Power grid carbon emission traceability method and device based on space-time matching and medium

Abstract

The invention relates to the cross technical field of electric power system monitoring and information technology, in particular to a power grid carbon emission traceability method and device based on space-time matching and a medium, which form a spatial topology expression by constructing a corresponding model of geographic coordinates and an electric power system network structure, and realize accurate mapping of carbon emission data to power grid nodes; multi-dimensional heterogeneous data is introduced, a unified data processing mechanism is established, and the data integration and standardization capability is improved; a response relation model between a time sequence and a space unit is established based on a data set, cross-scale matching is completed through a grid planning method, and the problem that carbon emission statistics and a power grid structure are disjointed is solved; and finally, combining the correlation structure, performing data fusion through scale conversion and weight superposition mechanisms, and outputting a fused data structure with spatial positioning and time continuity, thereby providing support for carbon emission path identification, power grid dispatching optimization and visual presentation.

Description

Technical Field

[0001] This invention relates to the field of power system monitoring and information technology, and in particular to a method, equipment and medium for tracing carbon emissions from power grids based on spatiotemporal matching. Background Technology

[0002] In the context of current energy structure transformation, the power industry, as a key sector for carbon emissions, directly impacts the achievement of carbon neutrality goals through its emission reduction efforts. With the increasing proportion of renewable energy generation and growing electricity demand, the decarbonization of the power system faces complex challenges. Carbon emission source tracing technology, as a core component, needs to cover the entire process from traditional energy combustion, renewable energy generation, and grid transmission and distribution. Existing source tracing models mainly include computational models based on physical processes, data-driven predictive models, and hybrid models combining both, aiming to quantify carbon emissions by simulating power generation methods, load characteristics, and grid operation. However, bottlenecks remain in building a unified data management system and standardized database, particularly in the integration of multi-source data (such as power plant emissions, grid load, and meteorological information) and the standardization of cross-regional carbon emission measurement, which urgently require breakthroughs.

[0003] However, existing technologies have core flaws: regional statistics are severely disconnected from power grid physical models, making it impossible to accurately correlate regional carbon emissions with power grid node data. Furthermore, it is difficult to match power generation, load, and carbon emission data across different spatiotemporal dimensions; there is a lack of spatiotemporal difference analysis capabilities between real-time and historical data, making it impossible to dynamically track anomalies or trend changes; the spatiotemporal interactions of multiple energy systems (such as wind power, photovoltaics, and energy storage) are not considered collaboratively, and heterogeneous data sources (electricity, transportation, meteorology, etc.) are difficult to effectively integrate due to differences in spatiotemporal scales, thus limiting the accuracy and timeliness of carbon emission tracing. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a spatiotemporal matching-based method for tracing carbon emissions from power grids to solve the core technical problems of the disconnect between regional statistics and the physical model of the power grid, and the failure of multi-source spatiotemporal data matching in the tracing of carbon emissions from power systems.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for tracing carbon emissions from power grids based on spatiotemporal matching, comprising:

[0008] A model corresponding to geographic coordinates and power system network structure is constructed to obtain a spatial topological representation;

[0009] Based on spatial topological representation, multidimensional heterogeneous data is introduced, and a data processing mechanism is established to generate datasets through standardized merging methods;

[0010] Using the dataset, a response relationship model between time series and spatial units is established. Through grid planning, cross-scale matching operations are completed to form an association structure.

[0011] Based on the association structure, data fusion is performed through scale transformation and weight superposition mechanisms, and the fused data structure is output.

[0012] As a preferred embodiment of the power grid carbon emission tracing method based on spatiotemporal matching described in this invention, the step of constructing a correspondence model between geographical coordinates and the power system network structure to obtain a spatial topology expression includes:

[0013] Determine the spatial information of nodes in the power grid structure and construct a spatial coordinate system;

[0014] Identify the constituent units of the power grid and establish their mapping relationships in a spatial coordinate system;

[0015] Based on the power grid connection relationships, a connection structure between nodes is constructed to form a spatial topology hierarchy;

[0016] By jointly representing spatial information and structural relationships, a spatial topological expression with locational and structural characteristics is formed.

[0017] As a preferred embodiment of the spatiotemporal matching-based power grid carbon emission tracing method described in this invention, the method involves: introducing multidimensional heterogeneous data based on spatial topology representation and establishing a data processing mechanism to generate a dataset through standardized merging, including:

[0018] Collect data related to multiple power grid sources and construct a unified data access format;

[0019] Establish a data processing mechanism to handle differences in data types and inconsistencies in structure, and perform cleaning and formatting operations on the input data;

[0020] Based on the data processing mechanism, a heterogeneous data alignment strategy is applied to the collected data to form a standardized dataset.

[0021] As a preferred embodiment of the power grid carbon emission tracing method based on spatiotemporal matching described in this invention, the step of establishing a response relationship model between time series and spatial units using a dataset includes:

[0022] The standardized dataset is divided into data partitions according to time and spatial location.

[0023] Construct spatiotemporal association rules and define association rules between different data partitions;

[0024] Based on association rules, a response relationship model between time series and spatial units is generated.

[0025] The beneficial effects of this preferred technical solution are as follows: by constructing a spatiotemporal coupling model based on standardized data partitioning, a precise correspondence between time series and spatial units is achieved, which can dynamically depict the spatiotemporal evolution of power system operating status and carbon emission factors; this model not only improves the accuracy and consistency of data matching, but also enhances the integration capability of multi-source data, effectively supports cross-scale data fusion and the establishment of carbon emission source tracing paths, and provides a more solid data support and structural foundation for subsequent scheduling optimization and carbon emission analysis.

[0026] As a preferred embodiment of the power grid carbon emission tracing method based on spatiotemporal matching described in this invention, the step of completing cross-scale data matching operations and forming an association structure through grid planning includes:

[0027] A spatiotemporal grid framework is established by dividing the standardized dataset into grids based on time and space dimensions.

[0028] Matching and comparing data at different time scales, extracting features, and establishing correspondences between time series data are represented as follows:

[0029]

[0030] Where, x i and y j Let i and j represent the i-th and j-th points of two time series data, respectively. DTW(i,j) represents the matching degree between the corresponding time points, i-1 represents the matching degree between the (i-1)-th point and the j-th point, and DTW represents the function used for dynamic time warping. The minimum matching degree among the data points near the two time series data points is the matching degree of the two time series data points.

[0031] Interpolation and scaling operations are performed on the spatial data to improve the spatial distribution information within the grid cells, as shown below:

[0032]

[0033] Among them, Z * (s0) is the estimated value of point s0, Z(s) i ) represents the value of a known data point, d(s0,s) i ) represents the estimated distance between the point and the known point, and p is the power exponent;

[0034] By combining the matching results of time and space dimensions, a data connection structure for cross-scale mapping features is constructed as the association structure for fusion operations.

[0035] The beneficial effects of this preferred technical solution are as follows: By constructing a unified spatiotemporal grid framework and integrating dynamic time warping and spatial interpolation methods, accurate matching and coordination between data at different time and spatial scales are achieved, effectively solving the inconsistency problem of multi-source data in terms of resolution, sampling frequency and spatial coverage; this cross-scale data connection structure improves the continuity and integrity of the data, providing a high-resolution, spatiotemporally coupled data foundation for subsequent carbon emission factor tracing, dynamic evolution analysis and scheduling response, thereby enhancing the accuracy and practicality of carbon emission analysis.

[0036] As a preferred embodiment of the spatiotemporal matching-based power grid carbon emission tracing method described in this invention, the heterogeneous data alignment strategy includes:

[0037] Multidimensional power grid operation data from power generation, load, and dispatching are categorized by source and their spatiotemporal features are extracted.

[0038] Multidimensional data is standardized and represented as follows:

[0039]

[0040] Among them, z i It is the standard deviation, x i These are the original data, μ is the mean of the data, and σ is the standard deviation;

[0041] Based on time index and spatial location index, the processed data is matched to establish the correspondence between multiple types of heterogeneous data in the spatiotemporal grid.

[0042] The aligned data is organized into a standardized dataset for response modeling and scale fusion operations.

[0043] As a preferred embodiment of the power grid carbon emission tracing method based on spatiotemporal matching described in this invention, the spatiotemporal association rules include:

[0044] The standardized data is segmented according to time order and different time periods are allocated; comparison rules are used to describe the time change process and identify the temporal characteristics between adjacent time periods.

[0045] Analyze data pairs from different time periods, determine comparison methods, and identify relationships between the data through these comparison methods.

[0046] As a preferred embodiment of the power grid carbon emission tracing method based on spatiotemporal matching described in this invention, the step of performing data fusion based on the correlation structure through scale transformation and weight superposition mechanisms, and outputting the fused data structure, includes:

[0047] A scale transformation function is constructed by fusing data at different time and spatial scales.

[0048] Based on the cross-matching results of spatial and temporal scales, a weighted fusion model is constructed, represented as:

[0049]

[0050] Among them, F 融合 F represents the fusion result. i For features at different scales, w i These are the corresponding weighting coefficients;

[0051] Based on the grid mapping relationship in the association structure, multi-source data under the same grid cell are collaboratively weighted and integrated; the output fused data structure retains the key information of carbon emission intensity and matching path of each grid cell.

[0052] The beneficial effects of this preferred technical solution are as follows: by constructing a scale transformation function and introducing a weighted fusion model, it can effectively unify the data expression methods under different time and spatial scales, and improve the fusion compatibility and matching accuracy between multi-source heterogeneous data; combined with the grid mapping relationship in the association structure, it realizes the collaborative weighted integration of multi-dimensional data within the same grid unit, thereby constructing a fusion data structure with continuity, accuracy and physical interpretability; this structure retains key elements such as carbon emission intensity and path information, providing more refined support and basis for carbon flow identification, source tracing and scheduling decisions.

[0053] In a second aspect, the present invention provides an electronic device, comprising:

[0054] Memory and processor;

[0055] 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 power grid carbon emission traceability method based on spatiotemporal matching.

[0056] Thirdly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the spatiotemporal matching-based power grid carbon emission tracing method.

[0057] Compared with existing technologies, the beneficial effects of this invention are as follows: By constructing a correspondence model between geographic coordinates and power system network structure, this invention establishes an expression method with spatial topological features, enabling carbon emission data to be accurately mapped to specific power grid nodes and lines, achieving precise location of carbon emission sources. By introducing multidimensional heterogeneous data and constructing a unified data processing mechanism, it solves the problem of difficulty in uniformly processing data from different sources and in different formats, improving the standardization and integration efficiency of data processing. Furthermore, by using a standardized dataset to establish a response relationship model between time series and spatial units, and by using a grid planning method to achieve cross-scale data matching, it solves the problem of scale inconsistency between macroscopic carbon emission statistics and microscopic power grid physical structure, effectively supporting the linkage modeling of power carbon data. Finally, based on the constructed association structure, data fusion is achieved through scale transformation and weight superposition mechanisms, outputting a fused data structure with continuity, spatial locationability, and temporal consistency. This fusion result can not only be used to identify high-carbon emission nodes and transmission paths, but also support power system scheduling optimization and visualization of carbon emission paths, thereby significantly improving the refinement and practicality of power grid carbon emission management. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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.

[0059] Figure 1 This is a schematic diagram of the overall process of the power grid carbon emission tracing method based on spatiotemporal matching according to an embodiment of the present invention. Detailed Implementation

[0060] 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.

[0061] Example 1, referring to Figure 1 As one embodiment of the present invention, a method for tracing carbon emissions from power grids based on spatiotemporal matching is provided, comprising:

[0062] S1: Construct a model of the correspondence between geographic coordinates and power system network structure to obtain a spatial topology representation;

[0063] S2: Based on spatial topology representation, multidimensional heterogeneous data is introduced, and a data processing mechanism is established to generate a dataset through standardized merging.

[0064] S3: Using the dataset, establish a response relationship model between time series and spatial units, and complete cross-scale matching operations of the data through grid planning methods to form an association structure;

[0065] S4: Based on the association structure, data fusion is performed through scale transformation and weight superposition mechanism, and the fused data structure is output.

[0066] It should be noted that the current process of tracing carbon emissions in the power system faces key problems such as heterogeneous multi-source data, inconsistent spatiotemporal scales, and disconnect between physical models and statistical data. In particular, given the complex spatial distribution and dynamic changes in power dispatch, it is difficult to achieve accurate tracking and correlation modeling of carbon emission flows.

[0067] Therefore, addressing the challenges of diverse carbon emission data sources, discrete distribution, significant temporal granularity differences, and difficulties in unifying spatial topology representation, this paper constructs an integrated solution path through steps S1-S4, encompassing geographic coordinates, power grid topology, multi-source data fusion, and cross-scale response modeling. By mapping the GIS coordinate system to the power grid structure, an expression model with spatial topological attributes is established, providing a foundation for data location and power grid node mapping. Heterogeneous data is introduced based on spatial topology representation, and a unified dataset is generated through a standardized merging mechanism. Response relationships between time series and spatial units are constructed, and efficient cross-scale data matching is achieved through grid planning, establishing a correlation structure with spatiotemporal coupling characteristics. Based on this correlation structure, multi-source data fusion output is achieved through scale transformation and weight superposition mechanisms, providing data support for subsequent carbon emission source tracing modeling and dynamic visual analysis.

[0068] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the above embodiment, a method for tracing carbon emissions from power grids based on spatiotemporal matching is provided.

[0069] In this embodiment of the application, step S1 involves constructing a correspondence model between geographic coordinates and the power system network structure to obtain a spatial topology representation, including:

[0070] S11: Determine the spatial information of nodes in the power grid structure and construct a spatial coordinate system;

[0071] S12: Identify the constituent units of the power grid and establish a mapping relationship in the spatial coordinate system;

[0072] S13: Construct the connection structure between nodes based on the power grid connection relationship to form a spatial topology hierarchy;

[0073] S14: Jointly represent spatial information and structural relationships to form a spatial topological expression with locational and structural characteristics.

[0074] It should be noted that by constructing a spatial topology representation, the organic integration of the power grid's physical structure and spatial geographic information is achieved, solving the problem of the lack of spatial correspondence between traditional carbon emission data and power grid nodes. The spatial topology representation can accurately depict the geographical distribution characteristics of each power node and its connection relationship, providing basic support for the location mapping of carbon emission factors in the power grid. It improves the spatial accuracy of subsequent data fusion and path identification, enabling carbon emission source tracing to have geographical location and network structure interpretability, effectively supporting the identification of high carbon emission areas, the visualization of power grid operation status, and carbon allocation and optimization analysis at the scheduling level.

[0075] Specifically, S11 to S14 include GIS coordinate transformation using the WGS84 ellipsoid model; the WGS84 coordinate system uses latitude and longitude coordinates, while the UTM coordinate system is based on the transverse Mercator projection.

[0076] Specifically, latitude and longitude converted to radians is expressed as follows:

[0077]

[0078] The calculation of the projection zone number involves dividing the UTM into multiple projection zones, each with a width of 6°, and dividing the longitude from -180° to +180° into 60 zones; the zone number is determined by longitude and is represented as follows:

[0079]

[0080] The central meridian (longitude of the central meridian in UTM) is represented as:

[0081] central meridian=(zone number-1)×6-180+3

[0082] Calculating the projected coordinates involves converting them to projected coordinates using the transverse Mercator projection formula;

[0083] The calculation of N, T, C, A, M, etc., is expressed as follows:

[0084]

[0085] The calculation of the horizontal (Easting) and vertical (Northing) coordinates is expressed as follows:

[0086] E = E0 + N·(A + (1-T + C)·B) 2 )

[0087] The final spatial topology of the transformation is expressed as UTM projection coordinates (X, Y) and projection zone number;

[0088] Where latitude rad represents latitude in radians, longitude rad represents longitude in radians, and zone number represents the projection zone number in UTM coordinates; centralmeridian represents the longitude of the central meridian of the UTM zone in UTM coordinates; N is the longitudinal coordinate, a is the semi-major axis of the Earth, e is the eccentricity of the Earth, and φ is latitude; E is the transverse coordinate, E0 is the longitude of the central meridian, A is the product of the longitude difference in radians and the latitudinal scaling, T is the square of the latitude tangent, C is the eccentricity correction term, and B is the calculated latitude-longitude difference.

[0089] Specifically, S11 to S14 also include the matching of GIS coordinates with power grid topology and the fusion of GIS and power grid topology data;

[0090] The matching of GIS coordinates with the power grid topology includes equipment matching, which matches nodes in the power grid topology with equipment in the GIS coordinates; direct matching is performed through equipment identifiers (such as substation numbers), or approximate matching is performed through geographical location; if multiple devices exist in the same location (such as multiple substations), they can be further distinguished based on other attributes of the devices (such as voltage level); line mapping: through the connectivity of coordinate data and the power grid topology, the transmission lines of the power grid are matched with the paths in the GIS coordinates; the starting and ending points of the lines are mapped to the actual geographical locations through the GIS coordinates of the equipment, forming a spatial model of the power grid;

[0091] Among them, the fusion of GIS and power grid topology data includes merging the GIS coordinate system with the power grid topology system to generate a comprehensive power grid model. In this model, GIS coordinates serve as the basis for node locations, while the power grid topology provides the power flow relationships between nodes. Through GIS technology, a map view can be provided for the power grid topology, helping power grid dispatchers to perform spatial analysis, monitoring, and fault location.

[0092] Specifically, the power grid topology consists of nodes (such as power plants and substations) and edges (transmission lines); in the power grid, the physical connections between nodes define the path of electricity flow; using power grid topology data and GIS coordinates, a specific spatial location is assigned to each node of the power grid (such as substations, load nodes, etc.) and matched with carbon emission data of the macro region.

[0093] In an optional implementation, the spatial topology representation in step S1 can also achieve dynamic topology modeling by introducing a linkage mechanism between a real-time geographic information system (real-time GIS) and a power grid SCADA system. Specifically, by mapping the node operating status (such as voltage, current, and switch status) collected in the SCADA system to the spatial coordinates corresponding to each node in the GIS in real time, a spatial topology structure that changes with the operating status can be dynamically constructed.

[0094] In another optional implementation, the spatial topology representation in step S1 can also be modified by multi-factor spatial correction through the overlay of topographic data and meteorological parameters. For example, in mountainous or coastal areas with complex topography, DEM (Digital Elevation Model) data can be acquired and height-corrected with the GIS coordinates of power grid equipment to optimize the spatial path representation of power lines. At the same time, the equivalent length of the transmission path can be dynamically adjusted by combining meteorological factors such as wind speed and temperature to more accurately reflect the differences in physical consumption or energy loss of the transmission path in carbon emission analysis.

[0095] In this embodiment of the application, step S2 introduces multidimensional heterogeneous data based on spatial topological representation and establishes a data processing mechanism to generate a dataset through a standardized merging method, including:

[0096] S21: Collect relevant data from multiple power grids and construct a unified data access format;

[0097] S22: Establish a data processing mechanism to handle differences in data types and inconsistencies in structure, and perform cleaning and formatting operations on the input data;

[0098] S23: Based on the data processing mechanism, perform a heterogeneous data alignment strategy on the collected data to form a standardized dataset.

[0099] Specifically, S21 to S23 involve preprocessing the spatiotemporal data to make it suitable for matching; common data preprocessing steps include data cleaning, filling in missing values, and denoising.

[0100] In an optional implementation, the data processing mechanism in step S2 can also introduce a rule-based template-based field mapping strategy to automatically identify and normalize the differences in field naming, unit expression, and semantics in data from different sources, so as to unify the expression of core indicators such as carbon emissions, power, and voltage in various data sources. For example, for power generation data from the dispatch center and the energy statistics system, “output power” and “unit electricity” are used respectively. By setting field mapping rules and unit conversion logic, the consistency of data fields and physical quantities can be achieved.

[0101] In another optional implementation, the data processing mechanism in step S2 can also introduce a time-series anomaly detection mechanism to identify and correct abrupt changes in the data of power grid nodes on the time axis, so as to avoid the impact of data drift on subsequent response relationship modeling. Specifically, in practical applications, some nodes may experience short-term load surges or carbon factor drops due to inaccurate acquisition equipment or communication delays. By setting sliding windows and dynamic threshold rules, combined with the historical statistical characteristics of nodes, the abnormal values ​​can be interpolated for repair or marked for removal.

[0102] In this embodiment of the application, step S23, based on the data processing mechanism, performs a heterogeneous data alignment strategy on the collected data to form a standardized dataset, including:

[0103] A1: Classify the multi-dimensional power grid operation data of power generation, load and dispatch side according to the source, and extract spatiotemporal features;

[0104] A2: Perform unified and standardized processing on multidimensional data;

[0105] A3: Based on time index and spatial location index, perform matching operations on the processed data to establish the correspondence between multiple types of heterogeneous data under the spatiotemporal grid;

[0106] A4: Organize the aligned data into a standardized dataset and perform response modeling and scale fusion operations.

[0107] Specifically, the formula for calculating A2 is as follows:

[0108]

[0109] Among them, z i It is the standard deviation, x i These are the original data, μ is the mean of the data, and σ is the standard deviation.

[0110] In an optional implementation, the heterogeneous data alignment strategy in step S23 can also be assisted by constructing a node adjacency matrix based on the power grid topology. Specifically, after performing spatiotemporal index matching on the generation side and dispatch side data, for some boundary node data with geographical offset or inconsistent sampling periods, an adjacency weight matrix is ​​constructed by introducing the physical connection relationship of the nodes, and the data of the deviating nodes is weighted and smoothed according to the principle of topological continuity, so as to improve the spatial consistency of heterogeneous data.

[0111] In another optional implementation, the heterogeneous data alignment strategy in step S23 can also optimize the data matching accuracy in the time dimension by combining dynamic time warping (DTW) with an adaptive adjustment mechanism for grid partitioning granularity. Before the standardized data enters the gridded mapping, the DTW window length and step size parameters are dynamically adjusted based on the time series stability of different power grid equipment or regional operation data. For rapidly changing areas of new energy output nodes or load sides with large fluctuations, the time slice granularity is refined and the local matching weight is enhanced.

[0112] In this embodiment of the application, step S3 utilizes the dataset to establish a response relationship model between time series and spatial units, and completes cross-scale data matching operations using a grid planning method to form an association structure, including:

[0113] S31: Divide the standardized dataset according to the time dimension and spatial location to establish data partitions;

[0114] S32: Construct spatiotemporal association rules and define association rules between different data partitions;

[0115] S33: Based on association rules, generate a response relationship model between time series and spatial units.

[0116] S34: Grid-divide the standardized dataset based on time and space dimensions to establish a spatiotemporal grid framework;

[0117] S35: Perform matching and comparison on data at different time scales, extract features, and establish correspondences between time series.

[0118] S36: Perform interpolation and scale unification operations on spatial data to improve the spatial distribution information within the grid cells;

[0119] S37: Combining the matching results of time and space dimensions, construct a data connection structure for cross-scale mapping features as the association structure for fusion operations.

[0120] It should be noted that by using a unified spatiotemporal grid framework, data with different temporal granularities and spatial distributions are effectively coordinated, enabling multi-source data with resolution differences and inconsistent coverage to be matched and fused under a unified structure. The dynamic time warping method improves the matching accuracy between time series data, while spatial interpolation and scale unification operations enhance the continuity and integrity of spatial data. Furthermore, by constructing a cross-scale data connection structure, a clear correspondence between the temporal change process and the spatial distribution state is established. The response relationship is optimized by combining a spatiotemporal regression model, and the mean square error and mean absolute error are used for verification to ensure the usability and accuracy of the matching structure in actual power grid scenarios.

[0121] Specifically, the S35 implementation formula is expressed as follows:

[0122]

[0123] Where, x i and y j Let i and j represent the i-th and j-th points of two time series data, respectively. DTW(i,j) represents the matching degree between the corresponding time points, i-1 represents the matching degree between the (i-1)-th point and the j-th point, and DTW represents the function used for dynamic time warping. The minimum matching degree among the data points near the two time series data points is the matching degree of the two time series data points.

[0124] The specific implementation formula for S36 is expressed as follows:

[0125]

[0126] Among them, Z * (s0) is the estimated value of point s0, Z(s) i ) represents the value of a known data point, d(s0,s) i ) represents the estimated distance between the point and the known point, and p is the power exponent;

[0127] Specifically, S31 to S37 also include methods for optimizing the objective using spatiotemporal regression models, expressed as:

[0128] y t =α+β1x t +β2z t +∈ t

[0129] The results of data matching are validated and evaluated. Evaluation metrics include the degree of matching (such as DTW, correlation coefficient) and the actual application effect of the matching results (such as prediction error, accuracy, etc.), expressed as follows:

[0130] Mean Square Error (MSE):

[0131]

[0132] Mean Absolute Error (MAE):

[0133]

[0134] Among them, y t It is the response variable, x t It's time data, z t This is spatial data, where α is a constant term, β1 and β2 are regression coefficients, ∈ t This is the error term; where y i This is the actual value. This is the predicted value, and n is the number of data points.

[0135] In an optional implementation, the response relationship model in step S3 can also be optimized by introducing a spatial adjacency matrix. By constructing a spatial adjacency matrix between power grid nodes, the topological connection relationship of each node is encoded into a structure vector and combined with the divided data partitions. When establishing a response relationship between the time series and spatial units, not only the time evolution trend of the data value itself is considered, but also its connectivity and neighborhood influence in the spatial structure are comprehensively considered. For example, before performing dynamic time warping matching, nodes with direct connection relationships or high topological similarity to the target node are preferentially screened.

[0136] In another optional implementation, the response relationship model in step S3 can also dynamically adjust the model parameters by introducing carbon emission patterns from historical operating scenarios. During the establishment of the spatiotemporal response model, a library of typical carbon emission evolution patterns is constructed based on historical power system dispatch data, load curves, and carbon emission factor records. In the actual modeling phase, the current standardized dataset is compared with historical patterns (e.g., using Pearson correlation coefficients), and the most similar historical pattern is selected as a reference template to initially adjust or constrain the regression model parameters (e.g., spatiotemporal regression coefficients).

[0137] In this embodiment of the application, step S32 involves constructing spatiotemporal association rules and defining association rules between different data partitions, including:

[0138] B1: The standardized data is segmented according to time order and different time periods are allocated; comparison rules are used to describe the time change process and identify the temporal characteristics between adjacent time periods;

[0139] B2: Analyze data pairs from different time periods, determine the comparison method, and identify the relationships between the data through the comparison method.

[0140] Specifically, B1 to B2 align different dimensions (such as regional data, power grid load data, carbon emission data, etc.) according to time based on the data timestamps, i.e., data time matching;

[0141] This refers to a technical method for establishing spatial correspondence between power system topology data and macro-regional carbon emission data based on a geospatial coordinate system, namely, spatial data matching.

[0142] In an optional implementation, the spatiotemporal correlation rules in step S32 can also be implemented through a segmented response matching mechanism based on load fluctuation trends. That is, based on the time series segmentation processing of standardized data, the time segments are dynamically labeled in combination with the changing trends of the load curve in different time periods (such as sudden rise, gradual change, and fall); then, for time periods with similar trend patterns, the corresponding carbon emission change paths are further extracted within the spatial unit to achieve trend-driven spatiotemporal matching and discrimination, thereby identifying the carbon emission response segments closely related to load changes.

[0143] In another optional implementation, the spatiotemporal association rule in step S32 can also be implemented through an adjacency grid reasoning mechanism based on the energy flow direction of nodes. That is, by using the connection relationship between nodes in the power grid topology, the actual path of electrical energy transmission in space is identified, the energy flow between upstream power generation nodes and downstream load nodes is mapped to the trajectory, and the trajectory is mapped to the transmission path of carbon emission factors. Then, combined with time series data, spatiotemporal pairing logic is constructed in adjacent grids to capture the delayed diffusion characteristics of carbon emission factors in space.

[0144] In this embodiment of the application, step S4, based on the association structure, performs data fusion through scale transformation and weight superposition mechanisms, and outputs the fused data structure, including:

[0145] S41: By fusing data at different time and spatial scales, a scale transformation function is constructed;

[0146] S42: Construct a weighted fusion model based on the cross-matching results of spatial and temporal scales;

[0147] S43: Based on the grid mapping relationship in the association structure, multi-source data under the same grid cell are collaboratively weighted and integrated; the output fused data structure retains the key information of carbon emission intensity and matching path of each grid cell.

[0148] It should be noted that by constructing a scale transformation function, the data discrepancies between different time scales (such as minute-level and hour-level) and spatial scales (such as node-level and regional-level) can be resolved, achieving equivalent data transformation under a unified computing framework and ensuring that various types of data have comparability and a basis for fusion. Simultaneously, the introduction of a weighted fusion model enables dynamic weighted integration of data from different scales according to their physical meaning, data quality, and spatiotemporal relevance, significantly improving the fusion compatibility and matching accuracy between multi-source heterogeneous data. Furthermore, through the grid mapping relationship in the association structure, data within the same grid cell is collaboratively integrated, maintaining not only the continuous expression of carbon emission intensity and physical pathways but also enhancing the spatial integrity and temporal consistency of the fusion results. The final output fused data structure possesses spatiotemporal consistency, supporting the tracking of carbon flow paths, carbon source allocation calculation, and the identification of carbon loads at key nodes.

[0149] Specifically, the S42 weighted fusion model is represented as follows:

[0150]

[0151] Among them, F 融合 F represents the fusion result. i For features at different scales, w i Its corresponding weighting coefficient.

[0152] Specifically, S41 to S43 also include the use of cross-scale correlation and fusion methods based on multi-scale analysis, that is, to fuse data at different time and spatial scales and to standardize the data so that they can be compared and analyzed on the same platform.

[0153] For example, suppose there are two scales of data, with a time scale of t and a spatial scale of x, and the data are D respectively. t (t) and D x (x), normalized representation is:

[0154]

[0155] Where, μ t ,μ x It is the mean of their respective datasets, σ t ,σ x It is the standard deviation. It is D t Differentiate (t) with respect to time t. It is D x (x) is differentiated with respect to spatial scale x;

[0156] Use methods based on convolutional neural networks (CNN) or graph convolutional networks (GCN) to capture cross-scale correlations and connect data at different scales;

[0157] For example, suppose there are two scales of input X. t and X x The fused data Z can then be processed using a weighted fusion method, represented as:

[0158] Z = w t ·X t +w x ·X x

[0159] Among them, w t and w x These are the weights of the corresponding scales, which determine the degree to which the two scales contribute to the final output; information from different scales is integrated through weighted averaging, splicing, or fusion.

[0160] In an optional implementation, the data fusion in step S4 can also be achieved through an iterative fusion method that introduces error feedback. That is, for the data structure after the initial fusion, the carbon emission intensity output by fusion is further compared with the actual observed value, and feedback correction is performed within the grid cell based on the observation error. Specifically, by introducing an error weight factor during the construction of the scale transformation function, the scale weight coefficient is dynamically adjusted according to the deviation result after each fusion, forming a feedback closed-loop mechanism for data fusion, so as to gradually reduce the error between the fusion result and the actual operating data, and make the output fusion data structure more in line with the actual needs of grid carbon emission traceability.

[0161] In another optional implementation, the data fusion in step S4 can also be achieved through a dynamic weight correction method based on typical operating conditions. Specifically, for typical operating conditions in the operation of the power system, such as high load periods and periods of large fluctuations in renewable energy output, characteristic operating condition databases are established respectively, and scale conversion and weight allocation rules under different operating conditions are predefined. In real-time operation, by identifying the operating condition type of the current spatiotemporal unit, the scale conversion function and weighted model parameters applicable to the current operating condition are dynamically selected, so that the fused carbon emission data structure better reflects the real carbon emission characteristics under different operating scenarios.

[0162] In summary, this invention establishes a dynamic correlation between the physical topology of the power grid and carbon emission data by integrating spatial coordinates and time-series data. Its core lies in accurately mapping multi-dimensional spatiotemporal attributes (such as geographical location, timestamps, and topological connectivity) with heterogeneous data (such as power grid operating status, carbon emission intensity, and energy flow paths) to reveal the evolutionary patterns of physical elements in complex systems over time and space. Through joint calibration of timestamps and spatial coordinates, the algorithm can synchronously correlate real-time power grid operating data (such as power load and inter-regional transmission power) with regional carbon emission dynamic factors (such as hourly carbon emission intensity), supporting refined spatiotemporal tracking and modeling of carbon flow in the power system. In the context of power grid and carbon emission big data, the algorithm can solve the scale mismatch problem between macro-regional (such as provincial administrative regions) carbon emission statistics and micro-node (such as substations and transmission lines) physical data, constructing a cross-scale linkage model of energy flow and carbon emissions through spatiotemporal interpolation, path overlay, and other technologies.

[0163] Example 3: The above is an illustrative scheme of a power grid carbon emission traceability method based on spatiotemporal matching.

[0164] This embodiment also provides an electronic device suitable for grid carbon emission traceability based on spatiotemporal matching, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the grid carbon emission traceability method based on spatiotemporal matching as proposed in the above embodiment.

[0165] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the spatiotemporal matching-based grid carbon emission traceability method proposed in the above embodiments.

[0166] The storage medium proposed in this embodiment and the method for implementing the power grid carbon emission traceability based on spatiotemporal matching 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.

[0167] Based on the above description of the implementation methods, those skilled in the art will clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0168] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not 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 tracing carbon emissions from power grids based on spatiotemporal matching, characterized in that, include: A model corresponding to geographic coordinates and power system network structure is constructed to obtain a spatial topological representation; Based on spatial topological representation, multidimensional heterogeneous data is introduced, and a data processing mechanism is established to generate datasets through standardized merging methods; Using the dataset, a response relationship model between time series and spatial units is established. Through grid planning, cross-scale matching operations are completed to form an association structure. Based on the association structure, data fusion is performed through scale transformation and weight superposition mechanisms, and the fused data structure is output.

2. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 1, characterized in that, The construction of the correspondence model between geographic coordinates and power system network structure yields a spatial topological expression, including: Determine the spatial information of nodes in the power grid structure and construct a spatial coordinate system; Identify the constituent units of the power grid and establish their mapping relationships in a spatial coordinate system; Based on the power grid connection relationships, a connection structure between nodes is constructed to form a spatial topology hierarchy; By jointly representing spatial information and structural relationships, a spatial topological expression with locational and structural characteristics is formed.

3. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 2, characterized in that, The method, based on spatial topology representation, introduces multidimensional heterogeneous data and establishes a data processing mechanism. Through standardized merging, a dataset is generated, including: Collect data related to multiple power grid sources and construct a unified data access format; Establish a data processing mechanism to handle differences in data types and inconsistencies in structure, and perform cleaning and formatting operations on the input data; Based on the data processing mechanism, a heterogeneous data alignment strategy is applied to the collected data to form a standardized dataset.

4. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 3, characterized in that, The process of establishing a response relationship model between time series data and spatial units using the dataset includes: The standardized dataset is divided into data partitions according to time and spatial location. Construct spatiotemporal association rules and define association rules between different data partitions; Based on association rules, a response relationship model between time series and spatial units is generated.

5. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 4, characterized in that, The process of using grid planning to perform cross-scale data matching operations and form an association structure includes: A spatiotemporal grid framework is established by dividing the standardized dataset into grids based on time and space dimensions. Matching and comparing data at different time scales, extracting features, and establishing correspondences between time series data are represented as follows: Where, x i and y j Let i and j represent the i-th and j-th points of two time series data, respectively. DTW(i,j) represents the matching degree between the corresponding time points, i-1 represents the matching degree between the (i-1)-th point and the j-th point, and DTW represents the function used for dynamic time warping. The minimum matching degree among the data points near the two time series data points is the matching degree of the two time series data points. Interpolation and scaling operations are performed on the spatial data to improve the spatial distribution information within the grid cells, as shown below: Among them, Z * (s0) is the estimated value of point s0, Z(s) i ) represents the value of a known data point, d(s0,s) i ) represents the estimated distance between the point and the known point, and p is the power exponent; By combining the matching results of time and space dimensions, a data connection structure for cross-scale mapping features is constructed as the association structure for fusion operations.

6. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 5, characterized in that, The heterogeneous data alignment strategy includes: Multidimensional power grid operation data from power generation, load, and dispatching are categorized by source and their spatiotemporal features are extracted. Multidimensional data is standardized and represented as follows: Among them, z i It is the standard deviation, x i These are the original data, μ is the mean of the data, and σ is the standard deviation; Based on time index and spatial location index, the processed data is matched to establish the correspondence between multiple types of heterogeneous data in the spatiotemporal grid. The aligned data is organized into a standardized dataset for response modeling and scale fusion operations.

7. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 6, characterized in that, The spatiotemporal association rules include: The standardized data is segmented according to time order and different time periods are allocated; comparison rules are used to describe the time change process and identify the temporal characteristics between adjacent time periods. Analyze data pairs from different time periods, determine comparison methods, and identify relationships between the data through these comparison methods.

8. The power grid carbon emission tracing method based on spatiotemporal matching as described in claim 7, characterized in that, The process of data fusion based on the association structure, through scaling and weighting mechanisms, and outputting the fused data structure includes: A scale transformation function is constructed by fusing data at different time and spatial scales. Based on the cross-matching results of spatial and temporal scales, a weighted fusion model is constructed, represented as: Among them, F 融合 F represents the fusion result. i For features at different scales, w i These are the corresponding weighting coefficients; Based on the grid mapping relationship in the association structure, multi-source data under the same grid cell are collaboratively weighted and integrated; the output fused data structure retains the key information of carbon emission intensity and matching path of each grid cell.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the power optimization method for multi-scale source-load matching according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the power optimization method for multi-scale source-load matching according to any one of claims 1 to 7.

Images