A power grid partitioning method

Abstract

The application discloses a power grid partition method, which comprises the following steps: acquiring a topological graph of an actual power grid structure; performing random partition on the topological graph of the actual power grid structure according to a set total partition quantity to obtain a particle swarm; calculating a load loss risk value generated by the particle swarm when the particle swarm suffers from external attack in an extreme case; and continuously optimizing the particle swarm by using an improved particle swarm algorithm, with the load loss risk value generated by the particle swarm when the particle swarm suffers from external attack in the extreme case as a target, so as to obtain a power grid partition scheme of the actual power grid structure when the load loss risk value generated by the particle swarm when the particle swarm suffers from external attack in the extreme case is reduced and tends to be stable. The scheme can improve the limit survival capability of the power grid by partitioning the power grid, so that the power grid can reduce the power cut range and the load loss when suffering from external attack, and the continuous power supply of important users can be ensured.

Description

Technical Field

[0001] This application relates to the field of power grid partitioning technology, specifically to a power grid partitioning method. Background Technology

[0002] Currently, power grid planning, construction, and operation are based on peacetime standards. Therefore, there is a lack of effective contingency plans and risk management measures for extreme situations such as external attacks. A targeted external attack on the power grid could lead to prolonged and widespread power outages, severely impacting users' electricity supply and causing inconvenience to their lives. Summary of the Invention

[0003] In view of this, this application provides a power grid partitioning method to solve the problem that a targeted external attack on the power grid may lead to long-term, large-scale power outages, which seriously affect users' electricity consumption and cause inconvenience to users' lives.

[0004] To achieve the above objectives, the following solution is proposed:

[0005] A power grid zoning method, comprising:

[0006] Obtain the topology map of the actual power grid structure; set the total number of partitions, and randomly partition the topology map of the actual power grid structure according to the total number of partitions to obtain a particle swarm; calculate the load loss risk value generated by the particle swarm under extreme conditions when it suffers an external attack; optimize the particle swarm using an improved particle swarm algorithm with the goal of reducing and stabilizing the load loss risk value generated by the particle swarm under extreme conditions when it suffers an external attack; when the load loss risk value generated by the particle swarm under extreme conditions when it suffers an external attack decreases and stabilizes, the power grid partitioning scheme of the actual power grid structure is obtained.

[0007] Preferably, obtaining the topology of the actual power grid structure includes: obtaining the power of each power plant node in the actual power grid structure; dividing each power plant node into upstream power supply nodes and downstream power supply nodes according to the power of each power plant node; obtaining the electrical distance between each power plant node as the line weight between each node, and constructing a power grid connectivity graph using the line weight; building a weighted adjacency matrix in the power grid connectivity graph to obtain a weighted connectivity graph; obtaining the shortest path between the upstream power supply node and the downstream power supply node using the shortest path algorithm based on the weighted connectivity graph; and marking the weighted connectivity graph using the shortest path between the upstream power supply node and the downstream power supply node to obtain the topology of the actual power grid structure.

[0008] Preferably, the step of randomly partitioning the topology map of the actual power grid structure according to the total number of partitions to obtain a particle swarm includes: dividing the topology map of the actual power grid structure into several target power supply areas; randomly partitioning the several target power supply areas in the topology map of the actual power grid structure according to the total number of partitions to obtain a first sub-particle swarm; randomly partitioning all power plant nodes in the actual power grid structure according to the total number of partitions to obtain a second sub-particle swarm; and combining the first sub-particle swarm and the second sub-particle swarm according to a preset ratio to obtain the particle swarm.

[0009] Preferably, the topology of the actual power grid structure includes upstream and downstream nodes; dividing the topology of the actual power grid structure into several target power supply areas includes: pre-setting a power supply radius; in the topology of the actual power grid structure, for each upstream node: taking the upstream node as the center, all isolated downstream nodes whose shortest path to the upstream node is less than the power supply radius are taken as target downstream nodes; wherein, the isolated downstream node is a downstream node that is not divided into a power supply area with any upstream node; dividing the upstream node and the target downstream node into a power supply area, thereby obtaining several power supply areas; determining whether the sum of the power supply capacity of all upstream nodes in the topology of the actual power grid structure is equal to the sum of the load capacity of all downstream nodes; if so, determining whether the sum of the power supply capacity of the upstream node and the load capacity of all downstream nodes in each power supply area is equal; if so, taking several power supply areas as the target power supply areas.

[0010] Preferably, the method further includes: when it is determined that the sum of the power supply capacity of the uplink node and the load capacity of all downlink nodes in each power supply area is not equal, adjusting the power supply radius and returning to the step of taking the uplink node as the center and selecting all isolated downlink nodes whose shortest path to the uplink node is less than the power supply radius as the target downlink node.

[0011] Preferably, the method further includes: when it is determined that the sum of the power supply capacity of all upstream nodes in the topology diagram of the actual power grid structure is not equal to the sum of the load capacity of all downstream nodes, then it is determined whether there are isolated downstream nodes in the topology diagram of the actual power grid structure. If so, the power supply radius is adjusted, and the method of taking the upstream node as the center and selecting all isolated downstream nodes whose shortest path to the upstream node is less than the power supply radius as the target downstream nodes is executed.

[0012] Preferably, it further includes: when it is determined that there is no isolated downstream node in the topology diagram of the actual power grid structure, then several of the power supply areas are taken as the target power supply areas.

[0013] Preferably, calculating the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks includes: constructing a fault set using the faults that occur when the particle swarm is subjected to external attacks under extreme conditions; obtaining the probability of the particle swarm experiencing each fault in the fault set; calculating the load loss amount generated by the particle swarm when experiencing each fault in the fault set; for each fault in the fault set, multiplying the probability of the fault occurrence by the load loss amount generated when the fault occurs to obtain a sub-value of the load loss risk generated by the particle swarm when subjected to external attacks; and summing the load loss risk sub-values ​​of each fault in the fault set to obtain the load loss risk value.

[0014] Preferably, calculating the load loss generated by the particle swarm when each fault in the fault set occurs includes: calculating the load loss generated by the particle swarm when each fault in the fault set occurs using a transient simulation method.

[0015] Preferably, the step of optimizing the particle swarm using an improved particle swarm algorithm, with the goal of reducing and stabilizing the risk value of load loss generated by the particle swarm under extreme conditions when subjected to external attacks, includes: updating the velocity of each particle in the particle swarm using the improved particle swarm algorithm; updating the position of each particle in the particle swarm using the updated velocity to obtain a new particle swarm; determining whether the risk value of load loss generated by the new particle swarm under extreme conditions when subjected to external attacks has decreased and stabilized; if yes, the optimization ends; if no, the step of updating the velocity of each particle in the particle swarm using the improved particle swarm algorithm is returned to execution.

[0016] As can be seen from the above technical solution, this application obtains the topology map of the actual power grid structure, sets the total number of partitions in the actual power grid structure, and then randomly partitions the topology map of the actual power grid structure according to the total number of partitions to obtain a particle swarm. It calculates the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks. Then, using an improved particle swarm algorithm, it optimizes the particle swarm with the goal of reducing and stabilizing the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks. When the load loss risk value generated by the particle swarm under extreme conditions decreases and stabilizes, a power grid partitioning scheme for the actual power grid structure is obtained. This scheme utilizes the actual power grid structure for partitioning, constructs a particle swarm, and then continuously optimizes the particle swarm using an improved particle swarm algorithm, aiming to reduce and stabilize the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks, thereby obtaining the optimal power grid partitioning scheme for the actual power grid structure. This improves the extreme survivability of the power grid, enabling the power grid to reduce the scope of power outages and load losses when subjected to external attacks, ensuring continuous power supply to important users. Attached Figure Description

[0017] Figure 1 An optional flowchart of the power grid partitioning method provided in the embodiments of this application;

[0018] Figure 2 A flowchart illustrating the method for dividing the target power supply area as provided in this application embodiment. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] The existing power system security defense system is mainly based on the three-level security and stability standards of the "Guidelines for Power System Security and Stability". my country has built a mature and effective three-line defense: relay protection, overload tripping and load stabilization devices, and low-frequency and low-voltage out-of-step disconnection devices. When the power grid suffers from conventional external attack risks, it is also divided into zones. However, the existing power grid zoning methods mainly prevent conventional power grid security risks and are insufficient to cope with external attacks under extreme circumstances. A targeted external attack on the power grid could lead to prolonged and widespread blackouts. The losses caused by a large-scale synchronous power grid blackout are often greater than those of a small-scale synchronous power grid, and in severe cases, may even affect national political stability and the stable operation of the social economy.

[0021] To address the aforementioned issues, this embodiment provides a power grid partitioning method, which will be discussed below. Figure 1 The power grid zoning method of this application is described, such as... Figure 1 As shown, the method includes:

[0022] S1: Obtain the topology diagram of the actual power grid structure.

[0023] When a real power grid needs to be partitioned for security defense, a topology map of the actual power grid structure can be obtained first, and then the actual power grid can be partitioned using the topology map.

[0024] S2: Set the total number of partitions, and randomly partition the topology of the actual power grid structure according to the total number of partitions to obtain the particle swarm.

[0025] The number of partitions can be preset according to the actual power grid coverage, coverage area and desired prevention effect. Then, the topology map of the actual power grid structure can be randomly partitioned according to this number of partitions to obtain the particle swarm.

[0026] S3: Calculate the risk of load loss when the particle swarm is attacked by external forces under extreme conditions.

[0027] This step calculates the load loss risk value generated when the particle swarm is attacked under extreme conditions, which is equivalent to calculating the load loss risk value generated when the actual power grid structure is attacked under extreme conditions.

[0028] S4: Optimize the particle swarm algorithm by using an improved particle swarm algorithm, with the goal of reducing and stabilizing the risk of load loss caused by the particle swarm when it suffers external attacks in extreme cases.

[0029] This application proposes an improved particle swarm optimization algorithm, which optimizes the obtained particle swarm by continuously improving the partitioning scheme of the particle swarm. The goal is to continuously reduce the load loss risk value generated by the partitioning scheme of the actual power grid structure under extreme conditions when it is subjected to external attacks, and finally reach a stable state.

[0030] S5: When the load loss risk value generated by the particle swarm under extreme conditions decreases and tends to stabilize, the power grid partitioning scheme of the actual power grid structure is obtained.

[0031] Specifically, during the continuous optimization process, if, after each optimization, the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks decreases and tends to stabilize, or decreases to the point of remaining basically unchanged, then the optimization process can be judged to be over. The zoning scheme obtained at this time is the final power grid zoning scheme of the actual power grid structure to be obtained.

[0032] Alternatively, if the load loss risk values ​​obtained are all equal within a certain number of optimization attempts, the optimization process can be considered complete. In this case, the resulting zoning scheme is the final power grid zoning scheme for the actual power grid structure.

[0033] As can be seen from the above technical solution, this application obtains a topology map of the actual power grid structure, randomly partitions the topology map according to a set total number of partitions to obtain a particle swarm optimization (PSO) algorithm, calculates the load loss risk value generated by the PSO under extreme conditions when subjected to external attacks, and then continuously optimizes the PSO algorithm with the goal of reducing and stabilizing the load loss risk value generated by the PSO under extreme conditions when subjected to external attacks. When the load loss risk value generated by the PSO under extreme conditions decreases and stabilizes, a power grid partitioning scheme for the actual power grid structure is obtained. This scheme improves the extreme survivability of the power grid by partitioning the power grid, reducing the scope of power outages and load losses when the power grid is subjected to external attacks, and ensuring continuous power supply to important users.

[0034] Specifically, embodiments of this application provide a method for obtaining a topology diagram of an actual power grid structure, which may include:

[0035] S11: Obtain the power of each power plant node in the actual power grid structure.

[0036] In a real power grid, there are multiple power plants, each with its own power capacity. For example, hydropower, thermal power, and renewable energy plants can all serve as substation nodes with power sources of 500kV and 220kV, providing active power support to other power plant nodes in the grid. The grid-connected power capacity of the i-th 500kV or 220kV power plant node can be expressed as... Where A represents the total number of 500kV or 220kV substation nodes in the power grid to which the substation node is connected, κ i,a Let P represent the output coefficient of the a-th power plant. The determination of this output coefficient requires comprehensive consideration of the adjustment margin of the three lines of defense mentioned above. S,a This represents the rated total capacity of the a-th power plant.

[0037] In reality, local users corresponding to the power grid act as load substations, consuming a portion of the grid-connected power. This application can consider only the power transmission of 500kV or 220kV substation nodes within the main grid structure of the actual power grid. Therefore, the output power of the substation nodes can be expressed as P. i =P Gi -P Li , where P LiP represents the power consumed by the downstream distribution network of the i-th 500kV or 220kV substation node. i It can reflect the power supply or load characteristics of plant nodes.

[0038] S12: Based on the power of each plant node, each plant node is divided into upstream nodes and downstream nodes.

[0039] In the above formula P i =P Gi -P Li In the middle, when P i When P > 0, it indicates that the node outputs power, and it is defined as an uploading node (power supply node). i When <0, it indicates that the node is inputting power internally, and it is defined as a power-sending node (load node).

[0040] S13: Obtain the electrical distance between each plant node as the line weight between each node, and use the line weight to construct the power grid connectivity diagram.

[0041] Because the actual power grid structure is quite complex, this step requires abstracting the actual power grid structure and obtaining the electrical distances between each power station node as the line weights between each node. The expression for this weight can be derived from the impedance matrix of the actual power grid.

[0042]

[0043] Among them, z ii , z jj , z ij Y is an element in the actual power grid impedance matrix. ii Let be the equivalent admittance between node i and node j. It is important to note that the elements here can include standby lines and loop lines that are not yet in use. These lines can be urgently deployed when the power grid suffers external attacks under extreme conditions, thereby improving the grid's extreme survivability.

[0044] Because the substation nodes in the embodiments provided in this application involve two voltage levels, 500kV and 220kV, the two voltage levels can be classified into the same voltage level. Then, the final line weight is obtained using the concept of impedance correction factor, the expression of which is:

[0045] w ij =Y ij (U ij / U N ) 2

[0046] The path weights between nodes can then be represented as:

[0047]

[0048] This weight matrix can be used to construct a power grid connectivity diagram of a real power grid structure.

[0049] S14: Construct a weighted adjacency matrix in the power grid connectivity graph to obtain a weighted connectivity graph.

[0050] An adjacency matrix is ​​constructed in the power grid connectivity graph. This matrix reflects the connection status between nodes in the actual power grid structure, i.e., the correlation between two nodes. In the adjacency matrix, if two nodes are directly connected, their corresponding elements are represented as 1; otherwise, they are represented as 0. Here, the elements corresponding to the aforementioned unused backup lines and loop lines in the adjacency matrix are set to 1. Then, the electrical distance obtained above is used as the edge weight to weight the adjacency matrix, resulting in:

[0051] Ew = E × W,

[0052] Where Ew represents an n×n symmetric matrix, reflecting the adjacency matrix after electrical distance weighting; E is the basic adjacency matrix, W is the electrical distance weight matrix; × is the Hadamard product, which represents the product of corresponding elements of the two matrices.

[0053] The power grid connectivity graph is modified using the weighted adjacency matrix to obtain a weighted connectivity graph.

[0054] S15: Based on the weighted connected graph, use the shortest path algorithm to find the shortest path between the sending node and the sending node.

[0055] S16: Mark the weighted connected graph using the shortest path between the sending node and the sending node to obtain the topology of the actual power grid structure.

[0056] Specifically, the process of randomly partitioning the topology of the actual power grid structure according to the total number of partitions to obtain the particle swarm optimization can include:

[0057] In the power grid partitioning method provided in this application embodiment, in order to avoid quickly getting trapped in a local solution during the subsequent optimization process, two methods are used to construct the particle swarm in this step, as follows:

[0058] 1) First, construct the first sub-particle swarm M1.

[0059] The obtained topology map of the actual power grid structure is divided into several target power supply areas. The specific division method is as follows: Figure 2 As shown:

[0060] S01: Preset power supply radius.

[0061] The power supply radius can be obtained by analyzing the shortest path between each node.

[0062] S02: Taking the upstream node as the center, all isolated downstream nodes whose shortest path to the upstream node is less than the power supply radius are taken as target downstream nodes.

[0063] In the topology diagram of the actual power grid structure, for each upstream node, taking that upstream node as the center, all isolated downstream nodes corresponding to the shortest path between that upstream node being less than the set power supply radius are taken as target downstream nodes. Here, isolated downstream nodes refer to downstream nodes that are not divided into a power supply area with any upstream node.

[0064] S03: Divide the up-sending node and the target down-sending node into a power supply area, thereby obtaining several power supply areas.

[0065] In this application, each target power supply area is set to contain only one up-feed node. Therefore, when dividing the up-feed node and down-feed node into regions, there may be isolated down-feed nodes. Thus, the power supply radius needs to be continuously modified during the division process.

[0066] S04: Determine whether the sum of the power supply capacity of all upstream nodes in the topology diagram of the actual power grid structure is equal to the sum of the load capacity of all downstream nodes. If yes, proceed to step S05; otherwise, proceed to step S08.

[0067] The purpose of this step is to maintain a balance between voltage supply and demand.

[0068] S05: Determine whether the sum of the power supply capacity of the upstream node and the load capacity of all downstream nodes in each power supply area is equal. If yes, proceed to step S06; otherwise, proceed to step S07.

[0069] S06: Select several power supply areas as target power supply areas.

[0070] S07: Adjust the power supply radius and return to step S02.

[0071] S08: Determine whether there is an isolated downstream node in the topology diagram of the actual power grid structure. If yes, proceed to step S07; otherwise, proceed to step S06.

[0072] This yields a topology diagram of the actual power grid structure divided into several target power supply areas. This set of target power supply areas can be represented as [G1, G2, ..., G...]. M M represents the number of data transmission nodes in the entire actual power grid structure, which can be represented by P. m Let P represent the output power of the m-th uplink node, and let N represent the total number of downlink nodes. n This represents the input power of the nth downstream node.

[0073] These target power supply areas are randomly partitioned according to the total number of partitions, that is, one or more target power supply areas are randomly divided into one area of ​​a total partition, thus obtaining a particle swarm.

[0074] Therefore, after randomly partitioning the topology of the actual power grid structure, a region may contain one or more target power supply areas. However, each region must contain at least one black-start power supply node, meaning that among the one or more power supply nodes in that region, at least one power supply node is a black-start power supply node. A black-start power supply refers to a system that, after a fault causes a complete power outage (excluding isolated small power grids that may still be operating), is in a completely "black" state. Without relying on other networks, it starts up through generators with self-starting capabilities, which in turn drive generators without self-starting capabilities, gradually expanding the system recovery range and ultimately restoring the entire system.

[0075] Furthermore, the actual power grid to be partitioned may be an inward-sending power grid, i.e., a receiving-end power grid. Therefore, it is impossible to achieve power source-load balance during partitioning. In this case, the role of external power supply needs to be considered, and the nodes that supply external power should be designated as upstream nodes. In addition, in step S08, when determining whether there are isolated downstream nodes in the topology diagram of the actual power grid structure, if so, the downstream node can be preferentially assigned to the power supply area of ​​the upstream node with the shortest path between it and the upstream node, and then step S02 can be executed.

[0076] 2) Reconstruct the second sub-particle swarm M2.

[0077] Based on the total number of partitions, all power plant nodes in the actual power grid structure are randomly partitioned to obtain the second sub-particle swarm.

[0078] In this step, it is not necessary to divide the actual power grid structure into target power supply areas. Instead, the grid is divided directly according to the total number of partitions to obtain the second sub-particle swarm.

[0079] Then, the first and second sub-particle groups are combined according to a preset ratio, such as 1:1, to obtain a particle group.

[0080] Optionally, when calculating the load loss risk value of a particle swarm under extreme external attacks, the following idea can be used to construct a formula for calculating the load loss risk value:

[0081] Using the power grid operating states under three fault types—total substation power outage at 500kV and 220kV substation nodes, main protection failure at a single substation, and simultaneous faults on multiple lines in important transmission channels—a fault set X = {X1, X2, ..., X} is constructed. h};

[0082] Obtain the probability of each fault in the fault set for the particle swarm; calculate the load loss generated by the particle swarm when each fault in the fault set occurs.

[0083] For each type of fault in the fault set, the probability of the fault occurring is multiplied by the amount of load loss generated when the fault occurs to obtain the load loss risk sub-value generated by the particle swarm when it is attacked by external forces; the load loss risk sub-values ​​of each fault in the fault set are added together to obtain the load loss risk value.

[0084] Therefore, the calculation formula is as follows:

[0085]

[0086] Where η represents the fault in the fault set, v η Let ΔS represent the probability that the particle swarm will experience the ηth fault in the fault set when it is subjected to an external attack. η This represents the risk value of load loss generated by the particle swarm when the ηth failure occurs.

[0087] The load loss generated by the particle swarm in each of the aforementioned fault sets can be calculated using transient simulation methods. During the transient simulation, the role of the three lines of defense in the power grid is considered. If the actual power grid structure remains stable after being protected by the three lines of defense under extreme conditions of external attack, then the load cut off by the three lines of defense represents the load loss of the power grid. If the actual power grid structure becomes unstable after being protected by the three lines of defense under extreme conditions of external attack, then it is considered that the actual power grid has lost all the loads in its structure.

[0088] Specifically, in steps S4 and S5: using an improved particle swarm optimization algorithm, with the goal of reducing and stabilizing the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks, the particle swarm is optimized; when the load loss risk value generated by the particle swarm under extreme conditions when subjected to external attacks reduces and stabilizes, the specific process of obtaining the power grid zoning scheme of the actual power grid structure may include:

[0089] Essentially, the power grid zoning problem is the process of allocating all power plant nodes within the power grid to different regions. Particle swarm coding is used to number all power plants in the actual power grid, forming a one-dimensional array that serves as the position information of the particles in the particle swarm.

[0090] The position information of the i-th particle can be denoted as: X i =(x i1 ,x i2 ,…x im ,…,x iN ),in,

[0091] Q represents the total number of power plant nodes in the actual power grid structure, and x represents the element in the location information. im =d, indicating that the plant node with number p belongs to the d-th partition, d∈{1,2,...,D}, where D is the total number of partitions. If x im =x in This means that power plant node p and power plant node q are assigned to the same power grid partition.

[0092] Define the velocity of the i-th particle as: V i =(v i1 ,v i2 ,…,v id ,…,v iN ), where v id A value of 0 or 1 indicates the correction amount for the node's partition number. Assume the current optimal solution for the i-th particle is: P i =(p i1 ,p i2 ,…,p im ,…,p iN The current optimal solution for the entire particle swarm is: Q = (q1, q2, ..., q m ,…,q N ).

[0093] By using the sigmoid activation function to map the particle's velocity to the interval [0, 1] and using it as a probability, we can obtain:

[0094]

[0095] Where w is the inertia coefficient, c1 and c2 are learning factors, and r1 and r2 are acceleration coefficients.

[0096] 1) Because the first sub-particle swarm M1 first divides the topology of the actual power grid structure into target power supply areas, and then performs random partitioning. The final power grid partition is defined as the upper layer, and the divided target power supply areas are defined as the lower layer. Therefore, when updating the position of the first sub-particle swarm M1 in the particle swarm using the update rate, the upper layer position information of the i-th particle is defined as X. i =(x i1 ,x i2 ,…x im ,…,x iN The lower-level position information of the i-th particle is F. i =(f i1 ,f i2 ,…,f im ,…,f iN ), where f im=g, where g∈{1,2,…,G} is the number of the target power supply area where the node is located.

[0097] Then the mapping relationship of the particle's position information between the upper and lower layers is represented as:

[0098] x id =B i (f ig )=d i , and f ig =d i .

[0099] Therefore, the formula for updating location information is defined as:

[0100]

[0101] For the i-th particle in M1, its lower-layer position information is f. ig ≠0, when v id When = 1, its position update formula is: in, The numbers are the target power supply regions adjacent to the target power supply region g where node d is located. The adjacency relationships of each target power supply region can be calculated from the boundary order set. This indicates that a target power supply area is randomly selected from the adjacent target power supply areas, and the power grid partition number is assigned to x. id .

[0102] The following is a specific example to illustrate the approach to updating the position of the first sub-particle swarm:

[0103] Nine target power supply areas have been divided in the topology diagram of an actual power grid structure, and they are numbered ①, ②, ③, ④, ⑤, ⑥, ⑦, ⑧, and ⑨ respectively. These ① to ⑨ are the lower-level numbers of the particles. Assuming that the total number of partitions is 3, these nine target power supply areas are randomly divided into three actual power grid structures, and the upper-level numbers of the particles are 1, 2, and 3.

[0104] Suppose that during a certain optimization, when partitioning the upper layer, the target power supply areas numbered ①, ③, and ⑧ are assigned to the power grid partition numbered 1, the target power supply areas numbered ④, ⑤, and ⑨ are assigned to the power grid partition numbered 2, and the target power supply areas numbered ②, ⑥, and ⑦ are assigned to the power grid partition numbered 3.

[0105] Among them, ① and ④ are both target power supply areas adjacent to ⑥.

[0106] Assuming the i-th particle's lower position is ⑥ and its upper position is 3, after a certain optimization, the load loss risk value generated by the combination of the first and second sub-particle groups under extreme conditions when subjected to external attacks is much smaller than the load loss risk value obtained after the previous optimization (i.e., it has not tended to stabilize). Then, the lower position of the i-th particle is updated to any target power supply area adjacent to ⑥ that does not belong to the same power grid partition. The target power supply areas adjacent to ⑥ are ① and ④. Neither ① nor ④ belongs to the power grid partition numbered 3. So, one is randomly selected, for example, ①, that is, the particle is assigned to the area with an upper position of ① and a lower position of 1.

[0107] 2) In the second sub-particle group M2, the lower-level position information of the i-th particle is f. ig =0, when v id When x = 1, its position update formula is x id (t+1)=rand(β ∈d ), and assign the power grid partition number to x id .

[0108] Specifically, assuming a total of 3 partitions, after partitioning, the particle is assigned to the power grid partition numbered 2. If the load loss risk value generated by the combination of the second sub-particle group and the second sub-particle group under extreme conditions when subjected to external attacks does not decrease to a stable level, then the particle is randomly assigned to another power grid partition other than 2.

[0109] In addition, to ensure that each grid section has at least one black-start power node, the grid section number of the black-start power node can be kept unchanged, i.e., v id =0.

[0110] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0112] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A power grid zoning method, characterized in that, include: Obtain the topology diagram of the actual power grid structure; Set the total number of partitions, and randomly partition the topology of the actual power grid structure according to the total number of partitions to obtain a particle swarm. Calculating the load loss risk value of the particle swarm under extreme conditions when subjected to external attack includes: constructing a fault set using the faults that occur when the particle swarm is subjected to external attack under extreme conditions; obtaining the probability of the particle swarm experiencing each fault in the fault set; calculating the load loss amount of the particle swarm when experiencing each fault in the fault set; for each fault in the fault set, multiplying the probability of the fault occurrence by the load loss amount to obtain a sub-value of the load loss risk of the particle swarm under external attack; and summing the load loss risk sub-values ​​of each fault in the fault set to obtain the load loss risk value. An improved particle swarm optimization algorithm is used to optimize the particle swarm, with the goal of reducing and stabilizing the risk value of load loss when the particle swarm is subjected to external attacks under extreme conditions. When the load loss risk value generated by the particle swarm under extreme conditions is reduced and tends to stabilize, a power grid partitioning scheme for the actual power grid structure is obtained.

2. The method according to claim 1, characterized in that, The process of obtaining the topology map of the actual power grid structure includes: Obtain the power of each power plant node in the actual power grid structure; Based on the power of each plant node, each plant node is divided into upstream power supply nodes and downstream power supply nodes; The electrical distance between each power plant node is obtained as the line weight between each node, and the power grid connectivity diagram is constructed using the line weight. A weighted adjacency matrix is ​​constructed in the power grid connectivity graph to obtain a weighted connectivity graph; Based on the weighted connected graph, the shortest path between the sending node and the sending node is obtained using the shortest path algorithm; The weighted connected graph is marked using the shortest path between the sending node and the sending node to obtain the topology of the actual power grid structure.

3. The method according to claim 1, characterized in that, The step of randomly partitioning the topology map of the actual power grid structure according to the total number of partitions to obtain a particle swarm includes: The topology of the actual power grid structure is divided into several target power supply areas; According to the total number of partitions, the target power supply areas in the topology diagram of the actual power grid structure are randomly partitioned to obtain the first sub-particle group; Based on the total number of partitions, all power plant nodes in the actual power grid structure are randomly partitioned to obtain the second sub-particle swarm. The first sub-particle group and the second sub-particle group are combined according to a preset ratio to obtain the particle group.

4. The method according to claim 3, characterized in that, The topology diagram of the actual power grid structure includes upstream and downstream power supply nodes; dividing the topology diagram of the actual power grid structure into several target power supply areas includes: Pre-set power supply radius; In the topology diagram of the actual power grid structure, for each data transmission node: Centered on the upstream power supply node, all isolated downstream power supply nodes whose shortest path to the upstream power supply node is less than the power supply radius are taken as target downstream power supply nodes; wherein, the isolated downstream power supply node is a downstream power supply node that is not divided into a power supply area with any upstream power supply node; The up-feed node and the target down-feed node are divided into a power supply area, thereby obtaining several power supply areas; Determine whether the sum of the power supply capacity of all upstream nodes in the topology diagram of the actual power grid structure is equal to the sum of the load capacity of all downstream nodes; If so, determine whether the sum of the power supply capacity of the upstream node and the load capacity of all downstream nodes in each power supply area is equal. If so, select several power supply areas as the target power supply areas.

5. The method according to claim 4, characterized in that, Also includes: If the sum of the power supply capacity of the upstream node and the load capacity of all downstream nodes in each power supply area is not equal, the power supply radius is adjusted, and the process returns to the step of taking the upstream node as the center and selecting all isolated downstream nodes whose shortest path to the upstream node is less than the power supply radius as the target downstream node.

6. The method according to claim 4, characterized in that, Also includes: If the sum of the power supply capacity of all upstream nodes in the topology of the actual power grid structure is not equal to the sum of the load capacity of all downstream nodes, then it is determined whether there are isolated downstream nodes in the topology of the actual power grid structure. If so, the power supply radius is adjusted, and the process returns to the step of taking the upstream node as the center and selecting all isolated downstream nodes whose shortest path to the upstream node is less than the power supply radius as the target downstream nodes.

7. The method according to claim 6, characterized in that, Also includes: If it is determined that there are no isolated downstream nodes in the topology diagram of the actual power grid structure, then several of the power supply areas are taken as the target power supply areas.

8. The method according to claim 1, characterized in that, The calculation of the load loss generated by the particle swarm when each type of fault in the fault set occurs includes: The load loss generated by the particle swarm when each fault in the fault set occurs is calculated using a transient simulation method.

9. The method according to claim 1, characterized in that, The improvement of the particle swarm optimization algorithm aims to reduce and stabilize the risk of load loss when the particle swarm is subjected to external attacks under extreme conditions. The optimization includes: The improved particle swarm optimization algorithm is used to update the velocity of each particle in the particle swarm. The position of each particle in the particle swarm is updated using the updated velocity to obtain a new particle swarm; Determine whether the load loss risk value generated by the new particle swarm under extreme conditions when subjected to external attacks decreases and tends to stabilize; If so, the optimization ends; If not, return to the step of updating the velocity of each particle in the particle swarm using the improved particle swarm algorithm.

Images