Power topology spatial-temporal characteristic carbon emission dynamic modeling method, system and device and storage medium

Abstract

The invention discloses an electric power topology spatial-temporal characteristic carbon emission dynamic modeling method, system and device and a storage medium, and the method comprises the steps: obtaining the electric quantity data of a power grid and the carbon emission data corresponding to the electric quantity consumption, and calculating the carbon emission factor of each node in the power grid; based on the carbon emission factor of each node of the power grid, information data of different data sources are fused, and a space-time incidence matrix is constructed; the node type weight and the topological connection weight of the space-time incidence matrix are optimized, the space-time matrix obtained through optimization serves as input, feature extraction is carried out to obtain a feature vector fusing space-time features, and the feature vector is mapped into a dynamic carbon emission factor. According to the method, a nonlinear weight decreasing method is introduced, so that the difficulty of falling into local optimum is effectively avoided, and the calculation precision and the calculation efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of carbon emission calculation of power system, and particularly relates to a power topology space-time feature carbon emission dynamic modeling method, system, device and storage medium. BACKGROUND

[0002] The traditional carbon emission tracking method based on power flow analysis mainly builds a node carbon potential equation to allocate the carbon emissions generated on the power generation side to the load side according to the power proportion. Among them, the typical methods are the proportion sharing method and the marginal carbon emission intensity method. The operation mechanism of this kind of method is highly dependent on the results of power flow calculation. In the actual operation process, it is necessary to repeatedly use the iterative method to solve the nonlinear equation set, for example, the commonly used Gauss-Seidel method. When the scale of the power grid continues to expand and the number of nodes exceeds 1000, the time consumed by the calculation will increase significantly. In recent years, genetic algorithm (GA), particle swarm optimization (PSO) and other swarm intelligence algorithms have been gradually introduced into the research field of carbon emission optimization problem. These algorithms generate the corresponding scheduling scheme by carefully constructing multi-objective functions, such as minimizing carbon emissions and minimizing power generation cost. However, this kind of method also exposes obvious defects in actual application. In the high-dimensional and non-convex solution space, the algorithm is easy to fall into the dilemma of local optimum and cannot find the global optimal solution. Moreover, this kind of method also lacks the ability to explicitly model the space-time dynamic characteristics of the power grid, which leads to that the accuracy of the final carbon emission allocation is greatly limited, and the actual error is often more than 5%, which is difficult to meet the high requirements of accurate calculation of carbon emissions.

[0003] Currently, existing carbon emission flow analysis technologies face a series of severe challenges in practical applications. Data generated during power system operation, such as power generation output, load demand, and line losses, exhibit significant spatiotemporal correlations. However, traditional analysis methods treat spatial topology and temporal evolution as two independent components. The direct consequence of this approach is a significant lag in updating dynamic carbon emission factors, typically on an hourly basis. This makes it impossible for the system to capture minute-level data fluctuations in a timely manner. For example, a sudden drop in wind and solar power output can trigger abrupt changes in carbon potential, but traditional methods struggle to respond effectively to such rapid changes. The setting of power source type weights (e.g., the weights for thermal, wind, and solar power) and topology weights (i.e., weights related to line conduction efficiency) largely relies on past experience or fixed formulas, such as allocating weights according to carbon emission intensity ratios. The biggest problem with this approach is its lack of adaptability. For example, in scenarios with high wind power penetration, traditional weighting models cannot dynamically reduce the excessive influence of thermal power weight on carbon potential based on actual conditions, ultimately causing significant deviations in carbon emission allocation and affecting the accuracy and reliability of carbon emission flow analysis. Summary of the Invention

[0004] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a method, system, device, and storage medium for dynamic modeling of carbon emissions from the spatiotemporal characteristics of power topology, solving the problems of low efficiency in parameter tuning and subjective weight setting in traditional methods.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, embodiments of the present invention provide a method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology, comprising: acquiring power data of the power grid and carbon emission data corresponding to the power consumption, and calculating the carbon emission factor of each node in the power grid;

[0007] Based on the carbon emission factors of each node in the power grid, information data from different data sources are fused to construct a spatiotemporal correlation matrix;

[0008] The node type weights and topological connection weights of the spatiotemporal correlation matrix are optimized. The optimized spatiotemporal matrix is ​​used as input to extract features and obtain a feature vector that integrates spatiotemporal characteristics. This feature vector is then mapped to a dynamic carbon emission factor.

[0009] As a preferred embodiment of the dynamic modeling method for carbon emissions based on the spatiotemporal characteristics of power topology described in this invention, the calculation of carbon emission factors for each node in the power grid includes:

[0010] The carbon emission factor for any node i in the power grid is determined by the active power output P of the local power plant. g and its carbon emission coefficient δ g And the active power P input from other nodes j to node i. ij And the carbon emission factor δ at the input node j A weighted average is performed to obtain the preliminary carbon emission factor.

[0011] As a preferred embodiment of the dynamic modeling method for carbon emissions based on the spatiotemporal characteristics of power topology described in this invention, it further includes: for any node i connected to all power sources k, its output P k Multiplied by the weight a of the power type k and carbon emission intensity ρ k ;

[0012] For all other nodes j connected to any node i, its transmission power P ji Multiplied by the topological weight β of the branch from node j to node i ji And the carbon emission factor of node j;

[0013] Carbon emission intensity per unit load includes the load at node i. and all input power P ji .

[0014] The beneficial effect of this preferred technical solution is that by adding power source type weight and topology weight, it can more accurately reflect the impact of different types of power sources and grid structures on carbon emissions.

[0015] As a preferred embodiment of the dynamic modeling method for spatiotemporal carbon emissions of power topology described in this invention, the method involves fusing information data from different data sources based on the carbon emission factors of each node in the power grid to construct a spatiotemporal correlation matrix, including:

[0016] The spatiotemporal correlation matrix includes spatial dimension, time slice and feature dimension.

[0017] By integrating the distribution of user load in time and space, the time series of power generation and consumption, and the grid topology, a high-dimensional data structure of the power system's operating status is constructed.

[0018] As a preferred embodiment of the dynamic modeling method for carbon emissions of spatiotemporal characteristics of power topology described in this invention, the method further includes: the spatiotemporal correlation matrix contains carbon emission factor information and weight settings for different types of power sources.

[0019] As a preferred embodiment of the dynamic modeling method for carbon emissions based on the spatiotemporal characteristics of power topology described in this invention, the optimization of the node type weights and topology connection weights in the spatiotemporal correlation matrix includes:

[0020] A vector containing node type weights and topology connection weights is randomly generated, and the weights are adjusted according to a preset security threshold.

[0021] Based on the weight adjustment results, the data is aggregated towards the optimal solution to obtain the optimized weights. The optimized weights are then written back to the spatiotemporal correlation matrix to update the data.

[0022] As a preferred embodiment of the dynamic carbon emission modeling method for spatiotemporal characteristics of power topology described in this invention, the optimized spatiotemporal matrix is ​​used as input, feature extraction is performed to obtain a feature vector that integrates spatiotemporal characteristics, and this feature vector is mapped to a dynamic carbon emission factor, including:

[0023] Based on the optimized matrix, a spatiotemporal graph convolutional network is used to simultaneously extract spatial topological features and temporal evolution patterns. Then, an attention mechanism is used to focus on key nodes and output a feature vector that integrates spatiotemporal characteristics. The feature vector is mapped to a dynamic carbon emission factor and combined with real-time power data to generate carbon emission prediction results.

[0024] The beneficial effects of this preferred technical solution are that it performs in-depth analysis on the optimized matrix, extracts feature vectors containing rich spatiotemporal information, focuses on key nodes through an attention mechanism, and finally generates dynamic carbon emission factors and carbon emission prediction results. By transforming from static data to dynamic prediction, it provides strong technical support for the refined management and low-carbon development of the power system.

[0025] Secondly, the present invention provides a dynamic modeling system for carbon emissions based on the spatiotemporal characteristics of power topology, comprising:

[0026] The carbon emission factor calculation module is used to acquire the power consumption data of the power grid and the carbon emission data corresponding to the power consumption, and to calculate the carbon emission factor of each node in the power grid.

[0027] The data fusion module is used to fuse information data from different data sources based on the carbon emission factors of each node in the power grid, and construct a spatiotemporal correlation matrix.

[0028] The optimization module is used to optimize the node type weights and topological connection weights of the spatiotemporal correlation matrix. The optimized spatiotemporal matrix is ​​used as input to extract features and obtain a feature vector that integrates spatiotemporal characteristics, and then it is mapped to a dynamic carbon emission factor.

[0029] Thirdly, the present invention provides an electronic device, comprising:

[0030] Memory and processor;

[0031] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the dynamic modeling method for carbon emissions of spatiotemporal characteristics of power topology are implemented.

[0032] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method for dynamic modeling of carbon emissions in the spatiotemporal characteristics of power topology.

[0033] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention solves the problems of insufficient accuracy of traditional static modeling, inability to respond to dynamic changes in power grid topology in real time, and reliance on fixed weight factors and static parameters in existing methods; compared with traditional hyperparameter optimization relying on grid search or manual experience and being prone to getting trapped in local optima, this invention introduces a nonlinear decreasing weight method, which effectively avoids getting trapped in local optima and improves computational accuracy and efficiency. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0035] Figure 1 This is a flowchart of a method for dynamic modeling of carbon emissions from the spatiotemporal characteristics of power topology according to an embodiment of the present invention;

[0036] Figure 2 This is a flowchart illustrating the sparrow search algorithm for solving a dynamic modeling method for carbon emissions based on the spatiotemporal characteristics of power topology, as described in one embodiment of the present invention. Detailed Implementation

[0037] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0038] Example 1, referring to Figures 1-2 As one embodiment of the present invention, this embodiment provides a method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology, such as... Figure 1 As shown, it includes:

[0039] S101: Obtain the power consumption data of the power grid and the carbon emission data corresponding to the power consumption, and calculate the carbon emission factor of each node in the power grid;

[0040] S102: Based on the carbon emission factors of each node in the power grid, information data from different data sources are fused to construct a spatiotemporal correlation matrix;

[0041] S103: Optimize the node type weights and topological connection weights of the spatiotemporal correlation matrix, take the optimized spatiotemporal matrix as input, perform feature extraction to obtain a feature vector that integrates spatiotemporal characteristics, and map it into a dynamic carbon emission factor.

[0042] It should be noted that traditional carbon emission tracking methods based on power flow analysis mainly construct nodal carbon potential equations to allocate carbon emissions generated on the power generation side to the load side according to power ratios. Examples include proportional sharing methods and marginal emission intensity methods. The operational mechanism of these methods is highly dependent on the results of power flow calculations. In practice, iterative methods are needed to solve nonlinear equations repeatedly. As the power grid scale continues to expand and the number of nodes exceeds 1000, the computation time increases significantly. Genetic algorithms, particle swarm optimization, and other swarm intelligence algorithms have gradually been introduced into the research field of carbon emission optimization problems. These algorithms generate corresponding scheduling schemes by carefully constructing multi-objective functions, such as incorporating objectives like minimizing carbon emissions and minimizing power generation costs. However, these methods also reveal significant shortcomings in practical applications. In high-dimensional and non-convex solution spaces, the algorithms easily get trapped in local optima and cannot find the global optimal solution. Furthermore, these methods lack the ability to explicitly model the spatiotemporal dynamic characteristics of the power grid, which greatly limits the accuracy of the final carbon emission allocation. The actual error often exceeds 5%, making it difficult to meet the high requirements for accurate carbon emission calculations.

[0043] This invention addresses the shortcomings of traditional static modeling, which suffers from insufficient accuracy. Existing methods rely on fixed emission factors and static parameters, failing to respond in real-time to dynamic changes in the power grid topology. Parameter tuning is inefficient, as traditional hyperparameter optimization relies on grid search or manual experience and is prone to getting trapped in local optima, increasing model deployment costs. Real-time monitoring speed is also insufficient; classification modeling based on traditional machine learning (such as SVM) cannot meet the demands of second-level real-time carbon flow tracking. This invention provides a sparrow search algorithm-driven carbon emission modeling approach. Through intelligent parameter optimization and dynamic weight allocation, it solves the problems of low parameter tuning efficiency and subjective weight setting in traditional methods, improving both model accuracy and computational efficiency.

[0044] In this embodiment of the invention, step S101, calculating the carbon emission factor of each node in the power grid, includes:

[0045] The carbon emission factor for any node i in the power grid is determined by the active power output P of the local power plant.g and its carbon emission coefficient δ g And the active power P input from other nodes j to node i. ij And the carbon emission factor δ at the input node j A weighted average is performed to obtain the preliminary carbon emission factor.

[0046] Specifically, the carbon emission factor is calculated as follows:

[0047]

[0048] Where, δ i P represents the carbon emission coefficient of the electricity generation at the i-th node; g δ represents the active power output of the power plant connected to the i-th consumer node; g It is the carbon emission coefficient of the power plant; P ij Ω represents the active power input from the j-th node to the i-th node. i Let δ be the set of branch connection nodes of the i-th node; j It is the carbon emission factor of the j-th node.

[0049] In this embodiment of the invention, step S101 further includes: for any node i connected to all power sources k, its output P k Multiplied by the weight a of the power type k and carbon emission intensity ρ k ;

[0050] For all other nodes j connected to any node i, its transmission power P ji Multiplied by the topological weight β of the branch from node j to node i ji And the carbon emission factor of node j;

[0051] Carbon emission intensity per unit load includes the load at node i. and all input power P ji .

[0052] Specifically, the carbon emission calculation formula after adding weights is expressed as follows:

[0053]

[0054] Among them, G i P represents the set of power sources connected to node i; k Indicates the output power of power source k; a k Represents the type weight of power source k; ρ k N represents the carbon emission intensity of power source k; i Represents the set of other nodes connected to node i; β ji P represents the topological weight from branch j to i; jiThis represents the transmission power from branch j to i; This represents the load of node i.

[0055] In this embodiment of the invention, step S102, which involves fusing information data from different data sources based on the carbon emission factors of each node in the power grid to construct a spatiotemporal correlation matrix, includes:

[0056] The dimensions of the spatiotemporal correlation matrix include spatial dimension, time slice, and feature dimension;

[0057] By integrating the distribution of user load in time and space, the time series of power generation and consumption, and the grid topology, a high-dimensional data structure of the power system's operating status is constructed.

[0058] It should be noted that the construction of the spatiotemporal correlation matrix integrates the distribution of user load in the spatiotemporal dimension reflected by marketing data, the time series of power generation and consumption reflected by metering data, and the grid topology connection relationship shown by dispatch data. The dimensions of this matrix include nodes (representing the spatial dimension), time slices, and features including power, carbon emission intensity, and topology weights.

[0059] Furthermore, when constructing the spatiotemporal correlation matrix, the initial values ​​for weight allocation are set as follows: the weight of the thermal power node is 1.0, the weight of the wind power node is 0.3, and the weight of the photovoltaic node is determined to be 0.2.

[0060] It should be noted that the initial weights were set after comprehensively considering the characteristics of various energy sources and their different roles in the power grid, thus providing a foundation for subsequent matrix-based analysis and calculations.

[0061] In this embodiment of the invention, step S102 further includes: the spatiotemporal correlation matrix contains carbon emission factor information and weight settings for different types of power sources.

[0062] In this embodiment of the invention, the optimization of the node type weights and topological connection weights in the spatiotemporal correlation matrix in step S103 includes:

[0063] A vector containing node type weights and topology connection weights is randomly generated, and the weights are adjusted according to a preset security threshold.

[0064] Based on the weight adjustment results, the data is aggregated towards the optimal solution to obtain the optimized weights. The optimized weights are then written back to the spatiotemporal correlation matrix to update the data.

[0065] In a preferred embodiment, the spatiotemporal matrix weight optimization based on the sparrow algorithm randomly generates a weight vector, including node type weights to distinguish the importance of nodes and topology connection weights to measure the grid connectivity. Discoverers adjust their own weights according to a preset safety threshold to ensure stable system operation. Followers aggregate towards the optimal solution based on the discoverer's results, accelerating the search. Individual perturbation weights are randomly generated from the population to break local optima and add search directions and diversity. After optimizing to obtain the weight w*, it is written back to the spatiotemporal matrix to update the data, laying the foundation for the next round of calculation and analysis.

[0066] It should be noted that the design of the Sparrow Search Algorithm (SSA) is primarily inspired by the foraging and anti-predation behaviors of sparrows. During foraging, sparrows employ two different behavioral strategies: acting as discoverers and joining other groups. Simultaneously, sparrows remain highly vigilant of their surroundings, constantly wary of potential predators. Discoverers, also known as predators, are responsible for finding food within the sparrow population, determining the foraging area and direction for the entire group. When the environment is safe (i.e., the warning value R2 is less than the safety value ST), the discoverer will conduct a large-scale search; conversely, when the environment is dangerous (i.e., R2 is greater than or equal to ST), the discoverer will lead the entire flock to a safe area.

[0067] In the iterative process of the Sparrow Search Algorithm (SSA), the discoverer typically uses a fixed weight when updating its position. However, this approach has drawbacks, easily causing the algorithm to converge prematurely to a local optimum, making it difficult to obtain the global optimum. To overcome this problem, we introduce a non-linear decreasing weight method to optimize the discoverer's position update mechanism. Through this optimization, the algorithm can explore the solution space more thoroughly, effectively avoiding getting trapped in local optima.

[0068] Specifically, after introducing nonlinear decreasing weights, the discoverer position update equation can be expressed as:

[0069]

[0070] Where ω is the update weight, and t represents the current iteration number. As the iteration number increases, the weight gradually decreases non-linearly. That is, in the later stages of the search, the scope of the discoverer gradually narrows, shifting to a more refined local search, thus enabling it to more accurately locate the vicinity of the global optimum. Iter represents the position information of the i-th sparrow in the j-th dimension. max Let α be the maximum number of iterations, Q be random numbers in the range [0,1] following a normal distribution, and L be the identity matrix.

[0071] Joiners, also called followers, obtain food resources by relying on discoverers. During foraging, joiners closely follow the discoverer to the location of food, and depending on changes in the situation, they may transition from joiners to discoverers. The position update method for joiners is related to the optimal position determined by the current discoverer and the worst-case position globally within the entire group. The specific position update formula is as follows:

[0072]

[0073] Among them, X worst This is the worst position globally.

[0074] The watchdog plays a crucial role in monitoring its surroundings to prevent the intrusion of potential threats such as predators. During its mission, once the watchdog detects danger, it swiftly and promptly issues an alarm signal. This alarm guides the entire sparrow colony to quickly escape the danger zone, ensuring the safety of the group. The watchdog's own position is determined according to specific rules and updated using a particular formula, as follows:

[0075]

[0076] The difference between the weight vectors in two adjacent iterations is calculated using Euclidean distance:

[0077]

[0078] Among them, w (k) and w (k-1) These represent the weight vectors for the k-th and (k-1)-th iterations, respectively.

[0079]

[0080] in, N is the population mean. pop Population size.

[0081] If D avg A value less than γ indicates insufficient population diversity, which may lead to a local optimum.

[0082] Reference Figure 2 In an optional embodiment, the SSA algorithm's dynamic carbon emission modeling calculation process includes steps 1 to 10:

[0083] Step 1: Input the unit type, load value, and active power output, and initialize the weights w and α according to the unit type and grid topology;

[0084] Step 2: Randomly generate a weight vector;

[0085] Step 3: Integrate the spatiotemporal distribution of user load, the time series of power generation and consumption, and the grid topology to construct a spatiotemporal correlation matrix;

[0086] Step 4: Introduce a non-linear decreasing weight method to optimize the location update mechanism of the discoverer, and at the same time use (3) to update the location of the discoverer;

[0087] Step 5: Followers gather toward the optimal solution based on the discoverer's results, speeding up the search. They will update their positions using (4) according to changes in the actual situation.

[0088] Step 6: Once the vigilant detects the presence of danger, they will quickly and promptly issue an alarm signal. The vigilant will use (5) to update their location.

[0089] Step 7: Randomly generate individual perturbation weights from the population to break local optima.

[0090] Step 8: After optimizing and obtaining the weight w*, write it back to the spatiotemporal matrix.

[0091] Step 9: Based on the optimized matrix, the spatial topological features and temporal evolution patterns are extracted simultaneously using a spatiotemporal graph convolutional network. The attention mechanism is used to focus on key nodes (such as high-weight thermal power units), and a feature vector that integrates spatiotemporal characteristics is output. The feature vector is mapped to a dynamic carbon emission factor, and carbon emission prediction results are generated by combining real-time power data.

[0092] Step 10: Use steps (6) and (7) to determine convergence. If convergence is achieved, stop the iteration; otherwise, return to step 4.

[0093] In this embodiment of the invention, step S103, which uses the optimized spatiotemporal matrix as input, performs feature extraction to obtain a feature vector that integrates spatiotemporal characteristics, and maps it to a dynamic carbon emission factor, includes:

[0094] Based on the optimized matrix, a spatiotemporal graph convolutional network is used to simultaneously extract spatial topological features and temporal evolution patterns. Then, an attention mechanism is used to focus on key nodes and output a feature vector that integrates spatiotemporal characteristics. The feature vector is mapped to a dynamic carbon emission factor and combined with real-time power data to generate carbon emission prediction results.

[0095] In an optional embodiment, the optimized spatiotemporal matrix is ​​used as input, and spatial topological features can be extracted by adjacency matrix aggregation.

[0096] In another alternative embodiment, the optimized spatiotemporal matrix is ​​used as input, and the temporal evolution pattern can be obtained through sliding window convolution.

[0097] It should be noted that this invention addresses the shortcomings of traditional static modeling, such as insufficient accuracy, inability to respond in real time to dynamic changes in the power grid topology, and reliance on fixed weight factors and static parameters in existing methods. Compared to traditional hyperparameter optimization, which relies on grid search or manual experience and is prone to getting trapped in local optima, this invention introduces a nonlinear decreasing weight method, effectively avoiding the predicament of getting trapped in local optima and improving computational accuracy and efficiency.

[0098] Example 2 is an embodiment of the present invention. This embodiment differs from the first embodiment in that it provides a dynamic modeling system for carbon emissions based on the spatiotemporal characteristics of power topology, comprising:

[0099] The carbon emission factor calculation module is used to obtain the power generation data of the power grid and the carbon emission data corresponding to the power consumption, and to calculate the carbon emission factor of each node in the power grid.

[0100] The data fusion module is used to fuse information data from different data sources based on the carbon emission factors of each node in the power grid, and construct a spatiotemporal correlation matrix.

[0101] The optimization module is used to optimize the node type weights and topological connection weights of the spatiotemporal correlation matrix. The optimized spatiotemporal matrix is ​​used as input to extract features and obtain feature vectors that integrate spatiotemporal characteristics, and then map them into dynamic carbon emission factors.

[0102] Specifically, when the modules of the power topology spatiotemporal feature carbon emission dynamic modeling system of this embodiment are executed, they implement the steps of the power topology spatiotemporal feature carbon emission dynamic modeling method in Embodiment 1, for example:

[0103] In one implementation, the power topology spatiotemporal characteristic carbon emission dynamic modeling system can perform the following steps:

[0104] The carbon emission factor for any node i in the power grid is determined by the active power output P of the local power plant. g and its carbon emission coefficient δ g And the active power P input from other nodes j to node i. ij And the carbon emission factor δ at the input node j A weighted average is performed to obtain the preliminary carbon emission factor.

[0105] For any node i connected to all power sources k, its output P k Multiplied by the weight a of the power type k and carbon emission intensity ρ k ;

[0106] For all other nodes j connected to any node i, its transmission power P ji Multiplied by the topological weight β of the branch from node j to node i jiAnd the carbon emission factor of node j;

[0107] Carbon emission intensity per unit load includes the load at node i. and all input power P ji .

[0108] The dimensions of the spatiotemporal correlation matrix include spatial dimension, time slice, and feature dimension;

[0109] By integrating the distribution of user load in time and space, the time series of power generation and consumption, and the grid topology, a high-dimensional data structure of the power system's operating status is constructed.

[0110] The spatiotemporal correlation matrix contains information on carbon emission factors and weight settings for different types of power sources.

[0111] A vector containing node type weights and topology connection weights is randomly generated, and the weights are adjusted according to a preset security threshold.

[0112] Based on the weight adjustment results, the data is aggregated towards the optimal solution to obtain the optimized weights. The optimized weights are then written back to the spatiotemporal correlation matrix to update the data.

[0113] Based on the optimized matrix, a spatiotemporal graph convolutional network is used to simultaneously extract spatial topological features and temporal evolution patterns. Then, an attention mechanism is used to focus on key nodes and output a feature vector that integrates spatiotemporal characteristics. The feature vector is mapped to a dynamic carbon emission factor and combined with real-time power data to generate carbon emission prediction results.

[0114] Example 3: This example provides an electronic device applicable to the dynamic modeling method of carbon emissions from the spatiotemporal characteristics of power topology, including:

[0115] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the dynamic modeling method for the spatiotemporal characteristics of carbon emissions in power topology, as proposed in the above embodiments.

[0116] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as proposed in the above embodiments.

[0117] The storage medium proposed in this embodiment and the method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0118] From the above description of the embodiments, those skilled in the art will clearly understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0119] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0120] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0122] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0123] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0124] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology, characterized in that, include: Acquire the power consumption data of the power grid and the carbon emission data corresponding to the power consumption, and calculate the carbon emission factor of each node in the power grid; Based on the carbon emission factors of each node in the power grid, information data from different data sources are fused to construct a spatiotemporal correlation matrix; The node type weights and topological connection weights of the spatiotemporal correlation matrix are optimized. The optimized spatiotemporal matrix is ​​used as input to extract features and obtain a feature vector that integrates spatiotemporal characteristics. This feature vector is then mapped to a dynamic carbon emission factor.

2. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 1, characterized in that, The carbon emission factors of each node in the calculated power grid include: The carbon emission factor for any node i in the power grid is determined by the active power output P of the local power plant. g and its carbon emission coefficient δ g And the active power P input from other nodes j to node i. ij And the carbon emission factor δ at the input node j A weighted average is performed to obtain the preliminary carbon emission factor.

3. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 2, characterized in that, It also includes: for any node i, the output P of all power sources k connected to it. k Multiplied by the weight a of the power type k and carbon emission intensity ρ k ; For all other nodes j connected to any node i, its transmission power P ji Multiplied by the topological weight β of the branch from node j to node i ji And the carbon emission factor of node j; Carbon emission intensity per unit load includes the load at node i. and all input power P ji .

4. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 3, characterized in that, Based on the carbon emission factors of each node in the power grid, information data from different data sources are fused to construct a spatiotemporal correlation matrix, including: The spatiotemporal correlation matrix includes spatial dimension, time slice and feature dimension. By integrating the distribution of user load in time and space, the time series of power generation and consumption, and the grid topology, a high-dimensional data structure of the power system's operating status is constructed.

5. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 4, characterized in that, Also includes: The spatiotemporal correlation matrix includes carbon emission factor information and weight settings for different types of power sources.

6. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 5, characterized in that, Optimizing the node type weights and topological connection weights of the spatiotemporal correlation matrix includes: A vector containing node type weights and topology connection weights is randomly generated, and the weights are adjusted according to a preset security threshold. Based on the weight adjustment results, the data is aggregated towards the optimal solution to obtain the optimized weights. The optimized weights are then written back to the spatiotemporal correlation matrix to update the data.

7. The method for dynamic modeling of carbon emissions based on the spatiotemporal characteristics of power topology as described in claim 6, characterized in that, The optimized spatiotemporal matrix is ​​used as input for feature extraction to obtain a feature vector that integrates spatiotemporal characteristics, and this vector is then mapped to a dynamic carbon emission factor, including: Based on the optimized matrix, a spatiotemporal graph convolutional network is used to simultaneously extract spatial topological features and temporal evolution patterns. Then, an attention mechanism is used to focus on key nodes and output a feature vector that integrates spatiotemporal characteristics. The feature vector is mapped to a dynamic carbon emission factor and combined with real-time power data to generate carbon emission prediction results.

8. A dynamic modeling system for carbon emissions based on the spatiotemporal characteristics of power topology, applied to the method described in any one of claims 1-7, characterized in that, include: The carbon emission factor calculation module is used to acquire the power consumption data of the power grid and the carbon emission data corresponding to the power consumption, and to calculate the carbon emission factor of each node in the power grid. The data fusion module is used to fuse information data from different data sources based on the carbon emission factors of each node in the power grid, and construct a spatiotemporal correlation matrix. The optimization module is used to optimize the node type weights and topological connection weights of the spatiotemporal correlation matrix. The optimized spatiotemporal matrix is ​​used as input to extract features and obtain a feature vector that integrates spatiotemporal characteristics, and then it is mapped to a dynamic carbon emission factor.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the dynamic modeling method for carbon emissions of spatiotemporal characteristics of power topology as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the dynamic modeling method for spatiotemporal carbon emissions of power topology as described in any one of claims 1 to 7.

Images