of the target energy storage system. Reverse calculation refers to combining all the above assessment results and reasoning backward to obtain the maximum available capacity that the energy storage system can provide in the current state. For example, in the aforementioned application scenario of the urban power grid, when facing an upcoming power consumption peak, it is possible to determine through reverse calculation which energy storage units can provide sufficient power support at critical moments, and which ones may not be able to meet the demand due to poor health status or other reasons. This accurate assessment of available capacity not only helps to optimize the dispatching of energy storage resources, but also effectively prevents problems such as insufficient power supply caused by misjudgment. To sum up, the process of dynamic assessment and reverse calculation of the available capacity of the target energy storage system based on the storage voltage-storage charge characteristic curve, rated capacity and performance records not only demonstrates a complete chain from characteristic curve modeling to capacity assessment, but also reflects the advanced concepts and technical means in modern energy storage management systems. In this process, each technical link is closely connected, jointly constituting a powerful assessment framework, enabling us to more accurately grasp the actual state of the energy storage system and make reasonable dispatching decisions in a timely manner to maintain the stability of the power grid frequency and excellent power quality.The application of this method is of great significance for optimizing the utilization efficiency of energy storage resources and improving the flexibility and reliability of power systems.

[0105] In a specific embodiment, based on the charge-discharge capacity loss characteristic parameters, a weighted statistics of the number of cycles of the target energy storage system is performed to obtain a capacity attenuation coefficient, including:

[0106] Perform time series analysis on the charge-discharge capacity loss characteristic parameters to obtain a charge-discharge cycle sequence;

[0107] Based on the charge-discharge cycle sequence, count the number of charging and discharging times of each cycle of the target energy storage system to obtain a cycle number data set;

[0108] Use the weighted moving average method to perform weighted average calculation on the cycle number data set to obtain a weighted cycle number;

[0109] Based on the weighted cycle number, perform fitting analysis on the capacity loss of the energy storage system to obtain a capacity attenuation curve;

[0110] Perform slope analysis on the capacity attenuation curve to obtain a capacity attenuation rate;

[0111] Based on the capacity attenuation rate and the weighted cycle number, calculate the quantity attenuation coefficient of the target energy storage system to obtain a capacity attenuation coefficient.

[0112] Specifically, the process of performing weighted statistics on the cycle count of the target energy storage system based on the charge-discharge capacity loss characteristic parameters to obtain the capacity attenuation coefficient is one of the key steps in the energy storage grid-forming control method for evaluating the health status of the energy storage system and predicting its future performance. This process not only involves time series analysis of the charge-discharge capacity loss characteristic parameters but also includes multiple technical aspects such as statistics of the cycle count dataset, weighted average calculation, fitting analysis, and finally the calculation of the capacity attenuation coefficient. The following is a detailed explanation of this complex scheme. First, perform time series analysis on the charge-discharge capacity loss characteristic parameters to obtain the charge-discharge cycle sequence. Time series analysis is a technique used to identify and quantify data patterns that change over time. Here, through time series analysis of the charge-discharge capacity loss characteristic parameters, the charge-discharge behavior of the energy storage system at different time periods and its impact on capacity can be captured. For example, in an urban power grid containing multiple lithium battery energy storage units, these characteristic parameters record the changes in battery capacity after each charge-discharge operation. Through time series analysis, we can connect these discrete data points to form a continuous charge-discharge cycle sequence, which intuitively shows the historical operation trajectory of the energy storage system and provides a basis for subsequent statistical work. Next, based on the charge-discharge cycle sequence, count the number of charging and discharging times for each cycle of the target energy storage system to obtain the cycle count dataset. Here, the cycle count refers to the number of times the energy storage system undergoes a complete charge-discharge process. Counting the number of charging and discharging times in each cycle helps us understand the usage frequency and intensity of the energy storage system. For example, in the aforementioned application scenario of the urban power grid, if a certain energy storage unit has experienced a large number of rapid charge-discharge operations in the past year, then these high-frequency short-cycle operations may have a greater impact on the total capacity than low-frequency long-cycle operations. By constructing the cycle count dataset, we can comprehensively understand the working history of the energy storage system and provide detailed input data for subsequent weighted average calculations. Further, use the weighted moving average method to perform weighted average calculation on the cycle count dataset to obtain the weighted cycle count. The weighted moving average method is a commonly used smoothing technique that assigns different weights to data in different time periods, thereby reducing the impact of short-term fluctuations and highlighting long-term trends. In the evaluation of energy storage systems, this method is particularly suitable for dealing with non-uniform charge-discharge behaviors caused by changes in environmental conditions or load demands. For example, in practical applications, charge-discharge activities may be more frequent in some periods and relatively less in others. Through the weighted moving average method, appropriate weights can be assigned to each cycle according to the importance and impact degree of each cycle, and then a weighted cycle count that can reflect the overall usage of the energy storage system can be calculated. Such a calculation result is not only more accurate but also more representative. On this basis, based on the weighted cycle count, perform fitting analysis on the capacity loss of the energy storage system to obtain the capacity attenuation curve.The fitting analysis aims to find the best mathematical model to describe the trend that the capacity of the energy storage system gradually decreases as the number of cycles increases. For example, linear regression, exponential functions, or other suitable models can be used to fit the relationship between the weighted number of cycles and the capacity loss. The capacity decay curve not only intuitively shows the life characteristics of the energy storage system but also provides an important reference for predicting future capacity changes. By comparing the capacity decay curves in different time periods, the performance of the energy storage system under different operating conditions can be evaluated, and a basis for optimizing its operation strategy can be provided. Then, slope analysis is performed on the capacity decay curve to obtain the capacity decay rate. Slope analysis is a method for quantitatively describing the capacity decay curve, which reflects the speed of capacity loss of the energy storage system after a certain number of cycles. For example, in the aforementioned example of the urban power grid, if the capacity decay curve of a certain energy storage unit shows a relatively steep slope, it indicates that the capacity loss speed of this unit is relatively fast, and more frequent maintenance or replacement may be required. On the contrary, if the slope is relatively gentle, it means that this unit has good durability and can maintain a high performance level for a long time. Through slope analysis, the aging rate of the energy storage system can be quantified, and a scientific basis for formulating a reasonable maintenance plan can be provided. Finally, based on the capacity decay rate and the weighted number of cycles, the capacity decay coefficient of the target energy storage system is calculated to obtain the capacity decay coefficient. The capacity decay coefficient is a comprehensive index that combines the results of slope analysis and the information of the weighted number of cycles to characterize the capacity change characteristics of the energy storage system throughout its life cycle. For example, in an actual application scenario, if a certain energy storage unit has undergone a large number of charge-discharge cycles and its capacity decay curve shows a significant downward trend, then by calculating the capacity decay coefficient, we can obtain a specific value to measure the current health status and remaining service life of this unit. Such accurate evaluation results are crucial for optimizing the scheduling of energy storage resources, extending the service life of equipment, and ensuring the stability of power supply. In summary, the process of obtaining the capacity decay coefficient by performing weighted statistics on the number of cycles of the target energy storage system based on the charge-discharge capacity loss characteristic parameters not only demonstrates a complete chain from time series analysis to capacity decay coefficient calculation but also reflects the advanced concepts and technical means in modern energy storage management systems. In this process, each technical link is closely connected, jointly constituting a powerful evaluation framework, enabling us to more accurately grasp the actual state of the energy storage system, make reasonable scheduling decisions in a timely manner, and maintain the stability of the power grid frequency and excellent power quality. The application of this method is of great significance for optimizing the use efficiency of energy storage resources and improving the flexibility and reliability of the power system.

[0113] In a specific embodiment, the formulating of the scheduling control plan based on the available capacity and the power grid operation trend includes:

[0114] The available capacity and the power grid operation trend are modeled for state association through a dynamic programming algorithm to obtain available capacity - power grid operation trend association data; wherein, the available capacity - power grid operation trend association data includes the energy storage capacity state and the load power deviation of the target power grid.

[0115] The available capacity - power grid operation trend association data is quantitatively evaluated by using a preset fuzzy neural network to obtain a state evaluation result; wherein, the state evaluation result includes a capacity adequacy parameter and a power grid stability parameter.

[0116] The state evaluation result is monitored. If any of the capacity adequacy parameter, the power grid stability parameter, and the load balance parameter in the state evaluation result is not within the preset range, multi - objective weight allocation is performed on the state evaluation result through a hybrid particle swarm optimization algorithm to obtain a scheduling priority sequence.

[0117] The scheduling priority sequence is transformed into a preliminary scheduling instruction sequence.

[0118] The preliminary scheduling instruction sequence is coordinated in time sequence through a distributed collaborative optimization algorithm to obtain a scheduling control plan.

[0119] Specifically, the process of formulating a scheduling control plan based on the available capacity and the power grid operation trend is a crucial step in the energy storage grid-forming control method. This process not only involves in-depth analysis of the available capacity of the energy storage system and the power grid operation trend, but also combines various advanced technologies and methods such as dynamic programming algorithms, fuzzy neural networks, hybrid particle swarm optimization algorithms, and distributed cooperative optimization algorithms to ensure that the final obtained scheduling control plan can effectively maintain the stability of the target power grid frequency. The following is a specific explanation of this complex scheme. First, the dynamic programming algorithm is used to perform state correlation modeling on the available capacity and the power grid operation trend, obtaining available capacity-power grid operation trend correlation data. The dynamic programming algorithm here is a mathematical optimization method for solving multi-stage decision-making problems. It can decompose complex system behaviors into a series of simple sub-problems and gradually solve the optimal solution. In this process, the dynamic programming algorithm will comprehensively consider the current available capacity of the energy storage system and the prediction of the future power grid operation trend, and construct a model reflecting the relationship between the two. For example, in an urban power grid with multiple wind farms, if it is predicted that the wind speed will decrease in the next few hours, resulting in a reduction in wind power output, the dynamic programming algorithm can help us understand which energy storage devices can be put into use at what time in this situation to make up for the possible energy gap. The result of this correlation modeling forms the so-called available capacity-power grid operation trend correlation data, including the energy storage capacity status and the load power deviation of the target power grid. These information are crucial for subsequent quantitative evaluation. Next, the preset fuzzy neural network is used to quantitatively evaluate the available capacity-power grid operation trend correlation data, obtaining a state evaluation result. The fuzzy neural network is an artificial intelligence technology that combines the advantages of fuzzy logic and neural networks. It can handle uncertainty and nonlinear problems and is very suitable for the complex environment of the power system. By quantitatively evaluating the correlation data, the fuzzy neural network can calculate two key parameters: the capacity adequacy parameter and the power grid stability parameter. The capacity adequacy parameter reflects whether the energy storage system has enough remaining capacity to cope with upcoming demand changes; while the power grid stability parameter measures the degree to which the power grid maintains the stability of frequency and other performance indicators under the current conditions. For example, in the application scenario of the aforementioned urban power grid, when a certain energy storage unit is close to the full charge state, its capacity adequacy parameter may be relatively low, indicating that it is difficult for this unit to provide additional support in the next period of time; on the contrary, if the frequency fluctuation of the power grid is small and the voltage level is stable, the power grid stability parameter is relatively high, indicating that the power grid is currently in a good operating state. Further, the state evaluation result is monitored. If any of the capacity adequacy parameter, power grid stability parameter, or load balance parameter in the state evaluation result is not within the preset range, the hybrid particle swarm optimization algorithm is used to perform multi-objective weight allocation on the state evaluation result, obtaining a scheduling priority sequence.The monitoring mentioned here refers to continuously tracking the changes in the status assessment results to ensure that they always meet the expected standards. Once it is found that some parameters deviate from the normal range, immediate measures need to be taken for adjustment. The hybrid particle swarm optimization algorithm is a heuristic search algorithm that can find the global optimal solution in multi-dimensional space and is suitable for dealing with complex optimization problems involving multiple objectives. In this case, the hybrid particle swarm optimization algorithm can determine how to prioritize the use of each energy storage unit and other regulation resources according to different weight allocation strategies, so as to generate a reasonable scheduling priority sequence. For example, in practical applications, if the load demand suddenly increases within a certain period, resulting in a decrease in the power grid stability parameter, then by adjusting the charge and discharge sequence and intensity of the energy storage unit, the stability of the power grid can be quickly restored. On this basis, the scheduling priority sequence is transformed into scheduling instructions to obtain a preliminary scheduling instruction sequence. The transformation of scheduling instructions refers to transforming the abstract priority sequence into specific operation commands, such as the set values of the charge and discharge power of the energy storage unit, the input or cut-off time, etc. This step ensures that the scheduling plan can be executed and also provides a basis for subsequent timing coordination. For example, in the aforementioned example of the urban power grid, according to the scheduling priority sequence, it can be decided to first enable those energy storage units with good health conditions and sufficient capacity to respond to peak loads, and at the same time arrange other units to carry out supplementary charging during low load periods. Such scheduling instructions not only help to improve energy utilization efficiency but also enhance the overall flexibility of the power grid. Finally, the preliminary scheduling instruction sequence is coordinated in time through a distributed cooperative optimization algorithm to obtain a scheduling control plan. The distributed cooperative optimization algorithm is a technology that can achieve synchronous optimization of multi-agent systems in a distributed environment and is particularly suitable for dealing with complex scheduling tasks across regions and departments. Through this method, it can be ensured that the operations between different energy storage units are coordinated with each other, avoiding the situation of overall sub-optimality caused by local optimality. For example, in a large urban power grid, multiple energy storage units are distributed in different substations, and each unit formulates preliminary scheduling instructions according to its own status assessment results. However, due to possible resource competition or uneven load distribution among the units, it is necessary to readjust the timing arrangement of these instructions through the distributed cooperative optimization algorithm to make the scheduling of the entire power grid more efficient and harmonious. The finally formed scheduling control plan can not only meet the immediate power demand but also maximize economic and social benefits. To sum up, the process of formulating a scheduling control plan based on the available capacity and the power grid operation trend not only demonstrates a complete chain from state correlation modeling to timing coordination but also reflects the advanced concepts and technical means in modern energy storage management systems. In this process, each technical link is closely connected, jointly constituting a powerful scheduling framework, enabling us to more accurately grasp the actual state of the energy storage system, make reasonable scheduling decisions in a timely manner, and maintain the stability of the power grid frequency and excellent power quality.The application of this method is of great significance for optimizing the utilization efficiency of energy storage resources and improving the flexibility and reliability of the power system.

[0120] The energy storage grid-forming control method in the embodiments of the present invention has been described above. Next, the energy storage grid-forming control system in the embodiments of the present invention will be described. Please refer to Figure 2 One embodiment of the energy storage grid-forming control system in the embodiments of the present invention includes:

[0121] An acquisition module 21, configured to perform real-time data acquisition on a target power grid through intelligent sensors to obtain power grid acquisition data;

[0122] A prediction module 22, configured to predict the power grid acquisition data through a preset big data analysis technology to obtain the power grid operation trend;

[0123] An evaluation module 23, configured to evaluate the energy state of a target energy storage system to obtain the available capacity of the target energy storage system;

[0124] A control module 24, formulating a scheduling control plan based on the available capacity and the power grid operation trend; wherein, the scheduling control plan is used to schedule and control the target power grid or the target energy storage system to maintain the frequency stability of the target power grid.

[0125] In this embodiment, for the specific implementation of each unit in the above system embodiment, please refer to that described in the above method embodiment, and details will not be elaborated here.

[0126] Refer to Figure 3 In the embodiments of the present invention, a computer device is further provided. The internal structure of the computer device can be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0127] Those skilled in the art can understand that Figure 3 the structure shown in

[0128] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0129] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or an external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0130] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, device, article, or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such a process, device, article, or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, device, article, or method including that element.

[0131] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, are equally included in the patent protection scope of the present invention.

Claims

1. A method for controlling energy storage network, characterized in that: The following steps are involved: Real-time data collection of the target power grid is performed through intelligent sensors to obtain power grid collection data; The power grid collected data is predicted by using a preset big data analysis technology to obtain a power grid operation trend; Evaluate the energy status of the target energy storage system to obtain the available capacity of the target energy storage system; Formulate a dispatching control plan based on the available capacity and the power grid operation trend; wherein the dispatching control plan is used to dispatch and control the target power grid or target energy storage system to keep the target power grid frequency stable; The predicting of the power grid collected data by using a preset big data analysis technology to obtain the power grid operation trend includes: Performing data preprocessing on the power grid collected data to obtain preprocessed electrical data; Performing feature extraction on the preprocessed electrical data to obtain electrical extraction features; wherein the electrical extraction features; Performing time series data decomposition on the electrical extraction features to obtain multi-scale time series electrical features; wherein the multi-scale time series electrical features include load change features, voltage time domain features, current time domain features, power quality features, and power factor dynamic characteristics; Performing causal analysis on the electrical characteristics of the multi-scale time series by using a migration entropy algorithm to obtain a feature association network; wherein the feature association network is used to reveal the dynamic operation association between the electrical characteristics of each multi-scale time series in the target power grid; Using an adaptive kernel density estimation method, the state transition probability of the target power grid is estimated based on the feature association network to obtain a power grid state transition probability table; wherein the power grid state transition probability table is used to quantify the conversion probability of the target power grid between different working conditions, thereby performing a quantitative dynamic probability prediction of the future power grid operation state; Performing Markov chain prediction on the target power grid based on the power grid state transition probability table to obtain a power grid state prediction sequence; A self-organized criticality analysis method is used to perform nonlinear dynamic analysis on the grid state prediction sequence to obtain a grid operation trend; wherein the grid operation trend includes voltage stability, frequency stability, load change trend, and power quality change trend.

2. The energy storage network control method according to claim 1, characterized in that: The smart sensor is a distributed optical fiber sensor array. The smart sensor is used to collect real-time data of the target power grid to obtain power grid collection data, including: Sampling key nodes of the target power grid through the distributed optical fiber sensor array to obtain original electrical parameter data; Performing wavelet packet transform analysis on the original electrical parameter data to obtain time-frequency electrical characteristic data; Using a preset tensor decomposition algorithm, a high-order singular value decomposition is performed on the time-frequency electrical characteristic data to obtain a key characteristic tensor; wherein the key characteristic tensor includes the time-frequency characteristics of power consumption fluctuation, the time-frequency distribution of power load and the time-frequency characteristics of power spectrum; The key feature tensor is subjected to dimensionality reduction processing to obtain power grid acquisition data.

3. The energy storage network control method according to claim 1, characterized in that: The energy state evaluation of the target energy storage system to obtain the available capacity of the target energy storage system includes: Real-time current data can be obtained by monitoring the target energy storage system through intelligent sensors; Using a segmented coulomb counter method, the accumulated charge of the target energy storage system is calculated based on the real-time current data to obtain an accumulated charge sequence; Performing multi-scale decomposition on the accumulated charge sequence by a multi-resolution analysis method to obtain a multi-scale accumulated charge feature; Dynamically adjusting the multi-scale cumulative charge characteristics using an adaptive filtering algorithm to obtain a cumulative charge estimation value; Performing storage voltage and storage charge relationship analysis on the target energy storage system based on the estimated value of the accumulated charge amount to obtain a storage voltage-storage charge characteristic curve; The rated capacity of the target energy storage system and the performance record of the target energy storage system are obtained from the database, and the available capacity of the target energy storage system is dynamically evaluated and reversely calculated through the storage voltage-storage charge characteristic curve, the rated capacity and the performance record to obtain the available capacity of the target energy storage system.

4. The energy storage network control method according to claim 3 is characterized in that: Based on the dynamic evaluation and reverse calculation of the available capacity of the target energy storage system through the storage voltage-storage charge characteristic curve, the rated capacity and the performance record, the available capacity of the target energy storage system is obtained, including: Using a piecewise exponential fitting algorithm to perform piecewise fitting processing on the storage voltage-storage charge characteristic curve to obtain a charge and discharge capacity loss characteristic parameter; Based on the charge and discharge capacity loss characteristic parameter, weighted statistics of the number of cycles of the target energy storage system are performed to obtain a capacity attenuation coefficient; Performing collaborative analysis on the target energy storage system based on the capacity attenuation coefficient and the performance record to obtain a health status of the target energy storage system; Based on the health status and the rated capacity, a dynamic capacity evaluation is performed on the current energy storage system to obtain a theoretical available capacity value; Acquiring a current temperature parameter through a preset temperature sensor, and performing capacity value compensation on the theoretical available capacity value based on the current temperature parameter to obtain a temperature corrected capacity value; Based on the temperature-corrected capacity value, the charge-discharge capacity loss characteristic parameter, the capacity attenuation coefficient and the health status, the available capacity of the target energy storage system is reversely calculated to obtain the available capacity of the target energy storage system.

5. The energy storage network control method according to claim 4, characterized in that: Based on the charge and discharge capacity loss characteristic parameter, weighted statistics of the number of cycles of the target energy storage system are performed to obtain a capacity attenuation coefficient, including: Performing time series analysis on the charge and discharge capacity loss characteristic parameters to obtain a charge and discharge cycle sequence; Based on the charge and discharge cycle sequence, counting the number of charge and discharge times of each cycle of the target energy storage system to obtain a cycle number data set; Performing weighted average calculation on the cycle number data set using a weighted moving average method to obtain a weighted cycle number; Based on the weighted number of cycles, a capacity loss of the energy storage system is fitted and analyzed to obtain a capacity decay curve; Performing slope analysis on the capacity decay curve to obtain a capacity decay rate; Based on the capacity decay rate and the weighted number of cycles, a capacity decay coefficient is calculated for the target energy storage system to obtain a capacity decay coefficient.

6. The energy storage network control method according to claim 1, characterized in that: The formulating of a dispatching control plan based on the available capacity and the power grid operation trend includes: The available capacity and the grid operation trend are modeled in a state association manner through a dynamic programming algorithm to obtain available capacity-grid operation trend association data; wherein the available capacity-grid operation trend association data includes energy storage capacity state and load power deviation of the target grid; A preset fuzzy neural network is used to quantitatively evaluate the available capacity-grid operation trend correlation data to obtain a state evaluation result; wherein the state evaluation result includes a capacity margin parameter and a grid stability parameter; The state assessment result is monitored, and if any of the capacity margin parameter, the grid stability parameter, and the load balance parameter in the state assessment result is not within a preset range, a multi-objective weight allocation is performed on the state assessment result by using a hybrid particle swarm optimization algorithm to obtain a scheduling priority sequence; Performing scheduling instruction conversion on the scheduling priority sequence to obtain a preliminary scheduling instruction sequence; The preliminary scheduling instruction sequence is time-coordinated through a distributed collaborative optimization algorithm to obtain a scheduling control plan.

7. An energy storage network control device, characterized in that: include: The acquisition module is used to collect real-time data of the target power grid through intelligent sensors to obtain power grid acquisition data; A prediction module, used to predict the collected data of the power grid by using a preset big data analysis technology to obtain the operation trend of the power grid; An evaluation module is used to evaluate the energy state of the target energy storage system and obtain the available capacity of the target energy storage system; A control module, which formulates a dispatching control plan based on the available capacity and the power grid operation trend; wherein the dispatching control plan is used to dispatch and control the target power grid or the target energy storage system to keep the target power grid frequency stable; The predicting of the power grid collected data by using a preset big data analysis technology to obtain the power grid operation trend includes: Performing data preprocessing on the power grid collected data to obtain preprocessed electrical data; Performing feature extraction on the preprocessed electrical data to obtain electrical extraction features; wherein the electrical extraction features; Performing time series data decomposition on the electrical extraction features to obtain multi-scale time series electrical features; wherein the multi-scale time series electrical features include load change features, voltage time domain features, current time domain features, power quality features, and power factor dynamic characteristics; Performing causal analysis on the electrical characteristics of the multi-scale time series by using a migration entropy algorithm to obtain a feature association network; wherein the feature association network is used to reveal the dynamic operation association between the electrical characteristics of each multi-scale time series in the target power grid; Using an adaptive kernel density estimation method, the state transition probability of the target power grid is estimated based on the feature association network to obtain a power grid state transition probability table; wherein the power grid state transition probability table is used to quantify the conversion probability of the target power grid between different working conditions, thereby performing a quantitative dynamic probability prediction of the future power grid operation state; Performing Markov chain prediction on the target power grid based on the power grid state transition probability table to obtain a power grid state prediction sequence; A self-organized criticality analysis method is used to perform nonlinear dynamic analysis on the grid state prediction sequence to obtain a grid operation trend; wherein the grid operation trend includes voltage stability, frequency stability, load change trend, and power quality change trend.

8. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images