Carbon emission correlation modeling method and system in combination with load space distribution characteristics

Abstract

The invention discloses a load spatial distribution characteristic combined carbon emission correlation modeling method and system, and relates to the technical field of carbon emission management and optimal scheduling, and the method comprises the steps: collecting node load data and geographic information data in a power system, carrying out the preprocessing, and building a correlation table of node loads and geographic characteristics; modeling the relevance between the load space distribution and the carbon emission factors based on a geographic weighted regression model and a spatial autoregression model, and constructing a relevance model between the load space distribution and the carbon emission factors; training and calibrating the correlation model, and evaluating the fitting effect and prediction precision of the model by adopting a cross validation method; and predicting the spatial distribution characteristics of the carbon emission of each region by using the correlation model, and implementing low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions. According to the invention, through combination of the geographically weighted regression model and the spatial autoregression model, spatial refined modeling of carbon emission factors is realized, and a high-carbon emission density region can be accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission management and optimal scheduling, and in particular to a carbon emission correlation modeling method and system combining load spatial distribution characteristics. Background Art

[0002] With the continuous improvement of carbon emission reduction requirements globally, the demand for reducing greenhouse gas emissions in all industries is becoming increasingly urgent; in the power industry, due to the diversification of power generation types and the differences in power generation characteristics in different regions, the carbon emission distribution of the power system also exhibits significant spatial characteristics; traditional carbon emission management methods generally calculate the global carbon emissions through the overall averaged carbon emission factors, without fully considering the load characteristics and power generation type differences of different nodes within the region; existing technologies usually ignore the carbon emission modeling of spatial distribution characteristics and are difficult to accurately describe the differentiated carbon emissions between regions.

[0003] In the power system, load-intensive areas are usually accompanied by higher carbon emissions; therefore, incorporating the load spatial distribution characteristics into the carbon emission model and combining with geographic information system (GIS) and spatial analysis technologies can further optimize the carbon emission management strategy of the power system; however, in the low-carbon scheduling and carbon emission management of the power system, existing technologies generally ignore the spatial correlation and the mutual influence between regions, only focusing on the optimization of the time dimension and failing to achieve a refined regional management effect.

[0004] The Geographically Weighted Regression (GWR) model is a regression method that can handle spatial heterogeneity, enabling weighted processing of regression coefficients for different geographical locations and being suitable for capturing the spatial differences in carbon emissions in the power system; while the Spatial Autoregression (SAR) model can represent the spatial interaction between regions, enabling the model to consider the correlation effects between adjacent nodes, and thus has high applicability in multi-node and multi-region systems; combining these two models can more accurately evaluate the regional carbon emission distribution of the power system and support regional low-carbon optimal scheduling.

[0005] Therefore, it is necessary to propose a carbon emission correlation modeling method combining load spatial distribution characteristics to solve the problem in the prior art that the regional carbon emission differences cannot be captured in a refined manner, so as to achieve accurate carbon emission assessment and low-carbon scheduling optimization based on spatial distribution characteristics. Summary of the Invention

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the problem to be solved by the present invention is how to provide a carbon emission correlation modeling method that combines the load space distribution characteristics. By incorporating the geographical location characteristics of the load into the carbon emission factor model, the problem of ignoring the spatial distribution characteristics in the existing carbon emission modeling is solved, realizing regional low-carbon scheduling and management, so as to more accurately identify the high-density areas of carbon emissions and provide support for the low-carbon optimization of the power system.

[0008] To solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, an embodiment of the present invention provides a carbon emission correlation modeling method that combines the load space distribution characteristics, which includes collecting node load data and geographical information data in the power system and performing preprocessing to establish an association table between the node load and geographical features; modeling the correlation between the load space distribution and the carbon emission factor based on the geographically weighted regression model and the spatial autoregressive model to construct an association model between the load space distribution and the carbon emission factor; training and calibrating the association model, and using the cross-validation method to evaluate the fitting effect and prediction accuracy of the model; using the association model to predict the spatial distribution characteristics of carbon emissions in each region, and implementing low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions.

[0010] As a preferred solution of the carbon emission correlation modeling method that combines the load space distribution characteristics of the present invention, wherein: the collection of node load data and geographical information data in the power system includes: collecting the load data of each node through the smart grid or the load monitoring system to ensure that the load data contains geographical location information; obtaining geographical information data related to the location of the load node, including climate data, land use type, and population density; the preprocessing includes: performing spatial matching on the load data and the geographical information data to establish an association table between the load and geographical features of each node; performing normalization processing on the load data and the geographical information data.

[0011] As a preferred solution of the carbon emission correlation modeling method that combines the load space distribution characteristics of the present invention, wherein: the association model between the load space distribution and the carbon emission factor includes: a geographically weighted regression model for modeling the change of the carbon emission factor with geographical location to capture the carbon emission differences in different regions; a spatial autoregressive model for describing the spatial autocorrelation characteristics of the carbon emission factor, and reflecting the mutual influence of carbon emissions between adjacent regions through the spatial weight matrix.

[0012] As a preferred solution of the carbon emission correlation modeling method that combines the load space distribution characteristics of the present invention, wherein: the geographically weighted regression model is used to describe the relationship between the carbon emission factor and the node location, and capture the carbon emission differences in different regions through the location-related regression coefficients. The calculation formula of the carbon emission factor EF(x,y) is as follows:

[0013] EF(x, y) = α(x, y) + β(x, y)L(x, y) + γ(x, y)T(x, y) + δ(x, y)H(x, y) + ε

[0014] Among them, L(x, y) is the load value of node (x, y); T(x, y) and H(x, y) are temperature and humidity respectively; α(x, y), β(x, y), γ(x, y), δ(x, y) are location-dependent regression coefficients; ε is the error term; the spatial autoregressive model is used to describe the spatial autocorrelation of carbon emission factors, and the calculation formula is as follows:

[0015]

[0016] Among them, ρ is the spatial autoregressive coefficient; w i is the spatial weight; N(x, y) is the neighborhood set of the node.

[0017] As a preferred embodiment of the carbon emission correlation modeling method combining load spatial distribution characteristics of the present invention, wherein: training and calibrating the correlation model includes the following steps: dividing the dataset according to geographical regions; processing location-related regression coefficients based on the geographically weighted regression model to optimize carbon emission factors in different regions; processing the node neighborhood relationship based on the spatial autoregressive model, and capturing the mutual influence of neighboring nodes through the spatial weight matrix; using the cross-validation method to evaluate the prediction accuracy of the model, and measuring the model error through the mean square error, and minimizing the sum of squared residuals.

[0018] As a preferred embodiment of the carbon emission correlation modeling method combining load spatial distribution characteristics of the present invention, wherein: the application of the correlation model includes the spatial distribution estimation of carbon emission factors and regional low-carbon scheduling optimization, and the specific steps are as follows: using the trained geographically weighted regression model and spatial autoregressive model to generate the spatial distribution of carbon emissions in the entire power grid system, and calculating the carbon emission factors of each node; based on the carbon emission characteristics of different regions, preferentially scheduling low-carbon power generation units in high-carbon emission regions or reducing their loads; increasing the load in low-carbon emission regions to improve the utilization rate of clean energy; the goal of the regional low-carbon scheduling optimization is to minimize carbon emissions and operating costs, and the specific formula is as follows:

[0019]

[0020] Among them, λ and μ are weight coefficients; is the carbon emission of node (x, y); C(P(x, y)) is the power generation cost of node (x, y).

[0021] As a preferred embodiment of the carbon emission correlation modeling method combining load spatial distribution characteristics according to the present invention, wherein: the correlation model further includes a scenario simulation function for predicting the carbon emission distribution under future load growth or clean energy introduction; the scenario simulation function includes scenario assumption, load prediction, and carbon emission distribution calculation, and the specific steps are as follows: predicting future loads based on different scenarios; using a geographically weighted regression model and a spatial autoregressive model to calculate the future spatial distribution of carbon emissions based on the assumed scenarios.

[0022] In a second aspect, to further solve the safety problems existing in carbon emission management and optimal scheduling, an embodiment of the present invention provides a carbon emission correlation modeling system combining load spatial distribution characteristics, which includes: a data acquisition module for collecting node load data and geographical information data in the power system, performing spatial matching and normalization processing, and establishing an association table between node loads and geographical features; a model construction module for modeling the correlation between load spatial distribution and carbon emission factors based on a geographically weighted regression model and a spatial autoregressive model, and constructing an association model between load spatial distribution and carbon emission factors; a model optimization module for training and calibrating the association model, and using a cross-validation method to evaluate the fitting effect and prediction accuracy of the model; a data prediction module for using the scenario simulation function in the association model to predict the spatial distribution characteristics of carbon emissions in each region, and implementing low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions.

[0023] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and wherein: when the computer program is executed by the processor, it implements any step of the carbon emission correlation modeling method combining load spatial distribution characteristics as described in the first aspect of the present invention.

[0024] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and wherein: when the computer program is executed by the processor, it implements any step of the carbon emission correlation modeling method combining load spatial distribution characteristics as described in the first aspect of the present invention.

[0025] Advantages of the present invention: By combining a geographically weighted regression model and a spatial autoregressive model, the present invention realizes the spatial fine-grained modeling of carbon emission factors, can accurately identify high carbon emission density regions, and provides more accurate data support for the low-carbon optimal scheduling of the power system; the present invention can achieve regional low-carbon scheduling based on the load distribution characteristics and carbon emission characteristics of different regions, optimize the load allocation strategies of different regions, and thus effectively reduce the overall carbon emissions; the model of the present invention can flexibly adapt to different load and environmental conditions during data collection and processing, and supports multiple data sources, such as meteorological sensors and GIS data. Description of the Drawings

[0026] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them:

[0027] Figure 1 It is the overall flowchart of the carbon emission correlation modeling method combining load spatial distribution characteristics in Embodiment 1.

[0028] Figure 2 It is the schematic diagram of the matching process between load spatial distribution and geographic information in Embodiment 1.

[0029] Figure 3 It is the schematic diagram of the spatial distribution process of carbon emission factors based on the geographically weighted regression model and the spatial autoregressive model in Embodiment 1.

[0030] Figure 4 It is the specific flowchart for regional low-carbon dispatch optimization in Embodiment 1.

[0031] Figure 5 It is the flowchart of scenario simulation prediction in Embodiment 1.

[0032] Figure 6 It is the schematic diagram of the structure of the computer device in Embodiment 3. Specific Embodiments

[0033] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the drawings in the specification.

[0034] In the following description, many specific details are set forth to facilitate a thorough understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0035] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that are mutually exclusive with other embodiments.

[0036] Embodiment 1

[0037] Refer to Figures 1 to 5 , which is the first embodiment of the present invention. This embodiment provides a carbon emission correlation modeling method combining load spatial distribution characteristics.

[0038] The existing carbon emission management methods mainly have the following problems: generally, the global carbon emissions are calculated through the overall averaged carbon emission factors, without fully considering the load characteristics and power generation type differences of different nodes within the region; the existing technologies usually ignore the carbon emission modeling of spatial distribution characteristics and it is difficult to accurately describe the differentiated carbon emissions between regions.

[0039] This application provides a solution that can effectively solve the above-mentioned problems. Next, multiple embodiments will be combined to elaborate in detail on how to implement the carbon emission correlation modeling method that combines the load spatial distribution characteristics.

[0040] Figure 1 The overall flowchart of the carbon emission correlation modeling method that combines the load spatial distribution characteristics is shown, including:

[0041] S1: Collect the node load data and geographical information data in the power system and perform preprocessing to establish an association table between the node load and geographical features.

[0042] Preferably, collecting the node load data and geographical information data in the power system includes: collecting the load data of each node in the power system through the smart grid or load monitoring system, ensuring that the load data contains geographical location information such as longitude and latitude or regional numbers, which are usually collected in real time by the smart grid and record the changes in power load at different time periods; the collected data not only contains the load value but also the geographical location information of the node, thus being able to provide support for the load spatial distribution.

[0043] Obtain the geographical information data related to the location of the load node, including climate data, land use type, and population density, for enhancing the spatial analysis ability of the model. These geographical information data can be collected through the Geographic Information System (GIS) or various environmental sensors.

[0044] Furthermore, as Figure 2 shown in the schematic diagram of the load spatial distribution and geographical information matching process, the preprocessing includes: performing spatial matching on the load data and geographical information data to establish an association table between the load of each node and geographical features. Through the matching process, the association relationship between the load of each node and its geographical features can be clarified, providing detailed data support for the subsequent carbon emission factor modeling.

[0045] To ensure the scale consistency between the characteristics of different data sources, the present invention performs standardization processing on the load data and geographical information data; through normalization or other data conversion methods, the data of different scales are unified into the same scale range, eliminating the influence on model training caused by data scale differences and improving the stability and accuracy during the model training process.

[0046] S2: Based on the geographically weighted regression model and the spatial autoregressive model, model the correlation between the load spatial distribution and the carbon emission factor, and construct an association model between the load spatial distribution and the carbon emission factor.

[0047] It should be noted that after the data collection and preprocessing are completed, the present invention constructs an association model between the load spatial distribution and the carbon emission factor to more accurately reveal the carbon emission characteristics of different regions in the power system, including establishing a carbon emission factor model based on the geographically weighted regression model (GWR) and using the spatial autoregressive model (SAR) for regional feature analysis.

[0048] Preferably, the association model between the load spatial distribution and the carbon emission factor includes: the geographically weighted regression model (GWR), which is used to model the change of the carbon emission factor with the geographical location to capture the carbon emission differences in different regions.

[0049] The spatial autoregressive model (SAR), which is used to describe the spatial autocorrelation characteristics of the carbon emission factor and reflects the mutual influence of carbon emissions between adjacent regions through the spatial weight matrix.

[0050] Specifically, as Figure 3 shown in the schematic diagram of the spatial distribution process of the carbon emission factor based on the geographically weighted regression model and the spatial autoregressive model, the geographically weighted regression model is used to describe the relationship between the carbon emission factor and the node location, and captures the carbon emission differences in different regions through the location-dependent regression coefficients. The calculation formula of the carbon emission factor EF(x, y) is as follows:

[0051] EF(x, y) = α(x, y) + β(x, y)L(x, y) + γ(x, y)T(x, y) + δ(x, y)H(x, y) + ε

[0052] where L(x, y) is the load value of the node (x, y); T(x, y) and H(x, y) are the temperature and humidity respectively; α(x, y), β(x, y), γ(x, y), δ(x, y) are the location-dependent regression coefficients; and ε is the error term.

[0053] Preferably, the advantage of the geographically weighted regression model is that it allows the regression coefficients at different geographical locations to be different, so as to reflect the non-uniformity of the carbon emission factor in space. Through the weighted regression at each node location, the spatial accuracy of the carbon emission factor model can be improved, and regionalized carbon emission assessment can be provided.

[0054] Specifically, the spatial autoregressive model is used to describe the spatial autocorrelation of the carbon emission factor. The spatial autoregressive model captures the spatial dependence between adjacent regions by introducing a spatial weight matrix, and the calculation formula is as follows:

[0055]

[0056] Among them, ρ is the spatial autoregressive coefficient, reflecting the spatial dependence of neighboring nodes; w i is the spatial weight, representing the influence of the neighbor node tree node (x, y); N(x, y) is the neighborhood set of the node.

[0057] Preferably, the advantage of the spatial autoregressive model is that it can quantify the interaction between adjacent nodes, making the carbon emission factor modeling more accurate, and is particularly suitable for the carbon emission distribution modeling of large-scale, multi-node, and multi-region power systems; combining the geographically weighted regression model and the spatial autoregressive model, the present invention can fully consider the spatial correlation characteristics in the carbon emission model to achieve the spatial dynamic correlation modeling of load and carbon emission.

[0058] S3: Train and calibrate the correlation model, and use the cross-validation method to evaluate the fitting effect and prediction accuracy of the model.

[0059] It should be noted that after the model is constructed, the present invention trains and calibrates the model to ensure the accuracy and stability of the carbon emission factor model. The model training includes dataset division, parameter optimization of the geographically weighted regression model and the spatial autoregressive model, as well as cross-validation and error analysis.

[0060] Preferably, training and calibrating the correlation model includes the following steps: dividing the dataset according to geographical regions to ensure the diversity of load and geographical features in different regions.

[0061] Based on the geographically weighted regression model (GWR), process the location-related regression coefficients and optimize the carbon emission factors in different regions.

[0062] Based on the spatial autoregressive model (SAR), process the node neighborhood relationship and capture the mutual influence of neighboring nodes through the spatial weight matrix.

[0063] Use the cross-validation method to evaluate the prediction accuracy of the model, and measure the model error through the mean square error (MSE), minimizing the sum of squared residuals to ensure the prediction effect of the carbon emission factor.

[0064] Specifically, dataset division means that for the convenience of model training and validation, the present invention divides the collected load and geographical information data into a training set and a test set according to regions, ensuring that the characteristics of different regions can be learned by the model; before data training, standardize the input load and geographical variables to ensure consistent feature scales to avoid numerical deviations in model training.

[0065] Specifically, the parameter optimization of the geographically weighted regression model refers to that during the training process of the geographically weighted regression model, first calculate the location-dependent regression coefficients for each node, and fit the model parameters by minimizing the residual sum of squares (RSS). The specific formula of the objective function is as follows:

[0066]

[0067] Among them, EF(x i ,y i ) is the carbon emission factor of node (i); L(x i ,y i ) is the node load value; (x i ,y i ) is the geographical location coordinates of the node; by minimizing RSS, the optimal fitting of the regression coefficients can be achieved, so as to obtain the predicted values of carbon emission factors at different geographical locations.

[0068] Specifically, the parameter optimization of the spatial autoregressive model refers to that during the training of the spatial autoregressive model, the present invention optimizes the autoregressive coefficient ρ and the load coefficient β of the model according to the neighborhood set N(x, y) and the spatial weight matrix w i to maximize the characterization accuracy of the interaction between adjacent nodes by the model; the spatial autoregressive model can more accurately reflect the spatial correlation between nodes by adjusting the weights between nodes within the neighborhood set.

[0069] Specifically, cross-validation and error analysis refer to that in order to evaluate the performance of the model, the present invention uses the cross-validation method to verify the GWR and SAR models. The preferred cross-validation methods include leave-one-out cross-validation LOOCV and K-fold cross-validation. The prediction error of the carbon emission factor is measured by calculating the mean squared error (MSE). The specific formula is as follows:

[0070]

[0071] Among them, EF i is the actual carbon emission factor; is the model predicted value; through cross-validation and error analysis, the present invention can further optimize the parameter settings of the model and improve the prediction accuracy of the model.

[0072] Furthermore, the applications of the correlation model include the estimation of the spatial distribution of carbon emission factors and the optimization of regional low-carbon scheduling. The specific steps are as follows: Use the trained geographically weighted regression model and spatial autoregressive model to generate the spatial distribution of carbon emissions in the entire power grid system, and calculate the carbon emission factors of each node. With the help of the geographically weighted regression model, the present invention can achieve the distribution estimation of carbon emission factors in the power system according to the load and geographical characteristics of each node; while the spatial autoregressive model provides the spatial correlation information between adjacent nodes, enabling the estimation results to not only consider the individual characteristics of each node but also capture the impact of adjacent nodes on carbon emissions. This estimation method can effectively identify the high-density areas of carbon emissions in the power system, laying a foundation for subsequent low-carbon scheduling optimization.

[0073] As Figure 4 shown in the specific flowchart for regional low-carbon scheduling optimization, based on the spatial distribution of carbon emissions, and adopting differentiated scheduling strategies to optimize the carbon emissions of the region according to the carbon emission characteristics of different regions, including optimization objectives, scheduling scheme design, multi-objective optimization, and scheduling update and adjustment.

[0074] Specifically, the optimization objective is to control the carbon emissions in high-carbon emission areas and improve the utilization rate of clean energy by adjusting the load distribution of each region. In the present invention, low-carbon power generation units are preferentially scheduled in high-carbon emission areas to reduce the dependence on traditional high-carbon emission power generation, while in low-carbon emission areas, the load is moderately increased to make full use of clean energy resources. The specific formula is as follows:

[0075]

[0076] where λ and μ are weight coefficients used to balance carbon emissions and operating costs; is the carbon emission of node (x, y); C(P(x, y)) is the power generation cost of node (x, y).

[0077] Specifically, the scheduling scheme design means that for high-carbon emission areas, the present invention preferentially schedules clean energy (such as wind power, solar power, etc.) and reduces the output of traditional thermal power; for areas with relatively low carbon emissions, the scheduling can moderately increase the load utilization rate to effectively utilize clean energy and reduce the overall carbon emissions; the design of the scheduling scheme for each region is based on the carbon emission factor estimation results and is dynamically adjusted according to the load demand between regions.

[0078] Specifically, multi-objective optimization means that when implementing low-carbon scheduling optimization, the present invention adopts a multi-objective optimization algorithm to enable the optimization process to consider both carbon emissions and operating costs simultaneously. The specific algorithm can adopt genetic algorithms, particle swarm optimization algorithms, or other common optimization methods to solve the multi-objective problem of carbon emission and cost balance; through the iterative optimization process, the optimal scheduling scheme is found, thereby effectively reducing carbon emissions and saving operating costs.

[0079] Specifically, scheduling update and adjustment refer to applying the optimal scheduling plan to the power system after the optimization process is completed to adjust the scheduling plans for each region; in high-carbon emission regions, gradually reduce regional carbon emissions by reducing the proportion of traditional power generation and increasing the utilization rate of clean energy; in low-carbon emission regions, improve the overall efficiency of the system by moderately increasing the load; the present invention can also dynamically adjust the scheduling plan according to real-time monitoring data to cope with load fluctuations and power generation changes.

[0080] Preferably, through the above scheduling optimization, the present invention can significantly reduce the emission intensity in high-carbon emission regions, and at the same time maximize the utilization of clean energy in low-carbon emission regions; regional scheduling optimization effectively balances the load and carbon emission characteristics of different regions, providing a practical solution for the low-carbon development of the power system.

[0081] S4: Use the correlation model to predict the spatial distribution characteristics of carbon emissions in each region, and implement low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions.

[0082] Preferably, the correlation model also includes a scenario simulation function for predicting the carbon emission distribution under future load growth or clean energy introduction.

[0083] Specifically, as Figure 5 shown in the scenario simulation prediction flow chart, in order to further verify the applicability and robustness of the present invention, the scenario simulation function includes scenario assumption, load prediction, and carbon emission distribution calculation.

[0084] Furthermore, scenario assumption and load prediction refer to establishing multiple future scenarios to predict the carbon emission distribution under future load growth and increasing clean energy proportion. In the scenario assumption, the following several typical situations can be considered: Scenario 1 is the load growth scenario, assuming that with the development of the economy and population, the power demand continues to grow and the load of each node increases.

[0085] Scenario 2 is the clean energy introduction scenario, assuming that the proportion of clean energy power generation gradually increases and the dependence on traditional thermal power generation is reduced.

[0086] Scenario 3 is the load migration scenario, assuming that the load in some high-carbon emission regions is transferred to low-carbon emission regions to further reduce the pressure on high-emission density regions.

[0087] Specifically, in the load growth scenario, the present invention estimates the future load increment of each node through a trend prediction model based on historical load data, and inputs these increment data into the GWR and SAR models to calculate the impact of load growth on the spatial distribution of carbon emissions; such scenario simulation can identify potential high-carbon emission density regions in the future, providing a reference basis for power system planning.

[0088] Specifically, in the scenario of introducing clean energy, the present invention assumes an increase in the power generation proportion of wind power and photovoltaic clean energy, gradually replacing traditional fossil energy power generation; through the GWR model and the SAR model, it simulates the impact of the introduction of clean energy on the carbon emission distribution in different regions, predicts which regions will have a significant reduction in carbon emissions as the proportion of clean energy increases, and identifies low-carbon potential regions.

[0089] Specifically, in the load migration scenario, the present invention optimizes the geographical distribution of the load, reduces the load density in high-carbon emission regions, and transfers the load to regions with lower carbon emission density; through model simulation analysis, it analyzes how the spatial distribution of carbon emissions will change under the scenario after load migration, and the improvement effect of load distribution adjustment on regional carbon emissions.

[0090] Furthermore, during the scenario simulation process, the present invention introduces a dynamic adjustment mechanism to continuously adjust the model parameters according to the real-time monitored data and scenario analysis results; through gradual adjustment, it simulates the future carbon emission distribution and provides data support for carbon emission management and policy formulation under different scenarios.

[0091] In summary, through the combination of the geographically weighted regression model and the spatial autoregressive model, the present invention realizes the spatial refinement modeling of carbon emission factors, can accurately identify high-carbon emission density regions, and provides more accurate data support for the low-carbon optimal scheduling of the power system; the present invention can achieve regional low-carbon scheduling based on the load distribution characteristics and carbon emission characteristics of different regions, optimize the load allocation strategies of different regions, thereby effectively reducing the overall carbon emissions; the model of the present invention flexibly adapts to different load and environmental conditions during data collection and processing, and supports multiple data sources, such as meteorological sensors and GIS data.

[0092] Embodiment 2 is an embodiment of the present invention, which provides a carbon emission correlation modeling system combined with load spatial distribution characteristics, including: a data acquisition module, which is used to collect node load data and geographical information data in the power system, perform spatial matching and normalization processing, and establish an association table between node loads and geographical features; a model construction module, which is used to model the correlation between the load spatial distribution and carbon emission factors based on the geographically weighted regression model and the spatial autoregressive model, and construct an association model between the load spatial distribution and carbon emission factors; a model optimization module, which is used to train and calibrate the association model, and evaluate the fitting effect and prediction accuracy of the model by using the cross-validation method; a data prediction module, which is used to predict the spatial distribution characteristics of carbon emissions in each region by using the scenario simulation function in the association model, and implement low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions.

[0093] Embodiment 3 is an embodiment of the present invention, which is different from the previous embodiment in that:

[0094] As shown Figure 6 If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0095] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in conjunction with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0096] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or more wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0097] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0098] Embodiment 4, an embodiment of the present invention, provides a carbon emission correlation modeling method combined with load spatial distribution characteristics. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0099] During the implementation of the present invention, the calculations and scenario simulation analyses of the GWR and SAR models are realized through a computing platform; the present invention deploys the models to the power system monitoring center, and through integrating the geographic information system (GIS) and the load monitoring system, realizes the real-time update and analysis calculation of data; the power system monitoring center can, through the model calculation results, conduct real-time monitoring of the spatial distributions of load and carbon emissions.

[0100] In system integration, the data stream management of the power monitoring center is connected to various data sources of the smart grid through wired and wireless communication methods to ensure the real-time collection and update of load data and geographic information data; the load data is transmitted into the monitoring center through the smart grid interface, while the geographic information data is input through the GIS interface and the meteorological sensing system. The data stream management platform preprocesses the data, completes data standardization and matching operations, and transmits the processed data to the model calculation module.

[0101] The power system monitoring center uses the calculation results to assist in low-carbon scheduling decisions and provides optimization solutions for the load scheduling and power generation control of each region; for example, for high carbon emission regions, low-carbon power generation units can be preferentially scheduled or the load can be reduced, while for low carbon emission regions, the load demand can be appropriately increased to improve the utilization rate of clean energy; the model also supports real-time carbon emission monitoring, and through calculating and analyzing the carbon emission trends of different regions, provides data support for the regional carbon management of the power system.

[0102] In terms of carbon emission optimization and low-carbon power dispatching, the present invention provides an effective means for the low-carbon optimization of the power system through a carbon emission modeling method that combines the load spatial distribution characteristics. The model can accurately identify high carbon emission density regions based on the carbon emission factors at each node and perform differential low-carbon dispatching based on the regional load distribution characteristics. Through practical applications, the present invention can effectively reduce the load emission intensity in high carbon emission regions and improve the utilization rate of clean energy within the system.

[0103] In terms of the refined management of regional carbon emissions, the present invention realizes the spatial dynamic correlation modeling of carbon emission factors, enabling the carbon emission management of the power system to shift from the traditional overall average mode to refined management. In practical applications, the present invention provides a scientific basis for the carbon emission control and emission reduction policies of each region, enabling the carbon emission management to be flexibly adjusted according to the emission characteristics and load conditions of different regions.

[0104] In terms of supporting the formulation of future carbon emission policies, the scenario simulation and future carbon emission prediction functions provided by the present invention can provide a basis for the carbon emission policies of the future power system by predicting different load growth scenarios and clean energy introduction scenarios. In the context of load growth, scenario simulation can help decision-makers identify potential high carbon emission density regions in advance and take early control and optimization measures; in the case of clean energy introduction scenarios, it can guide the reasonable allocation of the proportion of clean energy between regions to improve the effectiveness of low-carbon dispatching.

[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A carbon emission correlation modeling method combining load spatial distribution characteristics, characterized by: include: Collect node load data and geographic information data in the power system and pre-process them to establish a correlation table between node load and geographic features; Based on the geographically weighted regression model and the spatial autoregression model, the correlation between the spatial distribution of load and the carbon emission factor is modeled, and a correlation model between the spatial distribution of load and the carbon emission factor is constructed; The association model is trained and calibrated, and the fitting effect and prediction accuracy of the model are evaluated by a cross-validation method; The association model is used to predict the spatial distribution characteristics of carbon emissions in each region, and low-carbon scheduling optimization is implemented based on the load and carbon emission characteristics of different regions.

2. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 1 is characterized in that: The acquisition of node load data and geographic information data in the power system includes: Collect load data of each node through smart grid or load monitoring system to ensure that the load data contains geographical location information; Obtain geographic information data related to load node locations, including climate data, land use type, and population density; The pre-processing comprises: Spatially match the load data with the geographic information data to establish a load and geographic feature association table for each node; Normalize the load data and geographic information data.

3. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 2 is characterized in that: The correlation model between load spatial distribution and carbon emission factor includes: Geographically weighted regression model, used to model the variation of carbon emission factors with geographical location, so as to capture the differences in carbon emissions in different regions; The spatial autoregressive model is used to describe the spatial autocorrelation characteristics of carbon emission factors and reflect the mutual influence of carbon emissions between neighboring regions through the spatial weight matrix.

4. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 3 is characterized by: The geographically weighted regression model is used to describe the relationship between the carbon emission factor and the node location, and captures the carbon emission differences in different regions through the location-related regression coefficients. The calculation formula of the carbon emission factor EF(x,y) is as follows: EF(x,y)=α(x,y)+β(x,y)L(x,y)+γ(x,y)T(x,y)+δ(x,y)H(x,y)+ε Where L(x,y) is the load value of node (x,y); T(x,y) and H(x,y) are temperature and humidity respectively; α(x,y), β(x,y), γ(x,y), δ(x,y) are the position-dependent regression coefficients; ε is the error term; The spatial autoregressive model is used to describe the spatial autocorrelation of carbon emission factors. The calculation formula is as follows: Among them, ρ is the spatial autoregressive coefficient; w i is the spatial weight; N(x,y) is the neighborhood set of the node.

5. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 4 is characterized in that: Training and calibrating the association model includes the following steps: Divide the dataset by geographic region; Based on the geographically weighted regression model, the location-related regression coefficients are processed to optimize the carbon emission factors in different regions; The node neighborhood relationship is processed based on the spatial autoregressive model, and the mutual influence of neighboring nodes is captured through the spatial weight matrix; The prediction accuracy of the model was evaluated using the cross-validation method, and the model error was measured by the mean square error to minimize the residual sum of squares.

6. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 5 is characterized by: The application of the association model includes the spatial distribution estimation of carbon emission factors and regional low-carbon scheduling optimization. The specific steps are as follows: Use the trained geographically weighted regression model and spatial autoregressive model to generate the spatial distribution of carbon emissions of the entire power grid system and calculate the carbon emission factor of each node; Based on the carbon emission characteristics of different regions, low-carbon power generation units in high-carbon emission areas are prioritized or their load is reduced; load is increased in low-carbon emission areas to improve the utilization rate of clean energy; The goal of the regional low-carbon scheduling optimization is to minimize carbon emissions and operating costs. The specific formula is as follows: Among them, λ and μ are weight coefficients; E CO2 (x,y) is the carbon emission of node (x,y); C(P(x,y)) is the power generation cost of node (x,y).

7. The carbon emission correlation modeling method combined with load spatial distribution characteristics according to claim 6 is characterized by: The association model also includes a scenario simulation function for predicting carbon emission distribution under future load growth or clean energy introduction; The scenario simulation function includes scenario assumptions, load forecasting and carbon emission distribution calculation. The specific steps are as follows: Forecast future loads based on different scenarios; The geographically weighted regression model and spatial autoregression model are used to calculate the spatial distribution of future carbon emissions based on hypothetical scenarios.

8. A carbon emission correlation modeling system combining load spatial distribution characteristics, based on the carbon emission correlation modeling method combining load spatial distribution characteristics according to any one of claims 1 to 7, characterized in that: include, The data acquisition module is used to collect node load data and geographic information data in the power system and perform spatial matching and normalization processing to establish an association table between node load and geographic features; The model building module is used to model the correlation between load spatial distribution and carbon emission factors based on the geographically weighted regression model and the spatial autoregression model, and to build a correlation model between load spatial distribution and carbon emission factors; Model optimization module, which is used to train and calibrate the correlation model and use cross-validation method to evaluate the model's fitting effect and prediction accuracy; The data prediction module is used to use the scenario simulation function in the association model to predict the spatial distribution characteristics of carbon emissions in each region, and implement low-carbon scheduling optimization based on the load and carbon emission characteristics of different regions.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the carbon emission correlation modeling method combined with load spatial distribution characteristics described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the carbon emission correlation modeling method combined with load spatial distribution characteristics described in any one of claims 1 to 7 are implemented.

Images