A power grid energy storage regulation method and system considering source-storage-load collaborative interaction

Abstract

The present application relates to the technical field of power grid energy storage regulation, and particularly relates to a power grid energy storage regulation method and system considering source-storage-load collaborative interaction. The present application obtains power transmission node information of a power grid system and power consumption information of each terminal user, divides the power grid system into multiple virtual energy storage clusters according to the power transmission node information, analyzes load fluctuation characteristics of all terminal user power consumption data of each virtual energy storage cluster, determines a first importance degree of power demand of each load area, obtains a second importance degree of each load area under an abnormal state according to load power changes before and after power generation failure, and configures power supply priority of each load area according to the first importance degree and the second importance degree of each load area, and supplies power until power generation failure is repaired. The technical scheme can realize collaborative adjustment of energy storage stations, power supply stations and load ends after power generation failure of the power grid system, and effectively improves the accuracy of power distribution.

Description

Technical Field

[0001] This invention relates to the technical field of power grid energy storage regulation, and specifically to a power grid energy storage regulation method and system that considers the coordinated interaction between source, storage and load. Background Technology

[0002] The existing power grid system comprises a three-tiered structure: source-storage-load. Source refers to the origin of electricity production, i.e., power plants; including traditional thermal power, hydropower, and nuclear power, as well as rapidly growing intermittent renewable energy sources such as wind power and photovoltaics. Storage refers to energy storage stations, such as electrochemical energy storage (lithium-ion batteries, flow batteries, etc.), pumped hydro storage, and compressed air energy storage; it is the core hub for achieving coordinated interaction and a flexibly adjustable resource. Load refers to the electricity consumption side, i.e., the electricity demand side; including industrial, commercial, and residential electricity consumption, as well as adjustable loads such as electric vehicle charging stations, smart air conditioners, and interruptible production lines.

[0003] In the event of power outages such as generator failures, existing power grids must implement load shedding to maintain stable islanded operation. However, existing grid energy storage control schemes may misallocate limited emergency energy resources, misreporting the actual flexible loads that can be reduced, while cutting off the rigid loads that truly require stable power supply. This reduces the utilization efficiency of limited energy resources and may even lead to the paralysis of critical social functions, severely restricting grid resilience and the power restoration process. Summary of the Invention

[0004] To address the technical problem of improving the accuracy of stored energy allocation in power grid systems under power generation failures, the present invention aims to provide a power grid energy storage regulation method and system that considers source-storage-load synergistic interaction. The specific technical solution adopted is as follows: In a first aspect, embodiments of the present invention provide a grid energy storage regulation method considering source-storage-load coordinated interaction, the method comprising: To obtain information on the transmission nodes of the power grid system and the electricity consumption information of each end user, the power grid system is equipped with power plants, energy storage stations and power supply stations; The power grid system is divided into multiple virtual energy storage clusters based on the transmission node information, so that each virtual energy storage cluster can be powered by the corresponding energy storage station after a power generation failure. Based on the electricity consumption information of all end users in each virtual energy storage cluster, load fluctuation characteristics are analyzed to determine the primary importance of each load area to electricity demand in each virtual energy storage cluster. Based on the load power changes of each load area before and after the power generation failure, the second importance of each load area to the non-reducible power is obtained; Based on the first and second importance of each load area, the power supply priority for each load area is configured and power is supplied until the power generation failure is repaired.

[0005] In one optional embodiment, the power grid system is divided into multiple virtual energy storage clusters based on transmission node information, including: The node impedance is calculated based on the distribution of power grid nodes characterized by the transmission node information, so as to obtain the self impedance and mutual impedance of each node in the power grid system. Based on the self-impedance and mutual impedance of each node in the power grid system, the electrical node impedance, which characterizes the degree of voltage coupling between each transmission node, is obtained. Sort all electrical node impedances and merge them based on the sum of the combined power requirements of multiple adjacent nodes; After all adjacent nodes have been merged, the merged result is determined as the partitioning result of multiple virtual energy storage clusters.

[0006] In one optional embodiment, the electrical node impedance, characterizing the degree of voltage coupling between transmission nodes, is obtained based on the self-impedance and mutual impedance of each node in the power grid system, including: The cumulative impedance of the two transmission nodes is obtained by summing the self-impedances between them. The electrical node impedances of the two transmission nodes are obtained by using the absolute value of the difference between the cumulative impedance and the mutual impedance of the two transmission nodes.

[0007] In one optional embodiment, load fluctuation characteristic analysis is performed based on the electricity consumption information of all end users in each virtual energy storage cluster to determine the primary importance of each load area to electricity demand in each virtual energy storage cluster, including: Load curves are extracted from the electricity consumption information of each virtual energy storage cluster, and similarity calculations are performed on each load curve to classify each load curve into the load mode of the corresponding virtual energy storage cluster. The average value of all load curves of the same load mode in each virtual energy storage cluster is processed to obtain the target curve that characterizes the load characteristics of each load mode. The variance and mean of the load value of the target curve for each load mode are calculated to obtain the load variance and load mean square of the corresponding load mode. The ratio of the squared mean load to the variance of the load for each load mode is determined as the first importance of each load region in the corresponding virtual energy storage cluster.

[0008] In one optional embodiment, a similarity calculation is performed on each load curve to classify each load curve to the load pattern of the corresponding virtual energy storage cluster, including: Dynamic time warping distance calculation is performed on the electricity consumption time series characteristics of any two load curves to obtain a distance matrix characterizing the similarity between all load curves; The distance matrix is ​​input into a preset K-Medoids clustering model for iterative sample partitioning, and the load pattern corresponding to each load curve in the virtual energy storage cluster is obtained based on the partitioning results.

[0009] In one alternative embodiment, a second importance of each load region to the indivisibility of electricity is obtained based on the load power change of each load region before and after the generation failure, including: Based on the current load of each load area after a power generation failure and the minimum daily load before the power generation failure, the load margin index of the corresponding load area is obtained. The electricity load of each load area at the corresponding time before the power generation failure is determined as the synchronous demand load of the corresponding load area. The load reduction level of each load area is obtained based on the concurrent demand load of each load area and the current load after the power generation failure. Based on the load margin index and load reduction degree of each load area, the second importance of each load area to the inability to reduce electricity is obtained.

[0010] In one alternative embodiment, the second importance is negatively correlated with the load margin index and positively correlated with the degree of load reduction.

[0011] In one optional embodiment, power supply priority is configured for each load area according to its first and second importance, and power is supplied until the power generation fault is repaired, including: Based on the product of the first and second importance of each load area, the power supply priority of the corresponding load area during the power generation failure period is obtained; Based on the power demand, actual power, and descending order of all power supply priorities for each load area, the corresponding power supply is configured for each load area. Within a preset rolling time window, the load areas corresponding to the descending order of the results are divided into guaranteed areas and load reduction areas according to the preset ranking, and the power supply is configured accordingly until the power generation fault is repaired.

[0012] In one optional embodiment, obtaining the electricity consumption information of each terminal user includes: The electricity load of each terminal user is collected according to a preset frequency to obtain the load curve of the corresponding terminal user. The load curve is the curve of electricity load changing with electricity consumption time. The load curve and user tag of each end user are stored accordingly to obtain the electricity consumption information of each end user.

[0013] Secondly, embodiments of the present invention also provide a grid energy storage regulation system considering source-storage-load synergistic interaction, applied to any regulation method in the first aspect, the system comprising: The acquisition module is used to acquire information on the transmission nodes of the power grid system and the electricity consumption information of each end user. The power grid system is equipped with power plants, energy storage stations and power supply stations. The partitioning module is used to divide the power grid system into multiple virtual energy storage clusters based on the transmission node information, so that each virtual energy storage cluster can be powered by the corresponding energy storage station after a power generation failure. The determination module is used to perform load fluctuation characteristic analysis based on the electricity consumption information of all end users in each virtual energy storage cluster, so as to determine the first importance of each load area to the electricity demand in each virtual energy storage cluster. The module is used to obtain the second importance of each load area to non-reducible power consumption based on the load power changes of each load area before and after the power generation failure; The power distribution module is used to configure the power supply priority of each load area according to the first and second importance of each load area and supply power until the power generation failure is repaired.

[0014] The present invention has the following beneficial effects: The technical solution of this invention obtains the transmission node information and electricity consumption information of each end user in the power grid system. Based on the transmission node information, the power grid system is divided into multiple virtual energy storage clusters, so that each virtual energy storage cluster can be independently powered by its corresponding energy storage station when a power generation failure occurs, forming a distributed emergency power grid structure. The load fluctuation characteristics of all end users in each virtual energy storage cluster are analyzed. By calculating the fluctuation characteristics of each load curve, the first importance of each load area to the power demand is determined to characterize the dependence of the area on power in long-term operation. Then, based on the load power change before and after the power generation failure, the second importance of each load area under abnormal conditions is obtained to reflect the non-reduction of the load. Based on the first and second importance of each load area, the power supply priority of each load area is configured and power is supplied until the power generation failure is repaired. This technical solution enables coordinated adjustment of energy storage stations, power supply stations, and loads after a power generation failure occurs in the power grid system. Under limited energy storage conditions, it ensures priority power supply to critical loads and effectively improves the accuracy and timeliness of power distribution. Thus, it greatly improves the utilization efficiency of limited energy storage in extreme fault scenarios and enhances the adaptability of the smart grid to cope with emergencies. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating a grid energy storage regulation method considering source-storage-load synergistic interaction, provided as an embodiment of the present invention; Figure 2 A flowchart for calculating the first importance degree is provided in one embodiment of the present invention; Figure 3 A flowchart for calculating the second importance degree is provided in one embodiment of the present invention; Figure 4 This is a schematic diagram of a power grid energy storage control system that considers source-storage-load synergistic interaction, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a grid energy storage regulation method and system considering source-storage-load synergistic interaction proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] Currently, when a power grid system experiences a power generation failure, priority is determined by manually preset load type labels. These labels are static and cannot distinguish the inherent differences in electricity consumption behavior among different end-users within the same load category. Their true importance is not accurately quantified, leading to crude decisions regarding the power supply from energy storage stations. Furthermore, static prioritization completely ignores the temporal characteristics of the loads and their real-time operating status at the time of the failure, failing to reflect the actual electricity demand of a region at different times of the day. This results in the misallocation of limited emergency energy resources. The following section, with reference to the accompanying drawings, will specifically describe the specific scheme of a power grid energy storage control method and system that considers source-storage-load synergistic interaction, provided by this invention.

[0020] Please see Figure 1 , Figure 1This is a flowchart illustrating a grid energy storage regulation method considering source-storage-load coordinated interaction, provided as an embodiment of the present invention. This method can be applied to the operation of an energy storage dispatch terminal in a grid. The energy storage dispatch terminal can be a server or a computer, as long as it can run this method; no specific limitations are imposed here. The regulation method includes: S11. Obtain information on the transmission nodes of the power grid system and the electricity consumption information of each end user. The power grid system is equipped with power plants, energy storage stations and power supply stations.

[0021] Specifically, transmission node information represents the line connections between power plants, energy storage stations, and power supply stations (or distribution stations) in the power grid system. Power plants are used to transmit electricity to the power grid; energy storage stations store electrical energy through batteries and balance the supply and demand fluctuations of the power grid; power supply stations are nodes for power transmission and distribution, and can implement power transformation and power distribution to end users. The three stations work together to achieve a source-storage-load coordinated closed loop. Transmission node information can be obtained based on the power grid system's GIS (Geographic Information System) and PMS (Production Management System). Electricity consumption information represents the electricity consumption of each end user and can be derived based on user tags and recorded historical electricity consumption data.

[0022] Taking the acquisition of electricity consumption information for each end user as an example, the electricity load of each end user can be collected at a preset frequency to obtain the corresponding load curve. The load curve is a curve showing the change of electricity load over time. The load curve and user tag of each end user are stored accordingly to obtain the electricity consumption information of each end user. The user tag can be derived based on the attributes of the end user, such as classifying each end user into residential communities, hospitals, commercial complexes, industrial plants, etc., and marking the end user based on the corresponding user tag.

[0023] It should be noted that differentiated collection strategies need to be developed for different data types during the data acquisition phase. For example, when collecting electricity load data from each end user, real-time measurement data is obtained from SCADA (Supervisory Control And Data Acquisition) and smart meters. The system collects user-side electricity data at a preset frequency of once every 15 minutes to record the load consumption of each area. Daily electricity consumption data for all load areas (such as residential communities and hospitals) within the distribution network coverage area over the past 1-3 years is recorded at one data point every 15 minutes, forming a load curve of 96 data points per day, corresponding to the time periods of 0:00-0:15, 0:15-0:30, ..., 23:45-24:00, stored in a time-series database format. User tags are extracted from the user profile system and labeled with specific types based on user industry attributes and electricity consumption characteristics, such as hospitals, emergency command centers, commercial buildings, and residential households. User tags are updated in real time as user electricity consumption attributes change.

[0024] At this point, the transmission node information and electricity consumption information have been obtained, and we proceed to step S12.

[0025] S12. Divide the power grid system into multiple virtual energy storage clusters based on the transmission node information, so that each virtual energy storage cluster can be powered by the corresponding energy storage station after a power generation failure.

[0026] Specifically, after a large-scale power outage in the power grid system, the synchronous restoration of power supply across the entire grid is difficult and time-consuming. The optimal control strategy is to quickly form multiple isolated islands, i.e., virtual energy storage clusters, capable of autonomously supplying power to corresponding energy storage stations. This localized autonomous power supply ensures the operational needs of core loads. Transmission impedance analysis of energy storage stations can be performed based on transmission node information to minimize or keep the transmission impedance of individual energy storage stations within acceptable limits after a power generation failure in the power grid system, thus deriving each virtual energy storage cluster. Each virtual energy storage cluster includes multiple load areas, each corresponding to at least one end-user, which can be electricity-consuming entities such as hospitals, schools, and shopping malls.

[0027] When implementing virtual energy storage cluster partitioning, the distribution network needs to be abstracted into a mathematical graph model G, G = (V, E), meaning the mathematical graph model G includes a node set V and an edge set E. The node set V contains all bus nodes, load nodes, distributed generation nodes, and energy storage nodes in the distribution network, and needs to be further classified according to their functional attributes. Bus nodes are divided into PQ nodes (fixed voltage amplitude, known active / reactive power) and PV nodes (fixed active power and voltage amplitude, adjustable reactive power). Load nodes are classified according to importance into primary load nodes (e.g., hospital emergency departments, emergency command centers), secondary load nodes (e.g., commercial buildings, ordinary residential communities), and tertiary load nodes (e.g., landscape lighting, non-essential entertainment facilities). The edge set E represents the lines and switching equipment in the distribution network. Based on impedance calculations of each node, it is determined which transmission nodes are more reasonable to merge, and the merging result of multiple transmission nodes is determined as the corresponding single virtual energy storage cluster.

[0028] For example, step S12 includes sub-steps S12-1 to S12-4, which are described in detail below: S12-1. Calculate the node impedance based on the power grid node distribution represented by the transmission node information to obtain the self-impedance and mutual impedance of each node in the power grid system. Taking the partitioning of a virtual energy storage cluster based on a certain community as an example, each transmission node is regarded as an independent community, i.e., the initial community set is... , Given the total number of transmission nodes, we calculate the power supply and demand for each initial community, specifically including the energy storage capacity of distributed generation sources and the power of primary loads. This leads to the determination of the self-impedance of each transmission node and the mutual impedance between two transmission nodes. All self-impedances and mutual impedances can be represented based on the node impedance matrix of the power grid system, denoted as [the matrix is ​​missing in the original text]. .

[0029] S12-2. Based on the self-impedance and mutual impedance of each node in the power grid system, obtain the electrical node impedance, which characterizes the degree of voltage coupling between transmission nodes. The electrical node impedance characterizes the power transmission loss between two transmission nodes. When dividing a virtual energy storage cluster, transmission impedance should be optimized. The corresponding electrical node impedance can be derived based on the self-impedance and mutual impedance of two adjacent transmission nodes. The electrical node impedance can quantify the degree of electrical interconnection between nodes, providing a basis for the division of virtual energy storage clusters.

[0030] For example, the cumulative impedance of two transmission nodes can be obtained by summing their self-impedances; further, the electrical node impedance of the corresponding two transmission nodes can be obtained by the absolute value of the difference between the cumulative impedance and the mutual impedance of the two transmission nodes. The electrical node impedance is denoted as... , ,in, Let i be the self-impedance of transmission node i. Let J be the self-impedance of transmission node j. Let Z represent the mutual impedance between transmission node i and transmission node j. All impedance parameters can be obtained from the node impedance matrix Z. It should be noted that the electrical node impedance... The smaller the value, the higher the voltage coupling between transmission node i and transmission node j, the lower the power transmission loss. When both are assigned to the same virtual energy storage cluster, the grid voltage stability is stronger, the power regulation difficulty is lower, and the transmission loss in islanded power supply is smaller. Conversely, the larger the value, the lower the electrical node impedance. The larger the value, the greater the transmission loss between the two transmission nodes, making it unsuitable as a single virtual energy storage cluster.

[0031] It can be understood that the nodal impedance matrix Z is a fundamental matrix in power system analysis, and it is the inverse of the nodal admittance matrix Y, i.e. The node admittance matrix Y can be formed using standard methods based on the topological connection relationship of the power grid and the electrical parameters (such as resistance, reactance, and susceptance to ground) of each branch. Its diagonal elements are the sum of the admittances of all branches connected to node i, and the off-diagonal elements are the negative values ​​of the admittances of the branches between transmission node i and transmission node j.

[0032] S12-3. Sort all electrical node impedances and merge them based on the combined power demand of multiple adjacent nodes. Sort all electrical node impedances from smallest to largest, and prioritize merging the two transmission nodes with the smallest electrical node impedances. After merging, recalculate the total power supply and demand of the region to which the merged transmission nodes belong, and determine whether the power balance constraint is met. If the constraint is met, retain the merge result; if not, skip the combination and select the next group of adjacent transmission nodes with the smallest impedances for merging.

[0033] S12-4. After all adjacent nodes have been merged, the merged result is determined as the partitioning result of multiple virtual energy storage clusters. When all remaining transmission nodes cannot be merged with other transmission nodes, that is, when the unmerged communities cannot be merged with other communities to form a new community that satisfies the power balance constraint, the merging algorithm terminates. The community set obtained at this time is the virtual energy storage cluster partitioning scheme, and each virtual energy storage cluster contains at least one load area.

[0034] At this point, the grid system has been divided and multiple virtual energy storage clusters have been obtained. Proceed to step S13.

[0035] S13. Analyze the load fluctuation characteristics based on the electricity consumption information of all end users in each virtual energy storage cluster to determine the primary importance of each load area to electricity demand in each virtual energy storage cluster.

[0036] Specifically, in the current load management of the power grid system, especially when facing extreme faults requiring islanding operation and load control, it is impossible to distinguish the differences in the actual electricity consumption behavior of different end users within the same category of load. For example, a data center with stable energy consumption and a shopping mall with huge peak-valley differences are classified as the same type of commercial building, and their true importance for power supply is not accurately quantified. Secondly, static user tags completely ignore the time-series electricity consumption characteristics of the load itself and cannot reflect its real impact and value on the power grid at different times of the day. This leads to overly crude decision-making basis when formulating precise load reduction strategies, which may misreport flexible loads that can actually be peak-shaving while cutting off rigid loads that truly need stable power supply. This reduces the utilization efficiency of limited emergency energy and increases the risk of power grid restoration.

[0037] Load curves and user tags for corresponding end users can be extracted from electricity consumption information. Similarity calculations are performed based on the load fluctuation characteristics represented by the load curves, and further verification is conducted using user tags to derive the load pattern of each load area within each virtual energy storage cluster. This load pattern helps determine whether a corresponding load area requires priority emergency power supply. For example, commercial areas typically operate from 10:00 AM to 9:00 PM, with a load peak between 12:00 PM and 2:00 PM. Based on the load changes during operating hours, this area is a shopping mall, and its priority for power demand in the event of a power generation failure is lower than that of a hospital area. Therefore, by analyzing the load fluctuation characteristics of all end users across all virtual energy storage clusters, a corresponding priority can be determined. The priority characterizes the importance of power demand for each load area within the corresponding virtual energy storage cluster. A higher priority indicates greater social losses due to power outages, requiring priority power supply to such areas; conversely, a lower priority indicates a lower demand for power supply.

[0038] For example, please refer to Figure 2 , Figure 2 The flowchart for calculating the first importance is shown below. Step S13 includes sub-steps S13-1 to S13-4, which are described in detail below: S13-1. Extract load curves from the electricity consumption information of each virtual energy storage cluster, and perform similarity calculations on each load curve to classify them into the corresponding load patterns of the virtual energy storage cluster. Load curves characterize the load fluctuations of the corresponding end-users during electricity consumption. Based on the fluctuation characteristics, similarity calculations are performed. Load curves with similarity greater than a preset value are considered to have the same load pattern. Further analysis is conducted on these load curves to determine their respective load patterns. Through load pattern analysis, it is possible to identify attributes such as commercial buildings and medical units within the load area, and to analyze their electricity consumption pattern characteristics.

[0039] In practical applications, different end-users have varying electricity consumption scales, but their curve shapes may be similar. For example, the electricity load in multiple load areas may be high during the day and low at night. When comparing the similarity of load curves, comparing only numerical values ​​(mean, peak) can be affected by the scale of electricity consumption. Therefore, when determining the load pattern of load curves, Dynamic Time Warping (DTW) can be calculated for the electricity consumption time series characteristics of any two load curves to obtain a distance matrix characterizing the similarity between all load curves. The distance matrix is ​​then input into a preset K-Medoids clustering model for iterative sample partitioning, and the load pattern corresponding to each load curve in the virtual energy storage cluster is derived based on the partitioning results.

[0040] For example, when determining load patterns based on the K-Medoids clustering algorithm, a range of K values ​​(e.g., K from 2 to 10) can be set for testing to determine the optimal number of clusters; the number of clustering iterations is set to 50; and dynamic time-warped distance is used as the distance metric to solve the matching problem for different time series lengths or phase shifts. Based on the load curve of each end user, the daily load variation characteristics can be determined, and the load curve of each load region is taken as a time series sample. The distance matrix between samples is calculated to quantify the morphological differences between any two load curves. Using the distance matrix as input, the K-Medoids algorithm iteratively divides the samples for each K value, ensuring the highest morphological similarity among samples within the same cluster and the greatest morphological difference between different clusters. Finally, by calculating the silhouette coefficient of each cluster under different K values, the K value corresponding to the maximum silhouette coefficient (e.g., K=5) is selected as the optimal number of clusters; at this point, each cluster represents a different type of load pattern.

[0041] S13-2. For each virtual energy storage cluster, average all load curves under the same load mode to obtain target curves characterizing the load characteristics of each load mode. For each cluster, calculate the average of all load curves within that cluster to obtain the typical daily load curve for that mode, which is the target curve representing the load characteristics of the corresponding load mode. Where d is the cluster number, such as d=1 representing the weekday pattern and d=2 representing the weekend pattern; each data point of the target curve is denoted as t is the average load of all samples in this cluster at time t. This represents the load characteristics under this load mode.

[0042] Analysis reveals that the shape of the load curve (such as its volatility and peak-to-valley difference) reflects the functional attributes and electricity consumption characteristics of the load area. For example, a commercial area with drastically fluctuating power consumption should be considered differently from a hospital with stable power consumption. Quantifying the shape of the load curve allows for a more scientific assessment of its priority in fault scenarios.

[0043] S13-3. Calculate the variance and mean of the load values ​​for the target curve of each load mode to obtain the load variance and load mean squared for the corresponding load mode. Square each load value in the target curve; simultaneously, obtain the square of the average value of the target curve, denoted as the load mean squared. The squared form is used here to amplify differences and facilitate comparison. Similarly, obtain the variance of all data on the target curve, denoted as the load variance, which represents the degree of dispersion of the load values ​​around the average value. The larger the variance, the more severe the load fluctuation. It should be noted that, taking a daily load curve as an example, this curve contains 96 load values; the mean and variance can be calculated based on these 96 load values.

[0044] S13-4. The ratio of the squared load mean to the load variance for each load mode is determined as the first importance of each load region in the corresponding virtual energy storage cluster. The squared load mean can be used as the numerator, and the load variance as the denominator (to avoid division by zero, when the load variance is 0, the denominator can be set to a preset minimum positive number, such as 0.001). The ratio between these two values ​​represents the first importance of the load region m. The larger the value, the flatter the load curve (such as hospitals and data centers), which are usually rigid loads that require stable power supply at all times and have high basic importance; the smaller the value, the larger the peak-to-valley difference and the more volatile the load curve (such as commercial buildings), which are usually load areas with high demand response potential and have lower basic importance.

[0045] Based on the above operations, for each load region, its corresponding first importance can be obtained; this value does not change with the real-time operating status, and is stored in the load feature library of the multimodal resilience data pool as an inherent attribute of the load region, and then proceeds to step S14.

[0046] S14. Based on the load power changes of each load area before and after the power generation failure, obtain the second importance of each load area for the inability to reduce electrical energy.

[0047] Specifically, key information such as real-time population density in the load area and instantaneous operating status of critical facilities (such as hospital operating rooms) makes it difficult to determine power supply priority and the actual social and economic impact of power generation failures. This makes it impossible to achieve accurate and adaptive allocation of limited power, thereby reducing grid resilience and power supply recovery efficiency.

[0048] In a power system, during normal power demand periods, power is directly transmitted from power plants to end users, with surplus energy stored in energy storage stations. During power outages, power plants cease supplying electricity, and energy storage stations then supply power to end users. This also applies to areas with high electricity demand where demand exceeds supply during certain periods; in such cases, energy storage stations draw power stored during off-peak hours in those areas. Analysis reveals that electricity load inherently possesses an irreducible characteristic, meaning that the load of some end users changes relatively little compared to historical averages.

[0049] Therefore, for any load area, it is necessary to obtain the daily minimum load before the fault and the current load collected in real time after the fault. It should be noted that the current load, given the existence of a power system fault in the current area, fluctuates in real time due to factors such as equipment start-up and shutdown, personnel activity, and environmental changes. It encompasses the impact of all normal and abnormal factors, and in the case of a power generation fault, it is directly affected by events such as power outages, voltage drops, and manual equipment shutdowns. Based on the daily minimum load and the current load, the load power change of the corresponding load area can be determined. Based on the change, the second importance of each load area to the non-reducible power consumption can be further analyzed. This second importance determines whether load reduction is applicable to each load area, thus prioritizing the power supply to that area during subsequent power distribution.

[0050] For example, please refer to Figure 3 , Figure 3 The flowchart for calculating the second importance is shown below. Step S14 includes sub-steps S14-1 to S14-4, which are described in detail below: S14-1. Based on the current load of each load area after a power generation failure and the daily minimum load before the failure, obtain the load margin index for the corresponding load area. The current load can be obtained from the power supplied to the area by the corresponding energy storage station after the power generation failure. The daily minimum load is obtained based on the minimum value of the daily load curve; there is only one daily minimum load per day. This is achieved through the formula... Obtain the load margin index of load region m at the current time t. m is the sequence number of the corresponding load area. The current load at time t. The minimum daily load is represented by a denominator incremented by one to prevent the result from being meaningless due to a denominator of 0. This ensures that the resulting calculation error is within a controllable range. `norm()` represents the normalization of the calculation result. In this embodiment, `norm` normalization can specifically be, for example, maximum / minimum value normalization, without limitation. For the normalized ratio, the smaller the load margin index value, the lower the power consumption level in the area, indicating the presence of a large amount of essential equipment that cannot be shut down, such as hospital life support systems or server cooling systems. This indicates a lower load margin, higher rigidity, and stronger non-reduction capability, making it more fundamentally important to consider in subsequent allocation. Conversely, a larger load margin index value indicates greater potential for reduction and lower fundamental importance.

[0051] S14-2. Determine the electricity load of each load area at the corresponding time before the power generation failure as the synchronous demand load of the corresponding load area. Simultaneously, under the actual pressure of a power supply failure, it is necessary to identify which loads are the most resilient and the most difficult to reduce. Based on the load curve of the area's daily electricity consumption obtained in the above steps, determine the historical average load power at the current time after the failure relative to the same historical time, representing the typical load power at that moment, i.e., the synchronous demand load.

[0052] S14-3. Based on the concurrent demand load of each load region and the current load after a power generation failure, obtain the load reduction degree of the corresponding load region. This can be based on the formula... To obtain the load reduction level of load area m at the current time t. , The current load at time t. This represents the concurrent demand load at the current time t. Increasing the denominator by one prevents the result from being meaningless if the denominator is zero, thus keeping the calculation error within a controllable range. The load reduction degree quantifies the extent to which the load in this area has been forced to decrease due to a fault. If the ratio is close to 0, it indicates that the current load is close to its typical value, and the load reduction degree is relatively low (if the current load exceeds the concurrent demand load, the load reduction degree takes the minimum value of 0). Conversely, if the ratio is large, closer to 1, it indicates that the electricity load has been significantly reduced, and its current real-time rigidity is low.

[0053] S14-4. Based on the load margin index and load reduction degree of each load area, obtain the second importance of each load area for the inability to reduce electrical energy. The second importance of each load area can be obtained through weighted calculation based on the load margin index and load reduction degree. For example, based on the relationship between different load margin indices, load reduction degrees, and second importance, weight coefficients can be set for the load margin index and load reduction degree respectively, and a weighted calculation can be performed to obtain the second importance of each load area.

[0054] For example, to obtain the second importance, it is necessary to comprehensively consider the load margin index and load reduction degree of each load area. According to the analysis in S14-1, the lower the load margin index value, the more important it is (negative correlation); while according to S14-3, the higher the load reduction degree value, the more important it is (positive correlation). In a specific embodiment, to achieve this correlation, the load margin index can first be converted into an index positively correlated with importance, for example, through... The calculation is then performed. Subsequently, the converted index is averaged with the load reduction rate and normalized. An example formula is shown below: Obtain the second importance of load region m corresponding to the current time t. , This indicates normalization. This yields a comprehensive index of the non-reducibility of the region at the current moment. Regarding the second importance... The larger the value (i.e., the closer it is to 1), the higher the priority of power supply protection in the current region at the current moment; conversely, the lower the priority of power supply protection after a power generation failure.

[0055] At this point, the first and second importance of each load region have been obtained, and we proceed to step S15.

[0056] S15. Based on the first and second importance of each load area, configure the power supply priority for each load area and supply power until the power generation failure is repaired.

[0057] Specifically, the power supply priority for each load area can be determined by a weighted calculation based on the first and second importance. Alternatively, the power supply priority can be derived from the normalized calculation of the ratio or product of the first and second importance. This priority determines whether a load area requires priority power supply and allocates power to meet its full-load operation. A higher power supply priority indicates that the load area may be an important location such as a hospital, requiring maximum power supply. Conversely, a lower priority indicates that the load area may be a shopping mall, allowing for appropriate load reduction to meet its lighting and other power needs.

[0058] For example, step S15 includes sub-steps S15-1 to S15-3, which are described in detail below: S15-1. Based on the product of the first and second importance of each load area, obtain the power supply priority of the corresponding load area during the power generation failure period. According to the formula... Obtain the power supply priority of load area m at the current time t. The larger the value, the higher the importance of the area at the current moment, and the higher the priority for allocating power from the energy storage station to the area. The above power supply priority can be obtained for each load area.

[0059] S15-2. Based on the demand power, actual power, and descending order of all power supply priorities for each load area, configure the corresponding power supply for each load area. For a future time window (e.g., 4 hours), based on the SOC (State of Charge) of the distributed energy storage station and the random forest algorithm, predict the total available power and the demand power of each area at each future time t under islanded operation. The demand power can be estimated from historical load data for the same period. In this process, the power supply allocated by the energy storage station to a certain area is the decision variable for system control, representing the planned output power allocated by the system at time t; while the actual real-time load power collected is a state variable, reflecting the actual power consumption of end users after a power generation failure.

[0060] To achieve optimal utilization of energy storage resources, the goal is to maximize the comprehensive value of all load areas throughout the entire scheduling cycle. This involves multiplying the power supply priority of each area at each time point by the demand power and the actual power, then summing the results to obtain the total value coefficient F. The maximum value of F is then used as the optimization objective. Simultaneously, constraints are set within the corresponding time windows to ensure that the total power allocated to all areas does not exceed the available power of the energy storage station, and that the power allocated to each area is not less than zero and does not exceed the maximum load power of that area at the same historical time. Through iterative calculations, the total value coefficient F is maximized, thus obtaining the optimal power allocation scheme for each load area at future times. This corresponds to the optimal power allocation schedule, clarifying the power that should be allocated to each load area at each time point.

[0061] S15-3. Within a preset rolling time window, the load areas corresponding to the descending sort results are divided into guaranteed areas and load reduction areas according to the preset ranking, and the power supply is configured accordingly until the power generation fault is repaired. When the preset ranking M=1 / 3 is manually set, after sorting according to the dynamic power supply priority from high to low, the product of M and the total number of areas in the current virtual energy storage cluster yields several areas. Several areas are designated as guaranteed areas, requiring that the power allocated to these areas at any time is not lower than the absolute minimum load power in the historical data of these areas, in order to ensure the continuous operation of the basic load; the remaining areas are designated as load reduction areas, and power supply is preferentially reduced when there is a power shortage. The system further generates a load reduction command based on the difference between the typical load power and the optimal allocated power of each load area at the same historical time. This command is used to indicate the amount of load power to be reduced at time t. When the load reduction command is 0, it means that the power supply can be fully supplied; when it is positive, it means that the corresponding power needs to be reduced; if the load reduction command is equal to the typical load value, the power supply to this area is completely cut off at the current time. The power system periodically initiates rolling optimization calculations with a fixed rolling cycle of 15 minutes. Before each optimization, the energy storage status, available power, and power supply priority are updated, the total value coefficient F is recalculated, and the energy storage power is allocated to achieve dynamic optimal scheduling of power during fault repair until the power generation fault repair of the power system is completed.

[0062] Based on the same technical concept as the control method, this embodiment of the invention also provides a grid energy storage control system that considers source-storage-load coordinated interaction, applicable to any of the control methods described above. Please refer to... Figure 4 , Figure 4 This is a schematic diagram of the control system, which includes an acquisition module 401, a division module 402, a determination module 403, an acquisition module 404, and a power distribution module 405.

[0063] The acquisition module 401 acquires transmission node information and electricity consumption information of each end user in the power grid system, which includes power plants, energy storage stations, and power supply stations. The partitioning module 402 divides the power grid system into multiple virtual energy storage clusters based on the transmission node information, ensuring that each virtual energy storage cluster is supplied with power through its corresponding energy storage station after a power generation failure. The determination module 403 performs load fluctuation characteristic analysis based on the electricity consumption information of all end users in each virtual energy storage cluster to determine the primary importance of each load area's electricity demand within each virtual energy storage cluster. The acquisition module 404 obtains the secondary importance of each load area's non-reducible electricity demand based on the load power changes before and after a power generation failure. The distribution module 405 configures the power supply priority for each load area according to its primary and secondary importance and supplies power until the power generation failure is repaired.

[0064] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0065] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A grid energy storage regulation method considering source-storage-load synergistic interaction, characterized in that, The method includes: The system acquires information on transmission nodes and electricity consumption information of each end user in the power grid system, which is equipped with power plants, energy storage stations and power supply stations. The power grid system is divided into multiple virtual energy storage clusters based on the power transmission node information, so that each virtual energy storage cluster can be powered by the corresponding energy storage station after a power generation failure. Based on the electricity consumption information of all end users in each virtual energy storage cluster, load fluctuation characteristics are analyzed to determine the primary importance of each load area to electricity demand in each virtual energy storage cluster. Based on the load power changes of each load area before and after the power generation failure, the second importance of each load area to the non-reducible power is obtained; Based on the first and second importance of each load area, the power supply priority for each load area is configured and power is supplied until the power generation failure is repaired. The process of analyzing load fluctuation characteristics based on the electricity consumption information of all end users in each virtual energy storage cluster to determine the primary importance of each load area to electricity demand in each virtual energy storage cluster includes: Load curves are extracted from the electricity consumption information of each virtual energy storage cluster, and similarity calculations are performed on each load curve to classify each load curve into the load mode of the corresponding virtual energy storage cluster. The average value of all load curves of the same load mode in each virtual energy storage cluster is processed to obtain the target curve that characterizes the load characteristics of each load mode. The variance and mean of the load value of the target curve for each load mode are calculated to obtain the load variance and load mean square of the corresponding load mode. The ratio of the squared mean load to the variance of the load for each load mode is determined as the first importance of each load region in the corresponding virtual energy storage cluster. The method of obtaining the second importance of each load region for non-reducible power consumption based on the load power change of each load region before and after the power generation failure includes: Based on the current load of each load area after a power generation failure and the minimum daily load before the power generation failure, the load margin index of the corresponding load area is obtained. The electricity load of each load area at the corresponding time before the power generation failure is determined as the synchronous demand load of the corresponding load area. The load reduction level of each load area is obtained based on the concurrent demand load of each load area and the current load after the power generation failure. Based on the load margin index and load reduction degree of each load area, the second importance of each load area to the inability to reduce electricity is obtained.

2. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 1, characterized in that, The step of dividing the power grid system into multiple virtual energy storage clusters based on the transmission node information includes: Based on the power grid node distribution characterized by the transmission node information, the node impedance is calculated to obtain the self impedance and mutual impedance of each node in the power grid system. Based on the self-impedance and mutual impedance of each node in the power grid system, the electrical node impedance, which characterizes the degree of voltage coupling between each transmission node, is obtained. Sort all electrical node impedances and merge them based on the sum of the combined power requirements of multiple adjacent nodes; After all adjacent nodes have been merged, the merged result is determined as the partitioning result of multiple virtual energy storage clusters.

3. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 2, characterized in that, The process of obtaining the electrical node impedance, which characterizes the degree of voltage coupling between transmission nodes, based on the self-impedance and mutual impedance of each node in the power grid system, includes: The cumulative impedance of the two transmission nodes is obtained by summing the self-impedances between them. The electrical node impedance of the two transmission nodes is obtained by using the absolute value of the difference between the cumulative impedance and the mutual impedance of the two transmission nodes.

4. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 1, characterized in that, Similarity calculations are performed on each load curve to classify each load curve into the load pattern of the corresponding virtual energy storage cluster, including: Dynamic time warping distance calculation is performed on the electricity consumption time series characteristics of any two load curves to obtain a distance matrix characterizing the similarity between all load curves; The distance matrix is ​​input into a preset K-Medoids clustering model for iterative sample partitioning, and the load pattern corresponding to each load curve in the virtual energy storage cluster is obtained based on the partitioning results.

5. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 1, characterized in that, The second importance is negatively correlated with the load margin index, and positively correlated with the degree of load reduction.

6. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 1, characterized in that, The process of configuring power supply priorities for each load area based on its first and second importance, and supplying power to that load area until the power generation fault is repaired, includes: Based on the product of the first and second importance of each load area, the power supply priority of the corresponding load area during the power generation failure period is obtained; Based on the power demand, actual power, and descending order of all power supply priorities for each load area, the corresponding power supply is configured for each load area. Within a preset rolling time window, the load areas corresponding to the descending sort results are divided into guaranteed areas and load reduction areas according to preset positions, and power supply is configured accordingly until the power generation fault is repaired.

7. The grid energy storage regulation method considering source-storage-load synergistic interaction according to claim 1, characterized in that, Obtain electricity consumption information from each end user, including: The electricity load of each terminal user is collected according to a preset frequency to obtain the load curve of the corresponding terminal user. The load curve is the curve of electricity load changing with electricity consumption time. The load curve and user tag of each end user are stored accordingly to obtain the electricity consumption information of each end user.

8. A grid energy storage control system considering source-storage-load coordinated interaction, characterized in that, The system, applied to the control method according to any one of claims 1-7, comprises: The acquisition module is used to acquire information on the transmission nodes of the power grid system and the electricity consumption information of each end user. The power grid system is equipped with power plants, energy storage stations and power supply stations. The partitioning module is used to divide the power grid system into multiple virtual energy storage clusters according to the power transmission node information, so that each virtual energy storage cluster can be powered by the corresponding energy storage station after a power generation failure. The determination module is used to perform load fluctuation characteristic analysis based on the electricity consumption information of all end users in each virtual energy storage cluster, so as to determine the first importance of each load area to the electricity demand in each virtual energy storage cluster. The module is used to obtain the second importance of each load area to non-reducible power consumption based on the load power changes of each load area before and after the power generation failure; The power distribution module is used to configure the power supply priority of each load area according to the first and second importance of each load area and supply power until the power generation failure is repaired.

Images