Load processing method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a load processing method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring a topological structure of a power system; the edges for connecting the two nodes in the topological structure have corresponding connection weights; obtaining the load capacity of nodes at two ends of each edge in the topological structure, and updating the connection weight corresponding to each edge based on the load capacity; the updated connection weight is in negative correlation with the load capacity; dividing a plurality of nodes in the topological structure into a plurality of clusters based on the updated connection weight; wherein the connection weight corresponding to the edge for connecting the two nodes in the same cluster is greater than the connection weight corresponding to the edge for connecting the two nodes in different clusters; and in response to the situation that the power system does not meet the load demand, uniformly distributing the target load shedding amount to the nodes in the plurality of clusters, and performing load shedding on the nodes in the plurality of clusters. By adopting the method, secondary disturbance caused by low-frequency load shedding to the system can be reduced.

Description

Technical Field

[0001] The present application relates to the technical field of electric power, and particularly to a load processing method, device, computer device, computer-readable storage medium, and computer program product. Background Art

[0002] Under-frequency load shedding is a key measure to prevent the power system from experiencing frequency collapse and is an important part of the third line of defense in the power system. When the power system experiences a power deficit due to a fault, resulting in a frequency drop, the under-frequency load shedding device reduces the power deficit by quickly disconnecting some loads, thereby restoring the frequency stability of the system and avoiding large-scale power outages caused by frequency collapse.

[0003] Currently, in practical engineering applications, the formulation of under-frequency load shedding schemes mainly adopts an offline method, that is, the load shedding strategy is formulated in advance based on the historical data and preset conditions of the system. The parameter settings of this method include the number of action rounds, the frequency thresholds of each round, the action delay, and the load shedding amount of each round. Although the traditional under-frequency load shedding method has played an important role in ensuring the frequency stability of the system, its static nature may lead to unsafe operation of the system after load shedding or cause resource losses. Summary of the Invention

[0004] Based on this, it is necessary to provide a load processing method, device, computer device, computer-readable storage medium, and computer program product that can reduce the secondary disturbance brought by under-frequency load shedding to the system for the above technical problems.

[0005] In a first aspect, the present application provides a load processing method, including:

[0006] Obtain the topological structure of the power system; the edge connecting two nodes in the topological structure corresponds to a connection weight;

[0007] Obtain the load amounts of the nodes at both ends of each edge in the topological structure, and update the connection weights corresponding to each edge based on the load amounts; the updated connection weight is negatively correlated with the load amount;

[0008] Based on the updated connection weights, divide the multiple nodes in the topological structure into multiple clusters; among them, the connection weight of the edge connecting two nodes in the same cluster is greater than the connection weight of the edge connecting two nodes in different clusters;

[0009] In response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters.

[0010] In one embodiment, dividing multiple nodes in the topology into multiple clusters based on the updated connection weights includes:

[0011] Constructing a degree matrix based on the updated connection weights; the diagonal elements of the degree matrix are the sum of the connection weights corresponding to the edges connected by each node in the topology;

[0012] Performing spectral clustering on multiple nodes in the topology based on the degree matrix and a preset objective function to obtain multiple clusters; the objective function includes that the difference between the sums of the loads of all nodes in different clusters is less than a preset threshold.

[0013] In one embodiment, in response to the power system not meeting the load demand, evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in the multiple clusters includes:

[0014] In response to the power system not meeting the load demand, determining the load characteristics of the nodes in each of the multiple clusters; the load characteristics include at least one of the sensitivity to the change of the power system frequency and the load shedding cost.

[0015] Based on the load characteristics, determining the load shedding priorities of the nodes in each cluster; wherein, when the sensitivity of the first load to the change of the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load.

[0016] Evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in each of the multiple clusters according to their respective load shedding priorities.

[0017] In one embodiment, determining the load shedding priorities of the nodes in each cluster based on the load characteristics includes:

[0018] Dividing the loads of the nodes in each cluster into industrial loads, commercial loads, and residential loads;

[0019] Based on the load characteristics of the industrial loads, commercial loads, and residential loads of the nodes in each cluster, determining the load shedding priorities of the industrial loads, commercial loads, and residential loads of the nodes in each cluster.

[0020] In one embodiment, before in response to the power system not meeting the load demand, evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in the multiple clusters, it includes:

[0021] Obtain the inertia time constants of each operating unit in the power system, and determine the system equivalent inertia of the power system based on the inertia time constants;

[0022] Obtain the voltage frequencies of the buses where the respective units are located, and determine the power deficit of the power system based on the voltage frequencies, the inertia time constants, and the system equivalent inertia.

[0023] In one embodiment, the step of, in response to the power system not meeting the load demand, evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in the multiple clusters includes:

[0024] When the frequency of the power system is less than a first preset frequency, in response to the power system not meeting the load demand, evenly distribute a first proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters;

[0025] When the frequency of the power system is greater than the first preset frequency and less than a second preset frequency, in response to the power system not meeting the load demand, evenly distribute a second proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters; wherein, the first proportion is greater than the second proportion.

[0026] In a second aspect, the present application further provides a load processing device, including:

[0027] An acquisition module, configured to acquire the topological structure of a power system; an edge connecting two nodes in the topological structure corresponds to a connection weight;

[0028] An update module, configured to acquire the load amounts of the nodes at both ends of each edge in the topological structure, and respectively update the connection weights corresponding to the respective edges based on the load amounts; the updated connection weights are negatively correlated with the load amounts;

[0029] A partitioning module, configured to partition a plurality of nodes in the topological structure into a plurality of clusters based on the updated connection weights; wherein, the connection weight corresponding to an edge connecting two nodes in the same cluster is greater than the connection weight corresponding to an edge connecting two nodes in different clusters;

[0030] An excision module, configured to, in response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters.

[0031] In a third aspect, the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0032] In a fourth aspect, the present application further 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.

[0033] In a fifth aspect, the present application further provides a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the method described in any one of the above are implemented.

[0034] For the above load processing method, device, computer device, computer-readable storage medium and computer program product, by adjusting the connection weights between nodes according to the load information of nodes with the goal of load balancing based on the topological structure of the system, the connection weights between nodes with larger loads are relatively small and are dispersed in each cluster, realizing the initial load balance of each load partition. It can achieve the dispersed and balanced removal of the load amount that needs to be cut off during system failures in each load partition of the system, avoiding the secondary disturbance to the system caused by mainly concentrating on a certain area of the system for load removal during low-frequency load shedding in the related art, and improving the safety of system operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the following will briefly introduce the drawings required for use in the description of the embodiments of the present application or the related art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can be obtained based on these drawings.

[0036] Figure 1 It is a schematic flowchart of the load processing method in an embodiment;

[0037] Figure 2 It is a schematic flowchart of step S106 in an embodiment;

[0038] Figure 3 It is a schematic flowchart of step S108 in an embodiment;

[0039] Figure 4 It is a schematic diagram of the frequency characteristic curve of the load in an embodiment;

[0040] Figure 5 It is a schematic flowchart of step S1081 in an embodiment;

[0041] Figure 6Schematic diagram of the load processing method in another embodiment;

[0042] Figure 7 Schematic diagram of step S108 in another embodiment;

[0043] Figure 8 Structural block diagram of the load processing device in one embodiment;

[0044] Figure 9 Internal structure diagram of a computer device in one embodiment. Detailed implementation manners

[0045] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application 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 application, and are not used to limit the present application.

[0046] In one embodiment, as Figure 1 shown, a load processing method is provided. In this embodiment, it is exemplified that the method is applied to a terminal. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0047] Step S102: Obtain the topological structure of the power system; the edges connecting two nodes in the topological structure correspond to connection weights.

[0048] Among them, the connection weight can be determined by the power transmission capacity of the power line between two nodes. For example, the connection weight can be positively correlated with the power transmission capacity. In a possible implementation manner, the inalienable load areas in the system can be screened out first according to the importance of the load, and only the remaining load areas are partitioned. That is, only the nodes corresponding to the removable load areas are divided into multiple clusters. For example, according to the classification regulations of the load level in the national standard (GB50052-2009) and the actual requirements of system operation, the areas where first-level important loads such as hospitals, government agencies, and transportation hubs are located and other inalienable special load areas can be determined, and these areas are marked and screened out from the power transmission system with the lowest voltage level of the power grid.

[0049] Exemplarily, the topology of the lowest voltage level of the screened power transmission system can be modeled based on graph theory to obtain the topological structure G of the power system:

[0050] ;

[0051] ;

[0052] ;

[0053] Among them, is the set of nodes with the lowest voltage level in the power transmission system of the system; is the node with the lowest voltage level in the power transmission system ; is the total number of nodes with the lowest voltage level in the power transmission system; is the set of edges between the nodes with the lowest voltage level in the power transmission system, representing the power transmission lines in the system; are two nodes with the lowest voltage level in the power transmission system 、 the power line between.

[0054] The real-time power transmission capacity of the power line between two nodes with the lowest voltage level in the power transmission system can be used 、 to represent the connection weight of the edge to obtain the weight matrix of the edges in the system : :

[0055] ;

[0056] Among them, is a matrix. When there is no direct connection relationship between the nodes 、 with the lowest voltage level in the power transmission system or , the element in the matrix.

[0057] Step S104, obtain the load of the nodes at both ends of each edge in the topological structure, and update the connection weight corresponding to each edge based on the load; the updated connection weight is negatively correlated with the load.

[0058] Exemplarily, the connection weight between nodes can be adjusted by the real-time load of the nodes, so that the connection weight between nodes with larger loads is relatively smaller and it is easier to be divided into different clusters, realizing the initial balance of the load in each load partition. The updated weight matrix can include:

[0059] ;

[0060] Among them, is the new weight matrix; is the new weight element in the matrix. The updated connection weight The calculation formula of can include:

[0061] ;

[0062] Among them, is the node with the lowest voltage level of the power transmission system , and the real-time load. In another possible implementation, the updated connection weight can also be negatively correlated with the sum of the load amounts of the two end nodes.

[0063] Step S106: Based on the updated connection weights, divide the multiple nodes in the topology into multiple clusters; among them, the connection weight corresponding to the edge connecting two nodes in the same cluster is greater than the connection weight corresponding to the edge connecting two nodes in different clusters.

[0064] Exemplarily, based on the updated connection weights, the multiple nodes in the topology can be divided into multiple clusters through a community discovery algorithm. Among them, the community discovery algorithm is an important tool in graph analysis, used to reveal the potential structure in the network and can identify closely connected groups or regions in various fields. Common community discovery algorithms include modularity optimization algorithms, spectral clustering algorithms, label propagation algorithms, etc.

[0065] Step S108: In response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in multiple clusters.

[0066] Exemplarily, based on the power deficit of the system and the number of multiple clusters, the load shedding amount corresponding to each cluster in the multiple clusters can be determined. The calculation formula for the load shedding amount assigned to each cluster includes:

[0067] ;

[0068] Among them, is the evenly distributed load shedding amount in the ideal situation for each load partition.

[0069] Furthermore, to ensure that the target load shedding amount is evenly distributed as much as possible during the distribution process of the target load shedding amount, it can be distributed by combining the objective function and constraint conditions. The objective function can include:

[0070] ;

[0071] Among them, is the actually assigned load shedding amount for each load partition. The constraint conditions can include:

[0072] ;

[0073] ;

[0074] Among them, is the power deficit of the system, Pj is the total load of each load partition. In one possible implementation, after determining the objective function and constraints, the amount of load shedding allocated to each cluster can be optimized based on a linear programming algorithm. The linear programming algorithm is a mathematical optimization method aimed at solving the optimal solution through linear constraints and a linear objective function.

[0075] In the above load processing method, by adjusting the connection weights between nodes according to the topological structure of the system with the goal of load balancing based on the load information of the nodes, the connection weights between nodes with larger loads are relatively small and are dispersed in each cluster, achieving a preliminary balance of the loads in each load partition. It can achieve the dispersed and balanced shedding of the load to be shed when the system fails in each load partition of the system, avoiding the secondary disturbance to the system caused by mainly concentrating on a certain area of the system for load shedding during under-frequency load shedding in the related art, and improving the operational safety of the system.

[0076] In an exemplary embodiment, as Figure 2 shown, the above step S106 may include:

[0077] Step S1061, constructing a degree matrix based on the updated connection weights; the diagonal elements of the degree matrix are the sum of the connection weights corresponding to the edges connected by each node in the topological structure.

[0078] Among them, the spectral clustering algorithm is a kind of community discovery algorithm. The spectral clustering algorithm is a clustering method based on graph theory, suitable for dealing with non-convex shaped clustering problems. The core idea of the spectral clustering algorithm is to construct a similarity graph between data points, transform the data into a graph form, and use the eigeninformation of the Laplacian matrix of the graph for clustering. Specifically, first construct a similarity matrix, usually using a Gaussian kernel function to measure the similarity between data points; then, calculate the Laplacian matrix of the graph (for example, the standard Laplacian matrix or the normalized Laplacian matrix); then, solve the first k eigenvectors of the Laplacian matrix and use them as a low-dimensional representation; finally, use traditional clustering methods such as K-means to cluster the eigenvectors.

[0079] Exemplarily, the degree matrix can be a diagonal matrix, and the elements therein represent the degree of each node (the number of edges connected to each node). The elements in the degree matrix can be expressed as:

[0080] ;

[0081] The degree matrix can be expressed as:

[0082] ;

[0083] Among them, is the diagonal element in the degree matrix; is the set of nodes directly connected to node in the system.

[0084] Step S1062: Based on the degree matrix and a preset objective function, perform spectral clustering on multiple nodes in the topological structure to obtain multiple clusters; the objective function includes that the difference between the sums of the loads of all nodes in different clusters is less than a preset threshold.

[0085] Exemplarily, a Laplacian matrix can be constructed first :

[0086] ;

[0087] Then perform eigenvalue decomposition on the Laplacian matrix to solve its eigenvalues and eigenvectors. The solution formula for the eigenvectors includes:

[0088] ;

[0089] where is the eigenvalue of the Laplacian matrix ; is the eigenvector corresponding to the eigenvalue ; is the total number of eigenvalues and eigenvectors of the Laplacian matrix .

[0090] Then take the eigenvectors corresponding to the first smallest eigenvalues of the Laplacian matrix to form an eigenvector matrix:

[0091] ;

[0092] where is an eigenvector matrix with rows and columns. Each element in each column of the matrix represents the value of all nodes in this eigen-dimension. Each row element in the matrix forms the eigenvector of the corresponding node, that is, node corresponds to an -dimensional eigenvector , which can be regarded as the embedding representation of this node in the -dimensional space.

[0093] Then cluster the eigenvector matrix . For example, the K-means algorithm can be used for clustering to divide the nodes into clusters:

[0094] 。

[0095] To ensure load balancing during the clustering process, a load balancing constraint is introduced, and the goal is to make the total load of each cluster as close as possible. The objective function is obtained as follows:

[0096] ;

[0097] where, is the center of cluster ; is the total load of cluster ; is the total load of the system; is the weight coefficient, which is used to adjust the balance between load balancing and clustering quality.

[0098] Finally, the K-means algorithm can be used for clustering, and each cluster will correspond to a load partition of the system. Specifically, the way to use the K-means algorithm for clustering can include the following steps A1 to A4. Among them:

[0099] Step A1, initialize the clustering centers through K-means++, and intelligently select the initial clustering centers according to the distribution of feature vectors to reduce the local optimum problem during clustering convergence.

[0100] Among them, K-means++ is an improvement of the traditional K-means algorithm, aiming to improve the clustering quality and the convergence speed of the algorithm by improving the centroid initialization method. The traditional initialization method of K-means often randomly selects the centroid, which may cause the algorithm to fall into a local optimum solution, while K-means++ selects the initial centroid in a more intelligent way to improve this problem.

[0101] Step A2, calculate the distance between each node in the system and the current clustering centers. For node , calculate its Euclidean distance from each cluster center through the feature vector of the node. The calculation formula of the Euclidean distance includes:

[0102] ;

[0103] Assign each node to the cluster with the closest distance to it. After the assignment is completed, calculate the total load of each cluster:

[0104] ;

[0105] where, is the load of node assigned to cluster .

[0106] Step A3, after all nodes are assigned to the corresponding clusters, update the center of each cluster. The new cluster center is the average of the feature vectors of all nodes assigned to the cluster:

[0107] ;

[0108] where is the number of nodes assigned in cluster within.

[0109] Step A4, according to the objective function, perform continuous iteration until the clustering result is stable and the load balance within the cluster meets the requirements. At this time, the clustering process is completed, and each cluster obtained by clustering corresponds to a load partition. Finally, the transmission system with the lowest voltage level of the power grid is divided into load partitions.

[0110] In this embodiment, by combining the objective function to optimize the partition result, the load balance within each load partition can be further ensured.

[0111] In an exemplary embodiment, as Figure 3 shown, the above step S108 may include:

[0112] S1081, in response to the power system not meeting the load demand, determine the load characteristics of the nodes in each of the multiple clusters; the load characteristics include at least one of the sensitivity to the change in the power system frequency and the load shedding cost.

[0113] Exemplarily, the load characteristics of the nodes in each cluster can be determined based on the frequency regulation effect coefficient of the node load. The frequency regulation effect coefficient of the node load is used to measure the sensitivity of the node load to the change in the power system frequency. The determination method of the frequency regulation effect coefficient of the load may include:

[0114] First, determine the relational expression between the power system frequency and the active power of the load:

[0115] ;

[0116] where is the active power consumed by the load when the system frequency is ; is the active power consumed by the load when the system is at the rated frequency ; is the proportion of the load proportional to the nth power of the system frequency in ;

[0117] Then convert this relational expression into a per-unit value to obtain:

[0118] ;

[0119] Since the loads in the system that are proportional to the cube or higher power of the frequency change are very few, their effects can be ignored, and we can obtain:

[0120] ;

[0121] Differentiating it gives the calculation formula for the frequency regulation effect coefficient:

[0122] ;

[0123] where is the frequency regulation effect coefficient of the load. Please refer to Figure 4 , Figure 4 which is a schematic diagram of the frequency characteristic curve of the load in an embodiment. When the system frequency rises from to , , that is, compared with load 2, load 1 is more sensitive to frequency changes. Therefore, when the system frequency drops, load with a smaller

[0124] is preferentially disconnected. In this way, through the frequency regulation effect of the load itself, the frequency can be quickly restored, and it is beneficial to disconnect fewer loads. Exemplarily, the load shedding cost is the resource loss caused by disconnecting the load, which may include: economic cost, social cost, and restoration cost. The economic cost may include the direct resource loss caused by load shedding, which includes production stagnation, shutdown cost, business operation loss, etc.; the social cost mainly reflects the impact of load shedding on society and public interests, including the decline in the quality of life of residents, interruption of public services, etc.; the restoration cost refers to the time cost and labor cost required to restore the disconnected load, as well as the possible losses during the restoration process. The sum of the various losses caused by disconnecting various loads can be represented by the unit load shedding cost:

[0125] ;

[0126] where can be the unit load shedding cost of various loads; , , are respectively the unit economic cost, unit social cost, and unit restoration cost of disconnecting various loads; respectively represent industrial load, commercial load, and residential load. In another possible implementation, can also be the unit load shedding cost of each load.

[0127] S1082. Determine the load shedding priorities of the nodes in each cluster based on the load characteristics. Among them, when the sensitivity of the first load to the change in the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load.

[0128] Exemplarily, the comprehensive index of the load shedding sequence of each load partition can be determined according to the calculated frequency regulation effect coefficients and unit load shedding costs of various loads in each load partition. After normalizing the load frequency regulation effect coefficient and unit load shedding cost data of each unit, the comprehensive index can be expressed as:

[0129] ;

[0130] Among them, is the comprehensive index of the load shedding sequence in load partition j; and are the normalized load frequency regulation effect coefficients and unit load shedding costs of various loads in load partition j; and are the weight coefficients of the load frequency regulation effect coefficient and unit load shedding cost, which can be set according to actual needs. After calculating the comprehensive index of each type of load in load partition j, the corresponding load is shed in ascending order.

[0131] S1083. Evenly distribute the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in each cluster among the multiple clusters according to their respective load shedding priorities.

[0132] Further, as Figure 5 shown, the above step S1081 may include:

[0133] Step S811. Divide the loads of the nodes in each cluster into industrial loads, commercial loads, and residential loads.

[0134] Step S812. Determine the load shedding priorities of the industrial loads, commercial loads, and residential loads of the nodes in each cluster based on the load characteristics of the industrial loads, commercial loads, and residential loads of the nodes in each cluster.

[0135] Among them, the industrial load has the highest proportion in the system, the economic loss of power supply interruption is large, but the impact on residents' lives is small; the commercial load has a relatively low proportion in the system, the power supply interruption has a certain impact on residents' lives, but the economic loss is small; the residential load has a relatively high proportion in the power system, the power supply interruption has a great impact on residents' lives, but the economic loss is the smallest.

[0136] Further, for the load partition j, the total load within the area can be counted. Industrial load Commercial load Residential load , and the load shedding priority is sorted according to industrial load, commercial load or residential load. The load frequency regulation effect coefficient of the three types of loads in each load partition can be calculated, and taking the load frequency regulation effect coefficient as one of the indicators, the order of load shedding for each type of load within each load partition is determined. It is also possible to use the unit load shedding cost of each type of load as an indicator to determine the order of load shedding for each type of load within each load partition.

[0137] In this embodiment, through the multi-factor load shedding priority evaluation, it avoids the insecurity and losses caused by simply dividing the load shedding order mainly based on the load importance level in the related art. Further, by classifying the loads in each load partition into different types according to the load type and differentiating the load shedding priority for different types of loads within each load partition, the economy of safety control is improved.

[0138] In an exemplary embodiment, as Figure 6 shown, the above load processing method may further include:

[0139] Step S1071: Obtain the inertia time constant of each unit put into operation in the power system, and determine the system equivalent inertia of the power system based on the inertia time constant.

[0140] Exemplarily, when a system fault occurs, the number of units put into operation in the system at the current moment can be counted online. By obtaining the inertia time constant of each unit, the system equivalent inertia of the system can be calculated. The calculation formula of the system equivalent inertia may include:

[0141] ;

[0142] wherein, is the equivalent inertia time constant of the system; q is the number of units put into operation in the system in real time; is the inertia time constant of the operating unit j.

[0143] Step S1072: Obtain the voltage frequency of the bus where each unit is located, and determine the power deficit of the power system based on the voltage frequency, inertia time constant, and system equivalent inertia.

[0144] Exemplarily, the real-time inertia center frequency of the system can be determined first based on the inertia time constants of each operating unit in the system, the voltage frequency of the bus where it is located, and the system equivalent inertia of the system; then, based on the inertia center frequency, the system equivalent inertia, and the rated power, the power deficit of the system can be determined. The calculation formula for the power deficit can include:

[0145] ;

[0146] wherein, is the real-time inertia center frequency of the system, and f N is the rated power. The calculation formula for the inertia center frequency can include:

[0147] ;

[0148] wherein, is the real-time voltage frequency of the bus where the operating unit j is located.

[0149] In an exemplary embodiment, the load shedding rounds of the system low-frequency load shedding can be divided into an online basic action round and an offline special action round, and the specific parameters of the basic action round and the special action round can be set. As Figure 7 shown, the above step S108 may further include:

[0150] Step S1084, when the frequency of the power system is less than the first preset frequency, in response to the power system not meeting the load demand, evenly distribute the first proportion of the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in multiple clusters.

[0151] Step S1085, when the frequency of the power system is greater than the first preset frequency and less than the second preset frequency, in response to the power system not meeting the load demand, evenly distribute the second proportion of the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in multiple clusters; wherein, the first proportion is greater than the second proportion.

[0152] Further, the above step S1084 may include:

[0153] When the frequency of the power system is less than the first preset frequency, in response to the power system not meeting the load demand, determine the load characteristics of the nodes in each of the multiple clusters; the load characteristics include at least one of the sensitivity to the change in the power system frequency and the load shedding cost; based on the load characteristics, determine the load shedding priority of the nodes in each cluster; wherein, when the sensitivity of the first load to the change in the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; evenly distribute the first proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in each of the multiple clusters according to their respective load shedding priorities.

[0154] The above step S1085 may include:

[0155] When the frequency of the power system is greater than the first preset frequency and less than the second preset frequency, in response to the power system not meeting the load demand, determine the load characteristics of the nodes in each of the multiple clusters; the load characteristics include at least one of the sensitivity to the change in the power system frequency and the load shedding cost; based on the load characteristics, determine the load shedding priority of the nodes in each cluster; wherein, when the sensitivity of the first load to the change in the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; evenly distribute the second proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in each of the multiple clusters according to their respective load shedding priorities.

[0156] Exemplarily, the starting frequency of the basic operation round of the system low-frequency load shedding online can be set to 49.2 Hertz (Hz), and a system frequency change rate criterion is added:

[0157] ;

[0158] The load control amount of the basic operation round of the system low-frequency load shedding is 90% of the load amount that the system can shed; when the system frequency meets the starting condition, the power deficit of the system can be calculated ; evenly distribute the target load amount to be shed among the various load partitions in the system, that is, among the various clusters, and each load partition cuts the corresponding load in the order of the corresponding load shedding priority; after the load shedding action is completed, observe the system frequency recovery situation in real time. If the system frequency further drops, calculate the new power deficit of the system, and repeat the process of load distribution and shedding until the system frequency recovers.

[0159] In a possible implementation, the equivalent inertia of the system can be calculated first, and then it is determined whether the system frequency is less than 49.2 Hz and whether the system frequency change rate is less than 0; if not, the current basic round ends; if so, the system power deficit is calculated, load shedding is started, and it is determined again whether the system frequency is less than 49.2 Hz and whether the system frequency change rate is less than 0; if not, the current basic round ends; if so, the system power deficit is calculated again, and load shedding is started.

[0160] Exemplarily, the special action rounds of the system low-frequency load shedding offline can be set to two rounds, and the starting frequencies are set to 49.3 Hz and 49.5 Hz respectively. The load control amount of each round of the special action round is 5% of the total load that the system can shed; after the basic action round of the system is completed, after a delay of 15 s, the system frequency recovery situation is observed. If the system frequency hovers between 49.2 Hz and 49.5 Hz for a long time, the special action round is started.

[0161] In summary, in the above load processing method, by adjusting the connection weights between nodes according to the topological structure of the system with the goal of load balancing and in accordance with the load information of the nodes, the connection weights between nodes with larger loads are relatively smaller and are dispersed in each cluster, achieving preliminary load balancing of each load partition. It can achieve the dispersed and balanced shedding of the load to be shed when the system fails in each load partition of the system, avoiding the secondary disturbance to the system caused by mainly concentrating on a certain area of the system for load shedding in the related technology of low-frequency load shedding, and improving the operating safety of the system. Further, by combining the objective function, the load balance within each load partition can be further ensured. Further, by evaluating the load shedding priority of multiple factors, it avoids the unsafe situation and losses caused by simply dividing the order of load shedding mainly based on the load importance level in the related technology.

[0162] It should be understood that although the steps in the flowcharts involved in the above embodiments are shown in sequence according to the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0163] Based on the same inventive concept, an embodiment of the present application further provides a load processing device for implementing the load processing method involved above. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the load processing device provided below can refer to the limitations on the load processing method in the above text, and will not be repeated here.

[0164] In an exemplary embodiment, as Figure 8 shown, a load processing device 300 is provided, including: an acquisition module 301, an update module 302, a partitioning module 303, and a shedding module 304, where:

[0165] The acquisition module 301 is configured to acquire the topological structure of the power system; the edges connecting two nodes in the topological structure correspond to connection weights.

[0166] The update module 302 is configured to acquire the load amounts of the nodes at both ends of each edge in the topological structure, and respectively update the connection weights corresponding to each edge based on the load amounts; the updated connection weights are negatively correlated with the load amounts.

[0167] The partitioning module 303 is configured to partition a plurality of nodes in the topological structure into a plurality of clusters based on the updated connection weights; among them, the connection weight corresponding to the edge connecting two nodes in the same cluster is greater than the connection weight corresponding to the edge connecting two nodes in different clusters.

[0168] The shedding module 304 is configured to, in response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the plurality of clusters, and perform load shedding on the nodes in the plurality of clusters.

[0169] In an exemplary embodiment, the above partitioning module 303 may further be configured to:

[0170] Construct a degree matrix based on the updated connection weights; the diagonal elements of the degree matrix are the sum of the connection weights corresponding to the edges connected by each node in the topological structure;

[0171] Perform spectral clustering on a plurality of nodes in the topological structure based on the degree matrix and a preset objective function to obtain a plurality of clusters; the objective function includes that the difference between the sum of the load amounts of all nodes in different clusters is less than a preset threshold.

[0172] In an exemplary embodiment, the above shedding module 304 may further be configured to:

[0173] In response to the power system not meeting the load demand, determine the load characteristics of the nodes in each of the plurality of clusters; the load characteristics include at least one of the sensitivity to the change of the power system frequency and the load shedding cost;

[0174] Based on the load characteristics, determine the load shedding priority of the nodes in each cluster; wherein, when the sensitivity of the first load to the change in the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load;

[0175] Evenly distribute the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in each of the multiple clusters according to their respective load shedding priorities.

[0176] In an exemplary embodiment, the above-mentioned determining the load shedding priority of the nodes in each cluster based on the load characteristics may include:

[0177] Divide the loads of the nodes in each cluster into industrial loads, commercial loads, and residential loads;

[0178] Based on the load characteristics of the industrial loads, commercial loads, and residential loads of the nodes in each cluster, determine the load shedding priorities of the industrial loads, commercial loads, and residential loads of the nodes in each cluster.

[0179] In an exemplary embodiment, the above-mentioned load processing device 300 may further include a determining module for:

[0180] Obtain the inertia time constants of the various units put into operation in the power system, and determine the system equivalent inertia of the power system based on the inertia time constants;

[0181] Obtain the voltage frequencies of the buses where the various units are located, and determine the power deficit of the power system based on the voltage frequencies, inertia time constants, and system equivalent inertia.

[0182] In an exemplary embodiment, the above-mentioned cutting module 304 may further be used for:

[0183] When the frequency of the power system is less than the first preset frequency, in response to the power system not meeting the load demand, evenly distribute the first proportion of the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in the multiple clusters;

[0184] When the frequency of the power system is greater than the first preset frequency and less than the second preset frequency, in response to the power system not meeting the load demand, evenly distribute the second proportion of the target load shedding amount to the nodes in multiple clusters, and perform load shedding on the nodes in the multiple clusters; wherein, the first proportion is greater than the second proportion.

[0185] Each module in the above load processing device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form so that the processor can call and execute the operations corresponding to each of the above modules.

[0186] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structural diagram can be as Figure 9 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device 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 data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication 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, it implements a load processing method.

[0187] Those skilled in the art can understand that Figure 9 the structure shown in

[0188] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0189] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps in the above method embodiments.

[0190] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, it implements the steps in the above method embodiments.

[0191] Those of ordinary skill in the art can understand that all or part of the processes in the methods of 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 embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0192] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in the present application.

[0193] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A load processing method, characterized in that, The method includes: Obtain the topological structure of the power system; an edge connecting two nodes in the topological structure corresponds to a connection weight; Obtain the load of the nodes at both ends of each edge in the topological structure, and update the connection weight corresponding to each edge based on the load; the updated connection weight is negatively correlated with the load; Based on the updated connection weight, divide multiple nodes in the topological structure into multiple clusters; wherein, the connection weight corresponding to an edge connecting two nodes in the same cluster is greater than the connection weight corresponding to an edge connecting two nodes in different clusters; In response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters.

2. The method according to claim 1, characterized in that, The dividing the multiple nodes in the topological structure into multiple clusters based on the updated connection weight includes: Construct a degree matrix based on the updated connection weight; the diagonal element of the degree matrix is the sum of the connection weights corresponding to the edges connected by each node in the topological structure; Perform spectral clustering on the multiple nodes in the topological structure based on the degree matrix and a preset objective function to obtain multiple clusters; the objective function includes that the difference between the sums of the loads of all nodes in different clusters is less than a preset threshold.

3. The method according to claim 1, wherein The evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in the multiple clusters in response to the power system not meeting the load demand includes: In response to the power system not meeting the load demand, determine the load characteristics of the nodes in each cluster of the multiple clusters; the load characteristics include at least one of the sensitivity to the change in the power system frequency and the load shedding cost. Based on the load characteristics, determine the load shedding priority of the nodes in each cluster; wherein, when the sensitivity of the first load to the change in the power system frequency is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load; when the load shedding cost of the first load is lower than that of the second load, the load shedding priority of the first load is higher than that of the second load. Evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in each cluster of the multiple clusters according to their respective load shedding priorities.

4. The method according to claim 3, wherein The determining the load shedding priority of the nodes in each cluster based on the load characteristics includes: Divide the load of the nodes in each cluster into industrial load, commercial load, and residential load; Based on the load characteristics of the industrial load, commercial load, and residential load of the nodes in each cluster, determine the load shedding priorities of the industrial load, commercial load, and residential load of the nodes in each cluster.

5. The method according to claim 1, wherein Before the evenly distributing the target load shedding amount to the nodes in the multiple clusters and performing load shedding on the nodes in the multiple clusters in response to the power system not meeting the load demand, it includes: Obtain the inertia time constant of each operating unit in the power system, and determine the system equivalent inertia of the power system based on the inertia time constant; Obtain the voltage frequencies of the buses where the respective generating units are located, and determine the power deficit of the power system based on the voltage frequencies, the inertia time constant, and the equivalent inertia of the system.

6. The method according to claim 1, wherein In response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters, including: When the frequency of the power system is less than the first preset frequency, in response to the power system not meeting the load demand, evenly distribute the first proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters; When the frequency of the power system is greater than the first preset frequency and less than the second preset frequency, in response to the power system not meeting the load demand, evenly distribute the second proportion of the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters; wherein, the first proportion is greater than the second proportion.

7. A load processing device, characterized in that, The device includes: An acquisition module, configured to acquire the topological structure of the power system; the edges connecting two nodes in the topological structure correspond to connection weights; An update module, configured to acquire the load amounts of the nodes at both ends of each edge in the topological structure, and respectively update the connection weights corresponding to the respective edges based on the load amounts; the updated connection weights are negatively correlated with the load amounts; A partitioning module, configured to partition the multiple nodes in the topological structure into multiple clusters based on the updated connection weights; wherein, the connection weights of the edges connecting two nodes in the same cluster are greater than the connection weights of the edges connecting two nodes in different clusters; A shedding module, configured to, in response to the power system not meeting the load demand, evenly distribute the target load shedding amount to the nodes in the multiple clusters, and perform load shedding on the nodes in the multiple clusters.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, 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 the processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images