Unit combination scheduling method and device, computer equipment, medium and product

Abstract

The invention relates to a unit combination scheduling method and device, computer equipment, a medium and a product. The method comprises the following steps: acquiring topological connection information and system operation information corresponding to a target unit combination; the topological connection information is used for indicating a topological connection relationship in a target power network where the target unit combination is located; the system operation information is used for indicating system operation characteristics of the target power network; based on the topological connection information and the system operation information, performing spatial decoupling on the target power network to obtain a plurality of target decoupling sub-regions corresponding to the target power network; and determining a plurality of sub-scheduling results corresponding to the unit combinations in the plurality of target decoupling sub-regions, and determining a target scheduling result of the target unit combination in combination with the plurality of sub-scheduling results. By adopting the method, the scheduling processing efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the field of electrical engineering technology, and in particular to a method, device, computer equipment, medium and product for unit combination scheduling. Background Art

[0002] As the proportion of renewable energy in new power systems continues to increase, the uncertainty of renewable energy output makes the operation state of the power grid changeable, requiring dispatching decisions to be made in a shorter time, shortening the decision window. At the same time, it is urgent to give full play to the flexible adjustment capabilities of various adjustable resources. The scale of optimization models faced by the dispatching business of multi-resource power grids with sources, loads and storage has increased sharply, while the calculation time of the existing branch-and-bound mixed integer programming algorithm has exploded, making it difficult to meet the requirements of efficient optimization calculations of large power grids with multiple resources and multiple time scales.

[0003] Therefore, it is urgent to improve the efficiency of massive data processing and optimization analysis and calculation of ultra-large-scale power grids. Security Constrained Unit Commitment (SCUC) is a basic tool for optimizing power system operation. How to quickly solve this problem is the key to improving the efficiency of power system analysis and calculation. The computational complexity of the SCUC problem is in geometric series with the scale of the problem. In the context of the rapid increase in the scale of the problem, there is an urgent need for a solution that can improve the efficiency of processing unit combination scheduling. Summary of the invention

[0004] Based on this, it is necessary to provide a unit combination scheduling method, device, computer equipment, medium and product that can improve the scheduling processing efficiency in response to the above technical problems.

[0005] In a first aspect, the present application provides a unit combination scheduling method, comprising:

[0006] Acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0007] Based on the topological connection information and system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0008] Determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-regions, and determine the target dispatching result of the target unit combination in combination with the multiple target sub-dispatching results.

[0009] In one of the embodiments, based on the topological connection information and the system operation information, the target power network is spatially decoupled to obtain a plurality of target decoupled sub-areas corresponding to the target power network, including:

[0010] Acquire a power grid diagram model corresponding to the target power network based on topological connection information and system operation information;

[0011] The power grid graph model is partitioned by graph partitioning to obtain multiple target decoupled sub-areas.

[0012] In one embodiment, a power grid graph model is divided into graphs by using a graph division method to obtain multiple target decoupled sub-areas, including:

[0013] Determine the graph state matrix of the target power network based on the power grid graph model; the graph state matrix includes an adjacency matrix and a characteristic matrix, the adjacency matrix is ​​used to represent the topological connection information, and the characteristic matrix is ​​used to represent the system operation information;

[0014] The target power network is initially spatially divided using a weighted objective function to obtain a plurality of initial decoupled sub-regions corresponding to the target power network; the weighted objective function is determined based on an adjacency matrix; the weighted objective function is used to characterize the coupling and scale differences between the decoupled sub-regions;

[0015] The gain objective function is used to perform regional iterative update on multiple initial decoupled sub-regions to obtain multiple target decoupled sub-regions corresponding to the target power network; the gain objective function is determined based on the adjacency matrix and the characteristic matrix; the gain objective function is used to characterize the global gain of the decoupled sub-region transformation of the vertices in the power grid graph model.

[0016] In one embodiment, a gain objective function is used to perform regional iterative update on multiple initial decoupled sub-regions to obtain multiple target decoupled sub-regions corresponding to the target power network, including:

[0017] According to the gain objective function, the global gains corresponding to different decoupled sub-region transformations of each vertex are obtained, and a gain matrix is ​​generated;

[0018] According to the gain matrix, the regional update results of multiple decoupled sub-regions are obtained when the global gain is maximized;

[0019] Repeat the above process of generating the gain matrix and obtaining the regional update result until the preset iteration stop condition is met.

[0020] In one embodiment, determining multiple target sub-scheduling results corresponding to the unit combinations in the multiple target decoupling sub-regions respectively includes:

[0021] According to the multiple sub-SCUC models corresponding to the unit combinations in the multiple target decoupling sub-regions, the decision local variables of the multiple sub-SCUC models are updated to obtain the multiple initial sub-dispatching results corresponding to the multiple target decoupling sub-regions;

[0022] By optimizing the cross-region coupling variables between multiple target decoupling sub-regions, multiple initial sub-scheduling results are updated to obtain multiple updated sub-scheduling results;

[0023] Calculate the boundary consistency parameters between the multiple updated sub-scheduling results, and output multiple target sub-scheduling results when the boundary consistency parameters meet the preset convergence conditions.

[0024] In one embodiment, the method further comprises:

[0025] When the boundary consistency parameters do not meet the preset convergence conditions, the decision local variables and cross-region coupling variables of the sub-SCUC model are repeatedly updated until the boundary consistency parameters between the sub-scheduling results meet the preset conditions or reach the preset number of repeated iterations; the boundary consistency parameters include the original residual and the dual residual.

[0026] In a second aspect, the present application also provides a unit combination scheduling device, comprising:

[0027] An acquisition module is used to acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0028] A decoupling module is used to spatially decouple the target power network based on topological connection information and system operation information to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0029] The combination module is used to determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-areas, and determine the target dispatching result of the target unit combination by combining the multiple target sub-dispatching results.

[0030] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0031] Acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0032] Based on the topological connection information and system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0033] Determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-regions, and determine the target dispatching result of the target unit combination in combination with the multiple target sub-dispatching results.

[0034] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0035] Acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0036] Based on the topological connection information and system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0037] Determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-regions, and determine the target dispatching result of the target unit combination in combination with the multiple target sub-dispatching results.

[0038] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0039] Acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0040] Based on the topological connection information and system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0041] Determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-regions, and determine the target dispatching result of the target unit combination in combination with the multiple target sub-dispatching results.

[0042] The above-mentioned unit combination scheduling method, device, computer equipment, medium and product obtain the topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network; based on the topological connection information and the system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network; multiple sub-dispatching results corresponding to the unit combinations in the regions of the multiple target decoupled sub-areas are determined, and the target dispatching result of the target unit combination is determined in combination with the multiple sub-dispatching results. When dividing sub-areas, the related technology often only relies on the geographical location or the topological structure of the power grid, and lacks comprehensive consideration of the operating characteristics of the power system (such as the load level of the power grid, the distribution of power generation capacity, the network impedance, etc.). Due to the lack of full consideration of the load balance and resource allocation between the regions, some sub-areas may be too complex, containing a large number of equipment and constraints, resulting in excessive calculation burden; while other areas may be too simple, resulting in a waste of computing resources. The unit combination scheduling method provided in the present application obtains the topological connection information and system operation information of the power network, comprehensively considers the topological connection relationship between key equipment and the operation characteristics of the power system, realizes the balance and decoupling of the sub-region division of the power network, and can significantly reduce the coupling between regions, thereby improving the iteration efficiency and convergence speed when solving the SUCU model of the sub-region, greatly improving the calculation efficiency of the scheduling plan of the target unit combination of the whole network, and at the same time enhancing the applicability and reliability of the unit combination scheduling plan in large-scale complex power grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 An application environment diagram of a unit combination scheduling method in an embodiment;

[0045] Figure 2 A schematic diagram of a flow chart of a unit combination scheduling method in one embodiment;

[0046] Figure 3 A schematic diagram of a flow chart of steps for obtaining multiple target decoupled sub-regions in one embodiment;

[0047] Figure 4 A schematic flow chart of the steps of obtaining multiple target decoupled sub-regions in another embodiment;

[0048] Figure 5 A schematic flow chart of the step of determining multiple target sub-scheduling results in another embodiment;

[0049] Figure 6 Schematic diagram of the process of step C in one embodiment;

[0050] Figure 7 is a schematic flow chart of step D in one embodiment;

[0051] Figure 8 is a structural block diagram of a unit combination scheduling device in an embodiment;

[0052] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with 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.

[0054] The mathematical model of the SCUC problem is a complex large-scale mixed integer nonlinear programming (MINLP) problem. Due to its large-scale, mixed integer and NP-difficult characteristics, it is difficult for existing technologies to achieve a fast and direct solution to the SCUC problem. In the context of extensive load-side flexibility resources participating in unit combination calculations, the scale of the SCUC problem is growing rapidly. Most of the solvers in related technologies use branch-and-bound based mixed integer programming algorithms, which have slow solution speeds and are difficult to meet the timeliness requirements of power market clearing.

[0055] In the context of the current surge in the scale of problems, adopting the idea of ​​"divide and conquer" to solve the SCUC problem is a very promising technical route, that is, decomposing a large-scale problem into multiple sub-problems, and obtaining the solution of the SCUC problem by solving different sub-problems. Among them, the decomposition technology can adopt the following methods: (1) Lagrangian relaxation algorithm (LR) decomposes the SCUC problem into a series of easy-to-solve single-machine group sub-problems, which can speed up the solution speed and obtain a suboptimal solution with good quality. However, the LR method has a duality gap and is prone to oscillation during the iterative solution process. It cannot directly obtain a feasible solution to the SCUC problem, and it is even more difficult to obtain a global optimal solution. (2) Benders decomposition (BD) is to decompose the mixed integer programming problem into a main problem and a linear (or nonlinear) programming sub-problem for alternating solution. It can effectively handle complex constraints and variables and reduce the difficulty of solution. However, the Benders cut plane contains fewer continuous variables of the original problem, and its effectiveness is affected by the poor density of the cut plane, which leads to slow convergence of the BD method and a large number of iterations. (3) The alternating direction method of multipliers (ADMM) is an effective algorithm for solving convex programming problems with a block structure. It is essentially an augmented Lagrangian method that mainly modifies the multipliers in the model.

[0056] However, when dividing sub-regions, related technologies often only rely on geographical location or grid topology, lacking comprehensive consideration of the operating characteristics of the power system (such as grid load level, power generation capacity distribution, network impedance, etc.). Secondly, due to the lack of full consideration of load balance and resource allocation between regions, some sub-regions may be too complex, containing a large number of equipment and constraints, resulting in excessive computational burden; while other regions may be too simple, resulting in a waste of computing resources. The above problems lead to poor application of spatial decoupling methods in large-scale power grid optimization scheduling, lack of scientific rationality, low algorithm efficiency, and low reliability of scheduling schemes.

[0057] In order to improve the efficiency of the algorithm and the reliability of the scheduling scheme, an embodiment of the present application provides a unit combination scheduling method, which obtains the topological connection information and system operation information of the power network, comprehensively considers the topological connection relationship between key equipment and the operating characteristics of the power system, and achieves the balance and decoupling of the sub-region division of the power network. It can significantly reduce the coupling between regions, thereby improving the iteration efficiency and convergence speed when solving the SUCU model of the sub-region, and at the same time enhances the applicability and reliability of the unit combination scheduling scheme in large-scale complex power grids.

[0058] The unit combination scheduling method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. Among them, the terminal 102 can be but is not limited to various personal computers, laptops, smart phones, tablets, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0059] For ease of introduction, the following will take the implementation of the present application by a computer device as an example for description. The computer device is an electronic device with computing and processing capabilities. The computer device can be a single device, for example, Figure 1 The terminal 102 or server 104 in the application environment shown; or, the computer device may also be a collection of multiple devices, for example, the computer device may include the above Figure 1 The terminal 102 and the server 104 in the system shown, that is, the embodiment of the present application can be executed by the interaction of the above-mentioned terminal 102 and the server 104.

[0060] In an exemplary embodiment, Figure 2 As shown, a unit combination scheduling method is provided, including the following steps 202 to 206. Among them:

[0061] Step 202: Acquire topological connection information and system operation information corresponding to the target unit combination.

[0062] The target unit combination refers to the unit combination of the entire network to be scheduled for power generation in the power network. The topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located, and the system operation information is used to indicate the system operation characteristics of the target power network.

[0063] Exemplarily, the computer device obtains the connection relationship between key equipment in the target power network where the target unit combination is located, such as the connection relationship between key equipment such as busbars, generator sets, and substations. The computer device can also obtain the power system operation characteristics corresponding to the target power network where the target unit combination is located, such as grid load level, power generation capacity distribution, network impedance and other characteristics.

[0064] Step 204 : based on the topological connection information and the system operation information, spatially decouple the target power network to obtain a plurality of target decoupled sub-areas corresponding to the target power network.

[0065] Among them, spatial decoupling refers to the spatial division of the target power network.

[0066] The embodiments of the present application use topological connection relationships and system operation information, and not only rely on geographic location information or topological relationships, but can also fully consider operating characteristics such as grid load balancing and resource allocation between regions. When spatially dividing the target power network, based on topological connection relationships and system operation characteristics, it is possible to comprehensively consider reducing the coupling between sub-regions, ensuring the closeness and connectivity between sub-regions, balancing the computational burden between different regions, etc., so as to obtain multiple decoupled sub-regions.

[0067] Step 206, determining a plurality of sub-dispatching results corresponding to the unit combinations in the target decoupling sub-regions, and determining the target dispatching result of the target unit combination in combination with the plurality of sub-dispatching results.

[0068] Exemplarily, the computer device constructs a sub-SUCU model of the unit combination in the target decoupling sub-region for each target decoupling sub-region, and then determines the sub-dispatching result corresponding to the unit combination in the target decoupling sub-region according to the sub-SUCU model corresponding to the target decoupling sub-region. The sub-dispatching results of all target decoupling sub-regions are combined to form a unit combination plan for the entire network, that is, the target dispatching result is determined. The unit combination plan for the entire network includes the start and stop status of each unit at different time points and the power generation distribution results.

[0069] In the above-mentioned unit combination dispatching method, by obtaining the topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network; based on the topological connection information and the system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-regions corresponding to the target power network; multiple sub-dispatching results corresponding to the unit combinations in the regions of the multiple target decoupled sub-regions are determined, and the target dispatching result of the target unit combination is determined in combination with the multiple sub-dispatching results. When dividing sub-regions, the related technology often only relies on the geographical location or the topological structure of the power grid, and lacks comprehensive consideration of the operating characteristics of the power system (such as the load level of the power grid, the distribution of power generation capacity, the network impedance, etc.). Due to the lack of full consideration of the load balance and resource allocation between the regions, some sub-regions may be too complex, containing a large number of equipment and constraints, resulting in excessive computational burden; while other regions may be too simple, resulting in a waste of computing resources. The unit combination scheduling method provided in the embodiment of the present application obtains the topological connection information and system operation information of the power network, comprehensively considers the topological connection relationship between key equipment and the operation characteristics of the power system, realizes the balance and decoupling of the sub-region division of the power network, and can significantly reduce the coupling between regions, thereby improving the iteration efficiency and convergence speed when solving the SUCU model of the sub-region, thereby greatly improving the calculation efficiency of the scheduling plan of the target unit combination of the whole network, and at the same time enhancing the applicability and reliability of the unit combination scheduling plan in large-scale complex power grids.

[0070] In one embodiment, Figure 3 As shown, based on the topological connection information and system operation information, the target power network is spatially decoupled to obtain multiple target decoupled sub-areas corresponding to the target power network, including:

[0071] Step 302: Acquire a power grid diagram model corresponding to the target power network based on the topology connection information and the system operation information.

[0072] Step 304 , partitioning the power grid graph model using a graph partitioning method to obtain a plurality of target decoupling sub-regions.

[0073] Exemplarily, the computer device constructs a power grid graph model, and generates a power grid graph model for SCUC problem calculation based on data such as the power grid topology structure. The vertices of the graph represent key equipment (busbars, generator sets, substations) in the power grid, and the edges reflect the connection relationship between key equipment, where system operation information can be stored in the attributes of the corresponding vertices, such as vertex-related loads, power generation capacity, operating costs, etc. The computer device then divides the power grid graph model through a graph partitioning algorithm, thereby obtaining multiple target decoupled sub-areas after spatial partitioning.

[0074] In one embodiment, Figure 4 As shown in the figure, the power grid graph model is divided into graphs by using a graph partitioning method to obtain multiple target decoupled sub-areas, including:

[0075] Step 402: determining a graph state matrix of a target power network based on a power grid graph model.

[0076] Among them, the graph state matrix includes the adjacency matrix and the feature matrix. The adjacency matrix is ​​used to represent the topological connection information, and the feature matrix is ​​used to represent the system operation information.

[0077] Exemplarily, a graph state matrix is ​​constructed based on the power grid graph model, and the graph state matrix includes an adjacency matrix and a feature matrix. The adjacency matrix indicates whether the vertices are connected, and the feature matrix contains the correlation of the vertices (load, power generation capacity, operating cost, etc.). The adjacency matrix and the feature matrix are used as inputs, and the Kernighan-Lin graph partitioning algorithm is used for space partitioning.

[0078] For example, computer equipment has nodes and The vertices of the graph correspond to the nodes, and the edges of the graph correspond to the branches, building an undirected graph model ;in Represents a set of vertices; Represents the edge set. Based on the power grid graph model, the graph state matrix is ​​obtained. The characteristic matrix in the graph state matrix Depend on The eigenvector of the vertex Composition, including: parameter vector of production cost of production subject , For different types of production entities, including thermal power units, hydropower units, wind farms, photovoltaic power stations, virtual power plants and energy storage power stations; the production entity operation constraint parameter vector , including the maximum and minimum output of thermal power units, minimum continuous operation time, upward and downward climbing constraints, etc.; Suppose a day is divided into time period, For the power grid bus The load of time, If the vertex is a production entity, it only has production cost parameters and operation constraint parameters, and the load is zero. Conversely, if the vertex is a busbar, it only has load, and the production cost parameters and operation constraint parameters are zero. The adjacency matrix in the graph state matrix The dimension of , which is defined as follows:

[0079] ;

[0080] In the adjacency matrix, Represents a vertex and There is a connection relationship between them. It represents the vertex and There is no connection between them.

[0081] Step 404 , using a weighted objective function to perform initial spatial division on the target power network, to obtain a plurality of initial decoupled sub-regions corresponding to the target power network.

[0082] Among them, the weighted objective function is determined according to the adjacency matrix; the weighted objective function is used to characterize the coupling and scale differences between the decoupled sub-regions.

[0083] Exemplarily, the computer device performs graph partitioning on the power grid graph model by the Kernighan-Lin algorithm. First, a random initial partitioning is performed to randomly divide the entire network into a number of sub-areas, or, with the goal of minimizing the coupling between sub-areas and balancing the scale between the sub-areas, an initial spatial partitioning is performed to obtain a plurality of initial decoupled sub-areas.

[0084] For example, the Kernighan-Lin algorithm is used to partition the graph. In the initial space partitioning process, the goal is to: Divide into Sub-areas , to minimize the coupling between sub-regions while ensuring compactness and connectivity within the sub-regions. Define the partitioning scheme :

[0085] ;

[0086] in ,and .

[0087] The objective function is defined as minimizing the sum of cross-region edge weights (Cut):

[0088] ;

[0089] At the same time, the number of vertices in the sub-region should be as balanced as possible, with the following constraints:

[0090] ;

[0091] in is the maximum allowed difference in the number of vertices.

[0092] In order to comprehensively consider cross-regional coupling and sub-regional scale balance, a weighted objective function is defined:

[0093] ;

[0094] in is a function to measure the imbalance of sub-region size. and are the weights of the two parts of the objective function.

[0095] The weighted objective function is used to perform initial space division to obtain multiple initial decoupled sub-regions.

[0096] Step 406 , performing regional iterative update on the multiple initial decoupling sub-regions using the gain objective function to obtain multiple target decoupling sub-regions corresponding to the target power network.

[0097] Among them, the gain objective function is determined according to the adjacency matrix and the characteristic matrix; the gain objective function is used to characterize the global gain of the decoupled sub-region transformation of the vertices in the power grid graph model.

[0098] Exemplarily, after obtaining multiple initial decoupled sub-regions, the computer device performs iterative optimization, calculates the impact of the migration of vertices in each decoupled sub-region to other decoupled sub-regions on the partition boundary, selects the vertex pair with the largest gain for exchange, and updates the partition result until the gain converges or the maximum number of iterations is reached. Finally, the partition result is output to obtain multiple target decoupled sub-regions.

[0099] In a preferred embodiment, the gain objective function is used in step 406 to perform regional iterative update on multiple initial decoupling sub-regions to obtain multiple target decoupling sub-regions corresponding to the target power network, including: obtaining the global gains corresponding to different decoupling sub-region transformations of each vertex according to the gain objective function, and generating a gain matrix; according to the gain matrix, obtaining the regional update results of multiple decoupling sub-regions when the global gain is maximized; repeating the above process of generating the gain matrix and obtaining the regional update results until the preset iteration stop condition is met.

[0100] For example, after the initial partitioning of the computer equipment, Randomly divide into Sub-areas , to ensure that the number of vertices in each sub-region is as balanced as possible, the computer device calculates the gain matrix using the gain objective function, where for each vertex , calculate its movement to each different target area Gain ;

[0101] in, is the external cost of the vertex in other regions, that is, the vertex With other regions The sum of edge connectivity of internal vertices. For vertices in the current region The internal cost of the vertex The sum of edge connections to vertices in the same subregion.

[0102] After calculating the global gain, the gain matrix is ​​generated , where each element Represents a vertex Move to sub-area In each iteration, the vertex swap pair that maximizes the global gain is selected. Make the gain Maximum. Performing this exchange will change the vertex From area Move to area . Update the partitioning scheme , according to the new partitioning scheme, update the gain matrix . Iterate until the stopping condition is met. The stopping condition includes reaching the maximum number of iterations or No more increase. Output the optimal partitioning solution ;

[0103] In this embodiment, by graph modeling the power grid, the Kernighan-Lin algorithm is used to efficiently divide the power grid nodes, and the topological connection relationship, load characteristics and power generation distribution between nodes are comprehensively considered to achieve the balance and decoupling of sub-regional division. Different from the traditional method of directly obtaining sub-graphs based on geographical location or power grid topology, this algorithm tries to balance the computational burden between different regions, while significantly reducing the coupling between regions. The calculation scale of each sub-model when determining the sub-dispatching results is similar, and the iteration efficiency and convergence speed are greatly improved. The optimization effect of the division scheme is further improved from the calculation, and the cross-regional coupling is effectively reduced. At the same time, the applicability and reliability of the unit combination dispatching scheme in large-scale complex power grids are enhanced.

[0104] In one embodiment, Figure 5 As shown, multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-regions are determined, including:

[0105] Step 502, based on the multiple sub-SCUC models corresponding to the unit combinations in the multiple target decoupling sub-regions, the decision local variables of the multiple sub-SCUC models are updated to obtain the multiple initial sub-scheduling results corresponding to the multiple target decoupling sub-regions.

[0106] Step 504 , by optimizing the cross-region coupling variables between the multiple target decoupling sub-regions, the multiple initial sub-scheduling results are updated to obtain multiple updated sub-scheduling results.

[0107] Step 506 , calculating boundary consistency parameters between the multiple updated sub-scheduling results, and outputting multiple target sub-scheduling results when the boundary consistency parameters meet a preset convergence condition.

[0108] Among them, when the boundary consistency parameters do not meet the preset convergence conditions, the decision local variables and cross-regional coupling variables of the sub-SCUC model are repeatedly updated until the boundary consistency parameters between the sub-scheduling results meet the preset conditions or reach the preset number of repeated iterations; the boundary consistency parameters include the original residual and the dual residual.

[0109] Exemplarily, the computer device constructs a corresponding sub-SCUC model for each divided target decoupling sub-region. First, the sub-SCUC model is optimized and solved independently, the decision local variables of each sub-model are updated, and the preliminary unit combination scheme in the target decoupling sub-region, that is, the initial sub-scheduling result, is generated. Secondly, the boundary data between the sub-SCUC models interact, and the cross-region coupling variables (cross-region line power) of each sub-region are coordinated through the ADMM algorithm. The sub-region results are updated, and the scheduling scheme of each target decoupling sub-region is updated according to the data after the coupling variables are optimized. Finally, the boundary consistency of the scheduling scheme of each target decoupling sub-region is calculated to determine whether the convergence condition is met. If it is met or the number of iterations reaches the upper limit, the result integration is entered. If it is not met, the decision local variables and cross-region coupling variables of the above sub-models are updated iteratively until the convergence condition is met or the number of iterations reaches the upper limit, and then the updated scheduling optimization results corresponding to all target decoupling sub-regions are collected and integrated into the unit combination scheme of the whole network. The unit combination scheme of the whole network includes the start and stop status of each unit at different time points and the power generation distribution results.

[0110] In one embodiment, a unit commitment scheduling method is provided, the method comprising the following steps:

[0111] Step A: Build a safety constrained unit commitment problem model, obtain the grid operation scenario, and collect grid data such as grid topology, load forecast, and generator unit parameters. Build a safety constrained unit commitment (SCUC) problem model, which includes unit start-stop constraints, power balance constraints, and line transmission capacity constraints.

[0112] Step B, construct a power grid graph model, and generate a power grid graph model for SCUC problem calculation based on data such as the power grid topology. The vertices of the graph represent the key equipment in the power grid (busbars, generator sets, substations), and the edges reflect the connection relationship between the key equipment. Construct a graph state matrix based on the power grid graph model, and the graph state matrix includes an adjacency matrix and a feature matrix. Among them, the adjacency matrix indicates whether the vertices are connected, and the feature matrix contains the correlation of the vertices (load, power generation capacity, operating cost, etc.). Use the adjacency matrix and feature matrix as input to enter the Kernighan-Lin algorithm module.

[0113] For example, in step B, constructing the power grid diagram model and the state matrix includes: for a nodes and The vertices of the graph correspond to the nodes, and the edges of the graph correspond to the branches, building an undirected graph model ;in Represents a set of vertices; Represents the edge set. Based on the power grid graph model, the graph state matrix is ​​obtained. The characteristic matrix in the graph state matrix Depend on The eigenvector of the vertex Composition, including: parameter vector of production cost of production subject , For different types of production entities, including thermal power units, hydropower units, wind farms, photovoltaic power stations, virtual power plants and energy storage power stations; the production entity operation constraint parameter vector , including the maximum and minimum output of thermal power units, minimum continuous operation time, upward and downward climbing constraints, etc.; Suppose a day is divided into time period, For the power grid bus The load of time, If the vertex is a production entity, it only has production cost parameters and operation constraint parameters, and the load is zero. Conversely, if the vertex is a busbar, it only has load, and the production cost parameters and operation constraint parameters are zero. The adjacency matrix in the graph state matrix The dimension of , which is defined as follows:

[0114] ;

[0115] In the adjacency matrix, Represents a vertex and There is a connection relationship between them. It represents the vertex and There is no connection between them.

[0116] Step C, in the Kernighan-Lin algorithm module, the power grid model is partitioned. First, a random initial partition is performed to randomly divide the entire network into several sub-areas. Then, iterative optimization is performed to calculate the impact of each vertex migration on the partition boundary, select the vertex pair with the largest gain for exchange, and update the partition result until the gain converges or the maximum number of iterations is reached. Finally, the partition result is output to obtain multiple decoupled sub-areas, which are input into the ADMM solution module, and the SCUC model is solved independently for each sub-area. Please refer to Figure 6 , step C may include:

[0117] Step C1, initial partitioning, Randomly divide into Sub-areas , ensuring that the number of vertices in each sub-region is as balanced as possible.

[0118] Step C2, calculate the gain matrix, for each vertex , calculate its movement to each different target area Gain :

[0119] ;

[0120] in, is the external cost of the vertex in other regions, that is, the vertex With other regions The sum of edge connectivity of internal vertices. For vertices in the current region The internal cost of the vertex The sum of edge connections to vertices in the same subregion.

[0121] Generate gain matrix , where each element Represents a vertex Move to sub-area gain.

[0122] Step C2, select the optimal exchange. In each iteration, select the vertex exchange pair that can maximize the global gain. Make the gain Maximum. Performing this exchange will change the vertex From area Move to area .

[0123] Step C4, update the partitioning scheme , according to the new partitioning scheme, update the gain matrix . Iterate until the stopping condition is met. The stopping condition includes reaching the maximum number of iterations or No more increase. Output the optimal partitioning solution ;

[0124] Different from the traditional method of directly obtaining subgraphs based on geographical location or power grid topology, this algorithm tries to balance the computational burden between different regions, so that the computational scale of each submodel of the ADMM calculation module is similar, further improving the optimization effect of the partitioning scheme from the computational perspective and effectively reducing cross-regional coupling.

[0125] Step D, in the ADMM solution module, for each divided sub-region, a corresponding sub-SCUC model is constructed. First, the sub-model is optimized and solved independently to generate a preliminary unit combination plan in the region. Secondly, the boundary data between the sub-models interact, and the coupling variables (cross-regional line power) of each sub-region are coordinated through the ADMM algorithm. Update the sub-region results, and update the scheduling plan for each sub-region based on the data after the coupling variables are optimized. Finally, calculate the boundary consistency of the scheduling plan for each sub-region to determine whether the convergence conditions are met. If they are met or the number of iterations reaches the upper limit, enter the result integration module. If not, iterate and run step D. Please refer to Figure 7 , step D comprises:

[0126] Step D1, construct the sub-SCUC model, in the sub-graph In the system, the goal is to determine the start and stop status and power generation capacity of each unit in each time period to minimize the power generation cost while meeting various constraints for safe and stable operation of the power system.

[0127] Step D2: construct multi-region augmented Lagrangian functions ,in,

[0128] ;

[0129] in is the Lagrange multiplier used for the coordination of coupled constraints. is the regularization parameter and is the penalty coefficient for violating the coupling constraint.

[0130] Step D3: for each sub-region , with fixed Lagrange multipliers and global variables In the case of , the update formula is

[0131] ;

[0132] This step is solved independently for each region, which is equivalent to optimizing the local decision variables of the sub-region and the power flow of the inter-regional tie lines under the condition that the coupling variables in other regions are fixed.

[0133] Step D4, according to the cross-region power flow constraint, calculate and update the Lagrange multiplier, and the update formula is as follows:

[0134] ;

[0135] This step is used to gradually reduce the inconsistency of power flow in the interconnection lines between regions and accelerate global convergence.

[0136] Step D5, judging whether the ADMM algorithm converges based on the original residual and the dual residual, the convergence condition is as follows:

[0137] Raw residual Less than threshold :

[0138] ;

[0139] The original residual convergence condition reflects the degree of deviation from the original constraint, that is, whether the power flowing into and out of the tie line is equal.

[0140] Dual residual Less than threshold :

[0141] ;

[0142] The dual residual convergence condition reflects the degree of change of the dual variable update, that is, whether the dual variable (power flow of the inter-regional tie line) is still being updated.

[0143] The ADMM algorithm can efficiently decompose large-scale SCUC problems and achieve independent optimization of each region. At the same time, it ensures the consistency of global optimization through the coordination of coupling variables and Lagrange multipliers.

[0144] Step E: In the result integration module, the optimization results of all sub-regions are collected and integrated into a unit combination plan for the entire network. The unit combination plan for the entire network includes the start and stop status of each unit at different time points and the power generation allocation results.

[0145] In this embodiment, the Kernighan-Lin algorithm is used to efficiently divide the grid nodes. Different from the traditional method of directly obtaining subgraphs based on geographical location or grid topology, this algorithm tries to balance the computational burden between different regions, so that the computational scale of each submodel of the ADMM calculation module is similar, further improving the optimization effect of the division scheme from the computational perspective, effectively reducing cross-regional coupling, improving the iteration efficiency and convergence speed of the ADMM algorithm, and enhancing the applicability and reliability of the unit combination scheduling scheme in large-scale complex power grids. Decomposing large-scale mixed integer nonlinear programming problems into small-scale sub-problems for iterative solution overcomes the technical problem of slow solution speed of traditional solver methods in large-scale systems, and can meet the timeliness requirements of future virtual power plants and distributed power sources for extensive participation in power market clearing.

[0146] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0147] Based on the same inventive concept, the embodiment of the present application also provides a unit combination scheduling device for implementing the unit combination scheduling method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more unit combination scheduling device embodiments provided below can refer to the limitations of the unit combination scheduling method above, and will not be repeated here.

[0148] In an exemplary embodiment, Figure 8 As shown, a unit combination scheduling device is provided, including: an acquisition module 802, a decoupling module 804 and a combining module 806, wherein:

[0149] An acquisition module is used to acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network;

[0150] A decoupling module is used to spatially decouple the target power network based on topological connection information and system operation information to obtain multiple target decoupled sub-areas corresponding to the target power network;

[0151] The combination module is used to determine multiple target sub-dispatching results corresponding to the unit combinations in the multiple target decoupling sub-areas, and determine the target dispatching result of the target unit combination by combining the multiple target sub-dispatching results.

[0152] Each module in the above-mentioned unit combination scheduling device can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each of the above modules.

[0153] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to 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 that needs to be called when implementing the unit combination scheduling method. The input / output interface of the computer device is used to exchange information between the processor and an external device. 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, a unit combination scheduling method is implemented.

[0154] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0155] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0156] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0157] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0158] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0159] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and 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-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile 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. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0160] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, 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, they should be considered to be within the scope of this specification.

[0161] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A unit commitment scheduling method, characterized in that: The method comprises: Acquire topological connection information and system operation information corresponding to the target unit combination; the topological connection information is used to indicate the topological connection relationship in the target power network where the target unit combination is located; the system operation information is used to indicate the system operation characteristics of the target power network; Based on the topological connection information and the system operation information, spatially decoupling the target power network to obtain a plurality of target decoupled sub-areas corresponding to the target power network; Determine a plurality of target sub-scheduling results corresponding to the unit combinations in the regions of the plurality of target decoupling sub-regions, respectively, and determine a target scheduling result of the target unit combination in combination with the plurality of target sub-scheduling results.

2. The method according to claim 1, characterized in that The spatially decoupling the target power network based on the topological connection information and the system operation information to obtain a plurality of target decoupled sub-areas corresponding to the target power network includes: Acquire a power grid diagram model corresponding to the target power network based on the topology connection information and the system operation information; The power grid graph model is divided into graphs by using a graph division method to obtain the multiple target decoupling sub-areas.

3. The method according to claim 2, characterized in that The adopting a graph partitioning method to perform graph partitioning on the power grid graph model to obtain the multiple target decoupling sub-areas includes: Determining a graph state matrix of the target power network based on the power grid graph model; the graph state matrix includes an adjacency matrix and a characteristic matrix, the adjacency matrix is ​​used to characterize the topological connection information, and the characteristic matrix is ​​used to characterize the system operation information; Performing an initial spatial division on the target power network using a weighted objective function to obtain a plurality of initial decoupled sub-regions corresponding to the target power network; the weighted objective function is determined according to the adjacency matrix; the weighted objective function is used to characterize the coupling and scale differences between the decoupled sub-regions; A gain objective function is used to perform regional iterative update on multiple initial decoupling sub-regions to obtain the multiple target decoupling sub-regions corresponding to the target power network; the gain objective function is determined based on the adjacency matrix and the characteristic matrix; the gain objective function is used to characterize the global gain of the decoupling sub-region transformation of the vertices in the power grid diagram model.

4. The method according to claim 3, characterized in that The method of performing regional iterative updating on a plurality of initial decoupled sub-regions by using a gain objective function to obtain the plurality of target decoupled sub-regions corresponding to the target power network includes: According to the gain objective function, the global gains corresponding to the different decoupled sub-region transformations of each vertex are obtained, and a gain matrix is ​​generated; According to the gain matrix, obtaining regional update results of multiple decoupled sub-regions when the global gain is maximized; Repeat the above process of generating the gain matrix and obtaining the regional update result until the preset iteration stop condition is met.

5. The method according to claim 1, characterized in that Determining the multiple target sub-scheduling results corresponding to the unit combinations in the multiple target decoupling sub-regions includes: According to the multiple sub-SCUC models corresponding to the unit combinations in the multiple target decoupling sub-regions, the decision local variables of the multiple sub-SCUC models are updated to obtain multiple initial sub-scheduling results corresponding to the multiple target decoupling sub-regions; By optimizing the cross-region coupling variables between the multiple target decoupling sub-regions, the multiple initial sub-scheduling results are updated to obtain multiple updated sub-scheduling results; The boundary consistency parameters between the multiple updated sub-scheduling results are calculated, and when the boundary consistency parameters meet the preset convergence conditions, the multiple target sub-scheduling results are output.

6. The method according to claim 5, characterized in that The method further comprises: When the boundary consistency parameters do not meet the preset convergence conditions, the decision local variables and cross-region coupling variables of the sub-SCUC model are repeatedly updated until the boundary consistency parameters between the sub-scheduling results meet the preset conditions or reach the preset number of repeated iterations; the boundary consistency parameters include the original residual and the dual residual.

7. A unit combination dispatching device, characterized in that: The device comprises: An acquisition module, used to acquire topological connection information and system operation information corresponding to a target unit combination; the topological connection information is used to indicate a topological connection relationship in a target power network where the target unit combination is located; the system operation information is used to indicate system operation characteristics of the target power network; a decoupling module, configured to perform spatial decoupling on the target power network based on the topological connection information and the system operation information, and obtain a plurality of target decoupling sub-areas corresponding to the target power network; The combination module is used to determine multiple target sub-scheduling results corresponding to the unit combinations in the multiple target decoupling sub-regions, and determine the target scheduling result of the target unit combination in combination with the multiple target sub-scheduling results.

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

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

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

Images