Park carbon emission visual analysis system and method

Abstract

The invention provides a park carbon emission visual analysis system and method, and relates to the technical field of smart energy management, spectral feature analysis is performed on a satellite remote sensing image of a park, and emission source intensity data of an emission abnormal region is identified and obtained; constructing a spatial topological structure and an initial carbon flow model for representing the energy conversion and transmission relationship in the park; performing reverse correction on the initial carbon flow model by taking the emission source intensity data as spatial constraint to obtain a carbon flow tracking matrix, and further performing flow field inversion on the energy measurement data and the carbon flow tracking matrix of the park to obtain a dynamic carbon flow field; performing carbon flow network traceability according to the circulation path of the dynamic carbon flow field in the spatial topological structure to obtain a carbon emission affiliation mapping map; and performing dynamic visual display on the carbon emission affiliation mapping map. According to the invention, the method can achieve the dynamic tracing and visualization of the carbon flow based on data collaborative correction, improves the temporal-spatial resolution of the carbon emission management of the park, and reduces the carbon accounting error of the park.

Description

Technical Field

[0001] This application relates to the field of smart energy management technology, and more specifically, to a visualization analysis system and method for carbon emissions in industrial parks. Background Technology

[0002] Smart energy management is a core means of promoting the green and low-carbon transformation of the industrial sector. It integrates advanced information, communication, and data analysis technologies to achieve refined perception, optimized scheduling, and efficient control over the entire energy production, transmission, and consumption chain. Against this backdrop, industrial parks, as key carriers of industrial agglomeration and energy consumption, have made accurate monitoring, source analysis, and visualization of their carbon emissions a crucial foundation for achieving "dual carbon" targets and implementing energy-saving and carbon-reduction transformations.

[0003] Currently, carbon emission visualization analysis at the industrial park level mainly relies on two technical approaches: one is bottom-up accounting and static chart display based on ground energy metering data, but this technology depends on statistical reports and fixed emission factors, making it difficult to depict the real-time dynamic spatial distribution of carbon emissions; the other is top-down monitoring based on remote sensing data, but this technology is independent of the energy system within the park and cannot link monitored emissions to specific production equipment, processes, or energy flow paths for source tracing. In summary, existing technologies cannot reflect the time-varying carbon intensity under the system's operating status and the electricity market environment; visualization presentation is mostly based on two-dimensional charts or simple heat maps, lacking dynamic and interactive three-dimensional spatiotemporal source tracing capabilities based on the physical transfer relationship of carbon flow. Therefore, how to achieve dynamic source tracing and visualization of carbon flow based on data collaborative correction, improve the spatiotemporal resolution of carbon emission governance in industrial parks, and thus reduce carbon accounting errors in industrial parks is a challenge faced by the industry. Summary of the Invention

[0004] This application provides a visualization analysis system and method for carbon emissions in industrial parks, which can realize dynamic source tracing and visualization of carbon flows based on data collaborative correction, improve the spatiotemporal resolution of carbon emission control in industrial parks, and thus reduce carbon accounting errors in industrial parks.

[0005] Firstly, this application provides a method for visual analysis of carbon emissions in industrial parks, the method comprising the following steps: Acquire energy metering data and satellite remote sensing imagery of the park; The spectral features of the satellite remote sensing images are analyzed to identify emission source intensity data in areas of abnormal emissions, and a spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationship within the park are constructed. The emission source strength data is used as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix. The energy metering data and the carbon flow tracking matrix are then used to perform flow field inversion to obtain a dynamic carbon flow field. Based on the flow path of the dynamic carbon flow field in the spatial topology, carbon flow network source tracing is performed to obtain a carbon emission attribution mapping map. The carbon emission attribution mapping map is visualized using a visualization rendering driver.

[0006] In this embodiment, acquiring the park's energy metering data and satellite remote sensing imagery specifically includes: Energy metering data for the park is obtained through metering instruments; Satellite remote sensing images of the park were obtained through a satellite data platform.

[0007] In this embodiment, the process of analyzing the spectral features of the satellite remote sensing image to identify the emission source intensity data of the emission anomaly area specifically includes: Radiometric calibration and atmospheric correction were performed on the satellite remote sensing images to obtain surface reflectance image data; Spectral feature matching is performed on the surface reflectance image data to obtain enhanced value data; Morphological filtering is performed on the enhanced value data to obtain a spatial distribution mask of the emission anomaly area; Wind speed field data is acquired, and the enhancement value data within the spatially distributed mask is integrally calculated based on the wind speed field data to obtain the emission source intensity data of the emission anomaly area.

[0008] In this embodiment, constructing the spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationships within the park specifically includes: A spatial topology structure with equipment as nodes and energy flow as edges is constructed by analyzing the spatial location and connectivity of energy equipment in the park. The node attribute dataset is determined based on the energy conversion type of the nodes in the spatial topology. Based on the node attribute dataset and in accordance with the carbon conservation constraint, a set of node carbon potential balance equations is constructed, thereby generating an initial carbon flow model.

[0009] In this embodiment, the emission source strength data is used as a spatial constraint to perform reverse correction on the initial carbon flow model, resulting in a carbon flow tracking matrix, specifically including: The emission source intensity data is mapped to a spatial topology that characterizes the energy conversion and transmission relationships within the park to obtain a set of node constraints. Run the initial carbon flow model and calculate the residual sequence between the theoretical carbon flow distribution output by the initial carbon flow model and the node constraint set; The initial carbon flow model is assimilated and updated based on the residual sequence using an ensemble Kalman filter algorithm to obtain a reconstructed set of nodal carbon potential balance equations. The carbon potential balance equations of the reconstructed node are solved to obtain the carbon flow tracking matrix.

[0010] In this embodiment, the flow field inversion of the energy metering data and the carbon flow tracing matrix to obtain the dynamic carbon flow field specifically includes: By performing a time-series correlation between the energy metering data and the carbon flow tracing matrix, a system of linear equations is obtained. Boundary condition constraints are added to the linear equation system based on carbon conservation constraints, thereby constructing a carbon flow inversion model; The carbon flow inversion model is solved using a linear programming algorithm to obtain the carbon flow allocation value of each node at each time step. The spatiotemporal kriging interpolation algorithm is used to perform spatial continuity and temporal smoothing on all carbon flow assignment values ​​to obtain a dynamic carbon flow field.

[0011] In this embodiment, carbon flow network tracing is performed based on the flow path of the dynamic carbon flow field in the spatial topology to obtain a carbon emission attribution mapping map, specifically including: A carbon flow intensity distribution matrix is ​​constructed based on the spatial topology and the dynamic carbon flow field. The graph theory-based flow network decomposition algorithm determines the carbon flow transfer ratio from each source node to the downstream node in the carbon flow intensity distribution matrix, thereby obtaining the carbon flow transfer ratio matrix. The carbon contribution dataset is determined based on the carbon flow transfer ratio matrix and the dynamic carbon flow field. Spatial overlay and aggregation analysis were performed on the carbon contribution dataset to obtain a carbon emission attribution mapping map.

[0012] In this embodiment, the visualization of the carbon emission attribution mapping map based on visualization rendering specifically includes: The carbon emission attribution mapping map is converted into a three-dimensional carbon flow network spatiotemporal evolution model based on an open-source 3D geographic information rendering library. The spatiotemporal evolution model of the three-dimensional carbon flow network is visualized using a visualization rendering driver.

[0013] In this embodiment, the visualization rendering driver refers to the visualization engine execution process of real-time spatiotemporal data parsing and interactive 3D rendering.

[0014] Secondly, this application provides a park carbon emission visualization analysis system for performing a park carbon emission visualization analysis method, the analysis system comprising: The multi-source data sensing module is used to acquire energy metering data and satellite remote sensing images of the park; The carbon flow mapping module is used to perform spectral feature analysis on the satellite remote sensing images, thereby identifying emission source intensity data in areas of abnormal emissions, and constructing a spatial topology and initial carbon flow model that characterize the energy conversion and transmission relationships within the park. The flow field inversion module is used to use the emission source strength data as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix, and to perform flow field inversion on the energy metering data and the carbon flow tracking matrix to obtain a dynamic carbon flow field. The attribution and tracing module is used to trace the carbon flow network based on the flow path of the dynamic carbon flow field in the spatial topology and obtain a carbon emission attribution mapping map. The visualization rendering module is used to visualize the carbon emission attribution mapping map based on the visualization rendering driver.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The process involves acquiring energy metering data and satellite remote sensing imagery of the industrial park; performing spectral feature analysis on the satellite remote sensing imagery to identify emission source intensity data for areas with abnormal emissions; and constructing a spatial topology and initial carbon flow model to characterize energy conversion and transmission relationships within the park. The emission source intensity data is then used as a spatial constraint to perform inverse correction on the initial carbon flow model, resulting in a carbon flow tracing matrix. Flow field inversion is then performed on the energy metering data and the carbon flow tracing matrix to obtain a dynamic carbon flow field. Based on the flow path of the dynamic carbon flow field within the spatial topology, carbon flow network tracing is performed to obtain a carbon emission attribution mapping map. Finally, the carbon emission attribution mapping map is visualized using a visualization rendering driver.

[0016] Therefore, this application can realize dynamic carbon flow tracing and visualization based on data-driven collaborative correction. First, by simultaneously acquiring energy metering data and satellite remote sensing imagery, a multi-source heterogeneous data foundation is laid for integrating the internal system operation status of the park with external spatial observation information, overcoming the limitations of a single data source perspective. Second, by performing spectral feature analysis on remote sensing imagery to identify and quantify emission source intensity data, and constructing a spatial topology and initial carbon flow model, a preliminary correlation can be established between macroscopic abnormal emission location and microscopic energy system physical structure, which is conducive to establishing a computable network framework for accurate source tracing. Furthermore, by using emission source intensity data as spatial constraints to perform inverse correction on the initial model, a carbon flow tracking matrix is ​​obtained, and a dynamic carbon flow field is generated based on energy data flow field inversion. This approach helps solve the problem that traditional static models struggle to describe the spatiotemporal dynamics of carbon emissions, thus facilitating high-resolution temporal reconstruction of carbon flow states. Then, based on the flow path of the dynamic carbon flow field within the spatial topology, network tracing is performed to generate a carbon emission attribution mapping map. This accurately decomposes abstract emissions and traces them to specific equipment, processes, or responsible units, achieving a crucial leap from monitoring emissions to identifying responsibility. Finally, based on visualization rendering, the attribution mapping map is visualized, transforming complex calculation results into an intuitive and interactive spatiotemporal decision-making interface. This allows managers to gain real-time insights into the spatial distribution of carbon emissions, trace transfer paths, and assess control effects, thereby supporting the park's comprehensive improvement in carbon governance, enhancing its refined management and proactive intervention capabilities.

[0017] In summary, the technical solution adopted in this application can realize dynamic carbon flow tracing and visualization based on data collaborative correction, improve the spatiotemporal resolution of carbon emission control in the park, and thus reduce carbon accounting errors in the park. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for visualizing and analyzing carbon emissions in a park, as provided in this application. Figure 2 This is an exemplary flowchart based on the emission source intensity data for determining emission anomaly areas provided in this application; Figure 3 This is an exemplary flowchart for determining a carbon flow tracing matrix provided in this application; Figure 4 This is a module structure diagram of a park carbon emission visualization analysis system provided in this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a system and method for visualizing and analyzing carbon emissions in a park. The core of this system involves acquiring energy metering data and satellite remote sensing imagery of the park; performing spectral feature analysis on the satellite remote sensing imagery to identify emission source strength data for areas with abnormal emissions; and constructing a spatial topology and an initial carbon flow model characterizing energy conversion and transmission relationships within the park. The emission source strength data is then used as a spatial constraint to perform inverse correction on the initial carbon flow model, resulting in a carbon flow tracing matrix. Flow field inversion is then performed on the energy metering data and the carbon flow tracing matrix to obtain a dynamic carbon flow field. Based on the flow path of the dynamic carbon flow field within the spatial topology, carbon flow network tracing is performed to obtain a carbon emission attribution mapping map. Finally, the carbon emission attribution mapping map is visualized using a visualization rendering driver.

[0022] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is a flowchart of a method for visualizing and analyzing carbon emissions in a park according to this embodiment of the present application. The analysis method includes the following steps: In step S1, energy metering data and satellite remote sensing images of the park are acquired.

[0023] In this embodiment, acquiring the park's energy metering data and satellite remote sensing imagery specifically includes: Energy metering data for the park is obtained through metering instruments; Satellite remote sensing images of the park were obtained through a satellite data platform.

[0024] In practical implementation, firstly, smart meters for measuring the overall electricity consumption of the park can be installed at the main power supply line inlet; gas flow meters for measuring natural gas consumption can be installed at key nodes of the gas supply pipeline; and heat meters for measuring heat energy transmission can be installed on key pipelines of the centralized heating system. These instruments are connected to a higher-level data aggregation device via an on-site industrial communication network. The data aggregation device automatically reads the real-time readings of all instruments at a set, fixed cycle, packages the collected energy consumption values, and sends them to the park's data center server for centralized storage via wired or wireless network. The data is then stored on the server. The system stores a complete record of the consumption of various energy sources, including electricity, natural gas, and heat, at different points in time, as energy metering data. Then, an automated data acquisition program is developed to access public or commercial data service platforms that can provide satellite imagery data. The program sets the geographical area (i.e., the geographical boundary of the park) to be imaged, the required image type (e.g., requiring the image to contain spectral bands that can detect specific gases), and submits an acquisition request. After receiving the request, the data service platform retrieves the satellite imagery data files that meet the requirements by searching the database and sends them back via the network, thus obtaining the satellite remote sensing imagery.

[0025] It should be noted that the metering instruments in this application refer to a general term for sensor devices installed on the energy infrastructure of the park, used to directly measure the physical consumption of various energy sources (e.g., electricity, gas, heat); the satellite data platform refers to a network system or interface that provides query and download services for Earth observation satellite image data. Furthermore, energy metering data is structured time-series data representing the actual consumption of various energy sources such as electricity, natural gas, and steam / hot water, used to characterize the real-time operating status and energy flow scale of the energy system within the park; satellite remote sensing imagery refers to digital image files covering the geographical area of ​​the target park, capable of capturing the characteristic absorption signals of solar radiation by greenhouse gas molecules (such as methane) in the atmosphere.

[0026] In step S2, the spectral features of the satellite remote sensing image are analyzed to identify the emission source intensity data of the emission anomaly area, and a spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationship within the park are constructed.

[0027] Preferably, in this embodiment, reference Figure 2 As shown, this figure is an exemplary flowchart for determining emission source intensity data of emission anomaly areas based on the information provided in this application. In this embodiment, the emission source intensity data of emission anomaly areas can be identified by performing spectral feature analysis on the satellite remote sensing image using the following steps: First, in step S21, the satellite remote sensing image is subjected to radiometric calibration and atmospheric correction to obtain surface reflectance image data; Secondly, in step S22, spectral feature matching is performed on the surface reflectance image data to obtain enhanced value data; Then, in step S23, morphological filtering is performed on the enhanced value data to obtain a spatial distribution mask of the emission anomaly area; Finally, in step S24, wind speed field data is acquired, and the enhancement value data within the spatial distribution mask is integrally calculated based on the wind speed field data to obtain the emission source strength data of the emission anomaly area.

[0028] In practice, firstly, the acquired satellite remote sensing image is opened using remote sensing data processing software. The radiometric calibration function is then selected within the software. This function automatically converts the original brightness value of each pixel into a physical quantity that represents the true reflectivity of the Earth's surface based on the sensor calibration parameters recorded in the image file, thus completing the radiometric calibration. Next, the atmospheric correction tool in the remote sensing data processing software is used. This tool employs a built-in atmospheric model to automatically estimate and subtract the effects of atmospheric scattering and absorption in the image, outputting an image that more closely approximates the true reflectivity of the Earth's surface. This processed image is used as the surface reflectivity image data. Secondly, a specific band sensitive to methane gas is selected from the surface reflectivity image data. Specifically, the spectral analysis function in the remote sensing data processing software is used to calculate the spectral curve of each pixel in this band. The matching degree is then calculated between this spectral curve and the standard methane absorption spectral curve stored in the software's spectral library, resulting in a numerical matrix reflecting the degree of methane concentration anomaly in each pixel. This numerical matrix is ​​then used as the augmentation value data. Then, image morphological processing is performed on the enhanced value data: first, an "erosion" operation is performed to remove isolated, small-area high-value noise points in the image; then, a "dilation" operation is performed to connect closely spaced high-value areas into patches, identifying abnormal regions. An automatic threshold segmentation algorithm is then used to calculate the optimal segmentation threshold based on the numerical distribution of the entire image, marking pixels above this threshold as foreground, resulting in a black-and-white binary image. This binary image is used as a spatial distribution mask for the emission anomaly region. Finally, wind speed field data of the park's location at the time the image was captured is obtained from the meteorological data service interface. This wind speed field data includes wind direction and wind speed. By establishing a diffusion calculation model based on mass conservation, using the area defined by the spatial distribution mask as the range, the enhanced value data within this area as the concentration field, and the wind speed field data as the driving condition, spatial integration is performed on the concentration enhancement values ​​along the wind direction to calculate the total emission rate corresponding to the abnormal region. The calculated list containing geographical location and emission rate information is then used as the emission source strength data for the emission anomaly region.

[0029] It should be noted that the enhanced value data in this application is a two-dimensional numerical matrix, where the value at each position represents the anomalous intensity of the methane gas spectral signal relative to the average level of the image background, used as a relative indicator to identify potential emission points; the spatial distribution mask is a binary (0 or 1) image, where pixel regions with a value of 1 identify the coherent suspected emission spatial range finally determined after image denoising and region enhancement processing, used to accurately define the target area in subsequent calculations, and pixel regions with a value of 0 indicate weak emission intensity in that area; the comprehensive quality enhancement model is a physical model based on the Gaussian diffusion principle or similar theory, which can use the observed two-dimensional concentration anomaly distribution and wind speed information to quantify the mass of gas released into the atmosphere by the emission source per unit time through inverse calculation, thereby transforming the anomalous signal on the image into a physically meaningful emission rate; the emission source intensity data is a structured dataset, where each data point corresponds to an identified ground emission source, including geographic coordinates and estimated emission intensity, providing direct input for spatial correlation with energy system data and model correction.

[0030] In this embodiment, constructing the spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationships within the park specifically includes: A spatial topology structure with equipment as nodes and energy flow as edges is constructed by analyzing the spatial location and connectivity of energy equipment in the park. The node attribute dataset is determined based on the energy conversion type of the nodes in the spatial topology. Based on the node attribute dataset and in accordance with the carbon conservation constraint, a set of node carbon potential balance equations is constructed, thereby generating an initial carbon flow model.

[0031] In practice, the process begins by collecting the park's energy system architecture diagram, equipment layout diagram, and related technical data to identify the physical locations of all key energy equipment (such as gas turbines, electric boilers, photovoltaic arrays, energy storage batteries, and main workshops). Then, graph theory modeling is used to abstract each independent energy device or key load as a point, and the energy transmission media connecting these devices, such as cables, pipes, and steam networks, are abstracted as lines connecting these points. Additionally, a two-dimensional table (such as an adjacency matrix) or list (such as an edge list) is used in the computer to record the connection relationships between all the "points" and the "lines," thus constructing a network framework describing how energy flows physically. This network framework serves as the spatial topology. Next, each "point" in the spatial topology is labeled with its energy conversion type, such as "gas-fired power generation," "photovoltaic power generation," "electric heating," and "purely electricity-consuming." Based on each type, the corresponding standard energy conversion efficiency parameters and carbon emission factors are retrieved from the equipment database or technical manuals, such as the efficiency of gas-fired power generation and the carbon emissions per kilowatt-hour generated. By associating these parameters with each node in the form of a data table, an attribute table describing the energy conversion and emission characteristics of each node can be obtained. This attribute table serves as the node attribute dataset. Finally, based on the parameters provided by the node attribute dataset and in accordance with the carbon conservation constraint (i.e., the total carbon flow into a node must be equal to the total carbon flow out of the node), mathematical equations are established for each non-source node (such as energy-consuming load). In the equations, the "carbon potential" of each node (i.e., the amount of carbon emissions carried by each unit of energy flowing through the node) is set as an unknown, and the measured or designed energy flow is used as a known coefficient. The equations of all nodes are then combined to form a complete linear equation system, which is then used as the initial carbon flow model.

[0032] It should be noted that the spatial topology in this application is a network graph model composed of nodes and edges, used to depict the physical transmission path of energy from source to end, providing a "skeleton" or "map" for subsequent networked tracking and calculation of carbon flows; the node attribute dataset refers to a set of technical parameters used to define the behavioral characteristics and environmental impact characteristics of each node in the energy conversion process; the initial carbon flow model is a mathematical model built based on the physical law of carbon conservation, which can describe how carbon emissions flow and distribute with energy in the pre-set network skeleton in the form of mathematical equations.

[0033] In step S3, the emission source strength data is used as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix. The energy metering data and the carbon flow tracking matrix are then used to perform flow field inversion to obtain a dynamic carbon flow field.

[0034] Preferably, in this embodiment, reference Figure 3As shown, this figure is an exemplary flowchart for determining the carbon flow tracing matrix according to the present application. In this embodiment, the emission source strength data is used as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain the carbon flow tracing matrix. This can be achieved by the following steps: First, in step S31, the emission source intensity data is mapped to a spatial topology that characterizes the energy conversion and transmission relationships within the park, resulting in a set of node constraints. Secondly, in step S32, the initial carbon flow model is run to calculate the residual sequence between the theoretical carbon flow distribution output by the initial carbon flow model and the node constraint set; Then, in step S33, the initial carbon flow model is assimilated and updated based on the residual sequence using the ensemble Kalman filter algorithm to obtain the reconstructed nodal carbon potential balance equation set. Finally, in step S34, the carbon potential balance equations of the reconstructed node are solved to obtain the carbon flow tracking matrix.

[0035] In specific implementation, firstly, in GIS software, the latitude and longitude coordinates of each emission point in the emission source strength data are read, along with the coordinates of all nodes in the spatial topology. By calculating distances, one or more energy equipment nodes in the spatial topology that are geographically closest to each emission point are found, and the estimated emission rate value of that emission point is assigned to these nodes, forming a series of "node-observation" pairs. This series of pairings is used as a node constraint set. Preferably, in this embodiment, for each node constraint pair in the node constraint set, the observed value (emission rate retrieved from remote sensing) is compared with a baseline emission range obtained based on historical data of that node or statistics from similar equipment. If the observed value exceeds a preset percentage of the baseline range (e.g., exceeding the upper limit by 50% and falling below the lower limit by 30%), the observed data is deemed invalid. For anomalous data, an anomalous alarm containing the node's location, observation value, and baseline information can be recorded and output. This anomalous data can be paused or marked for participation in subsequent model calibration processes, triggering a manual review task on the workflow platform and prompting relevant management personnel to conduct on-site verification or data source confirmation. Secondly, the acquired energy metering data for the current time period can be used as boundary conditions, input into the initial carbon flow model for solution calculation, obtaining the carbon emissions predicted by the initial carbon flow model for each node. The carbon emissions predicted by the initial carbon flow model on the node constraint set are then subtracted one by one from the corresponding observation values ​​in the node constraint set to obtain a set of differences. These differences are then arranged and combined in order to obtain a residual sequence. Finally, the ensemble Kalman filter algorithm is used to optimize the initial carbon flow model. The pseudocode for the optimization process of the ensemble Kalman filter algorithm on the initial carbon flow model is as follows: enter: 1. The initial parameter set X_f^a = {x_1, x_2, ..., x_N}, where each x_k represents a set of possible values ​​for the initial carbon flow model parameters (such as the node carbon emission factor), and N is the number of members in the set.

[0036] 2. Observed value y (i.e., emission rate value in the node constraint set).

[0037] 3. The observation operator H maps the model state (predicted node carbon emissions) to the observation space.

[0038] 4. Observation error covariance matrix R.

[0039] Output: The analyzed parameter set X_a is the updated set of model parameters that better reflects the observed data.

[0040] step: 1. Prediction step: For each set member k = 1 to N: a. Run the initial carbon flow model, using the current parameter x_k and the current energy metering data, to obtain the model's state forecast (the predicted distribution of nodal carbon emissions).

[0041] b. Apply the observation operator H to obtain the predicted value H(x_k) of the member in the observation space (the node corresponding to the node constraint set).

[0042] 2. Calculate set statistics: a. Calculate the mean of the forecast ensemble: \bar{x}_f = (1 / N) * Σ x_k b. Calculate the forecast ensemble perturbation matrix: X_f = [x_1 - \bar{x}_f, x_2 - \bar{x}_f, ...,x_N - \bar{x}_f] c. Calculate the mean of the forecast observation set: y_f = (1 / N) * Σ H(x_k) d. Calculate the forecast observation perturbation matrix: Y_f = [H(x_1) - y_f, H(x_2) - y_f, ..., H(x_N) - y_f] 3. Analysis Step (Update): Calculate the Kalman gain matrix K: K = (1 / (N-1)) * X_f * Y_f^T * [ (1 / (N-1)) * Y_f * Y_f^T + R ]^{-1} For each set member k = 1 to N: a. Calculate the observation perturbation of the member: d_k = y - H(x_k) + e_k, where e_k is a random perturbation drawn from the distribution N(0,R) to characterize the observation uncertainty.

[0043] b. Update the parameter of this member: x_k^a = x_k + K * d_k The analysis set is obtained by combining all the updated parameters x_k^a into a new set X_a.

[0044] This application utilizes an ensemble Kalman filter algorithm to represent adjustable parameters (such as carbon emission factors at certain nodes) in the initial carbon flow model as a set of multiple possible values. Then, it leverages the observation-prediction bias reflected in the residual sequence and automatically calculates and updates the weight probability of each value in this parameter set using the mathematical rules of the ensemble Kalman filter algorithm. This allows the adjusted model parameters to better match the observed data. The updated optimal parameter set is then used to reconstruct the node carbon potential balance equations, which are then used as the reconstructed node carbon potential balance equation set. Finally, the reconstructed node carbon potential balance equation set is mathematically solved. For example, based on the edge connections in the spatial topology and known energy flow data, the carbon flow from any node i to its connected node j is calculated, and all carbon flows are represented by a matrix. The element in the i-th row and j-th column of the matrix represents the carbon flow transfer intensity from node i to j, and this matrix is ​​then used as the carbon flow tracking matrix.

[0045] It should be noted that the node constraint set in this application is a set of data pairs formed by geographically aligning external emission observation data with the internal network nodes of the energy system. This set provides an independent validation benchmark from outside the system for model calibration. The residual sequence is a series of differences between the model predictions and actual observations, serving as a feedback signal that drives the automatic adjustment of model parameters. The ensemble Kalman filter algorithm is an advanced data assimilation technique. Its core idea is to maintain a set of probability distributions for model parameters and continuously update this distribution using observation data, thereby allowing the mathematical model to gradually approximate the state of the real system. It is a key algorithm for realizing the "reverse correction" mechanism model using remote sensing observation data. The reconstructed node carbon potential balance equation set is a mathematical model with parameters that better match the current actual observations, obtained after data assimilation and calibration. The elements in the carbon flow tracing matrix are used to characterize the transmission relationship and intensity of carbon emissions between any two connected nodes in the park's energy network.

[0046] In this embodiment, the flow field inversion of the energy metering data and the carbon flow tracing matrix to obtain the dynamic carbon flow field specifically includes: By performing a time-series correlation between the energy metering data and the carbon flow tracing matrix, a system of linear equations is obtained. Boundary condition constraints are added to the linear equation system based on carbon conservation constraints, thereby constructing a carbon flow inversion model; The carbon flow inversion model is solved using a linear programming algorithm to obtain the carbon flow allocation value of each node at each time step. The spatiotemporal kriging interpolation algorithm is used to perform spatial continuity and temporal smoothing on all carbon flow assignment values ​​to obtain a dynamic carbon flow field.

[0047] In practical implementation, firstly, energy metering data (such as total hourly electricity consumption and gas consumption) is correlated with the carbon flow tracing matrix. Specifically, for each time point, the overall energy input and output data of the park (from energy metering data) is used as the known boundary flow, and the unit carbon flow transfer relationship between nodes represented by the elements in the carbon flow tracing matrix is ​​used as fixed coefficients within the system. This establishes a system of linear equations with all unknown carbon flows in the network as variables. Secondly, the energy data reflected in the energy metering data and measured at the park's energy inlet and outlet are converted into carbon emissions, serving as the known input and output conditions that the entire system of equations must satisfy, i.e., boundary constraints. This completes the system by incorporating carbon conservation constraints at all nodes and boundary constraints. An integer system is used as the carbon flow inversion model. Then, a linear programming algorithm is employed to solve the carbon flow inversion model. This algorithm aims to minimize the total "virtual loss" of carbon emission transfer within the network or satisfy other physical rationality requirements. Under all constraints, it calculates a set of optimal solutions. Specifically, for each time point, a specific value is obtained, representing the precise amount of carbon emissions flowing from one node to another adjacent node at that moment. These flow values ​​between all nodes are summarized to obtain the carbon flow allocation value for each node at each time point. Finally, a spatiotemporal kriging interpolation algorithm is used to process the carbon flow allocation value. The pseudocode for this process is as follows: Input: A dataset of carbon flow assignment values, where each data point contains (longitude, latitude, time, carbon emission intensity). Output: Dynamic carbon flow field (continuous grid data covering the entire geographical area and time period of the park) step: Construct a spatiotemporal sample point set S = {s_i = (x_i, y_i, t_i, z_i)}, where (x_i, y_i) are the geographic coordinates of the node, t_i is the timestamp, and z_i is the carbon emission intensity of the node.

[0048] Establish a spatiotemporal variogram model: a. Calculate the spatiotemporal distance h_s (spatial Euclidean distance) and h_t (time difference) between all sample point pairs.

[0049] b. Fit a combined spatiotemporal variogram γ(h_s, h_t), for example, using a separable model: γ(h_s,h_t) = γ_s(h_s) + γ_t(h_t), where γ_s and γ_t are the variograms of pure space and pure time, respectively.

[0050] For the target interpolation point p_0 = (x_0, y_0, t_0) (i.e., any location and time within the park to be determined): a. Calculate the spatiotemporal covariance C(p_0,s_i) between the target point p_0 and all sample points s_i based on the spatiotemporal variability function.

[0051] b. Constructing the Kriging equations: [C(s_i, s_j)] [λ_j]= [C(p_0, s_i)] where [C(s_i, s_j)] is the covariance matrix between sample points, [λ_j] is the kriging weight to be determined, and [C(p_0, s_i)] is the covariance vector between the target point and the sample points.

[0052] c. Solve the above system of equations to obtain the weight λ_j.

[0053] d. Calculate the carbon emission intensity estimate for target point p_0: z_0 = Σ (λ_j * z_j).

[0054] By traversing the entire park's geographical grid and all continuous time points, step 3 is repeated to generate a continuous carbon emission intensity field covering the entire area and all times, i.e., a dynamic carbon flow field.

[0055] This application uses a spatiotemporal kriging interpolation algorithm to treat the carbon flow allocation value of each node at each time step (which can be converted into the carbon emission intensity of that node) as a sample point discretely distributed in three-dimensional space (two-dimensional geographic coordinates plus one-dimensional time). Based on the spatial location and time label of all sample points, as well as the statistical correlation between nodes, the spatiotemporal kriging interpolation algorithm constructs an interpolation model that can calculate the carbon emission intensity of any point within the geographical scope of the park at any consecutive time step. Then, the calculated carbon emission intensity field data, which covers the entire park and changes continuously with time, is used as a dynamic carbon flow field.

[0056] It should be noted that the linear equations in this application are a set of mathematical equations composed of the transitivity defined by the carbon flow tracking matrix and the boundary conditions defined by the energy metering data; the carbon flow inversion model is a mathematical model formed by describing the physical constraints (carbon conservation) that the carbon flow of each node must balance and the fixed input and output conditions determined by measured data; the linear programming algorithm is a mathematical optimization method that finds the optimal solution of the linear objective function under given linear constraints, and can be used to find the one that best conforms to physical laws (such as minimizing dissipation) among many possible carbon flow allocation schemes; the carbon flow allocation value is a numerical value representing the specific distribution of carbon emissions on each connecting path of the park's energy network at a specific time; the spatiotemporal kriging interpolation algorithm is a geostatistical algorithm that can simultaneously consider the correlation of data points in the spatial and temporal dimensions, and is used to reconstruct a continuous and smooth spatiotemporal field from discrete and sparse observation point data; the dynamic carbon flow field defines the carbon emission intensity of each "location-time" point on continuous geographic spatial coordinates and continuous time axes, which is conducive to achieving high-resolution and continuous characterization of the spatiotemporal dynamic evolution of carbon emissions in the park.

[0057] In step S4, carbon flow network tracing is performed based on the flow path of the dynamic carbon flow field in the spatial topology to obtain a carbon emission attribution mapping map.

[0058] In this embodiment, carbon flow network tracing is performed based on the flow path of the dynamic carbon flow field in the spatial topology to obtain a carbon emission attribution mapping map, specifically including: A carbon flow intensity distribution matrix is ​​constructed based on the spatial topology and the dynamic carbon flow field. The graph theory-based flow network decomposition algorithm determines the carbon flow transfer ratio from each source node to the downstream node in the carbon flow intensity distribution matrix, thereby obtaining the carbon flow transfer ratio matrix. The carbon contribution dataset is determined based on the carbon flow transfer ratio matrix and the dynamic carbon flow field. Spatial overlay and aggregation analysis were performed on the carbon contribution dataset to obtain a carbon emission attribution mapping map.

[0059] In practical implementation, firstly, a blank matrix with the number of rows and columns equal to the total number of nodes is created based on the spatial topology. The actual carbon flow at a specific moment is extracted from the dynamic carbon flow field, and the flow value is filled into the corresponding position in the matrix according to the node number, generating a matrix describing the carbon flow network distribution at that moment. This matrix then serves as the carbon flow intensity distribution matrix. Secondly, a graph-based flow network decomposition algorithm (such as downstream tracing) is used to process this matrix. Specifically, starting from each carbon emission source node, the total carbon emissions of the source node are distributed to all downstream nodes along the network edges according to the flow ratio of the current edge, resulting in a proportional matrix. This matrix can then be used to represent the carbon flow intensity distribution matrix. The proportion matrix serves as the carbon flow transfer proportion matrix. Then, the total carbon emissions of each source node in the dynamic carbon flow field at different times are multiplied by the carbon flow transfer proportion matrix to obtain the contribution of each node. The set of all contributions is then used as the contribution dataset. Finally, the carbon contribution dataset is overlaid with the park's geographic layer in GIS software. That is, nodes are mapped to their actual locations and associated with their respective management units. The contributions received by all nodes within the same unit are summed to obtain the total carbon emissions and source composition of the unit's consumption end. This data is then presented in a combination of maps and charts to obtain the carbon emission attribution mapping map.

[0060] It should be noted that the carbon flow intensity distribution matrix is ​​an instantaneous snapshot depicting the intensity of carbon flow distribution on the network edges at a specific moment; the graph theory-based flow network decomposition algorithm is used to scientifically decompose upstream carbon flow to downstream based on flow ratio; the carbon flow transfer ratio matrix is ​​a steady-state ratio relationship output by the graph theory-based flow network decomposition algorithm, used to describe the inherent carbon flow distribution law; the carbon contribution dataset records the contribution of each node to other nodes at each moment; spatial overlay and aggregation analysis are geographic information operations that match and summarize point data with surface layers; the carbon emission attribution mapping map is used to represent the spatial attribution relationship of carbon emissions from the production end to the consumption end, providing a basis for precise carbon management.

[0061] In step S5, the carbon emission attribution mapping map is visualized based on the visualization rendering driver.

[0062] It should be noted that, in this embodiment, the visualization rendering driver refers to the visualization engine execution process of real-time spatiotemporal data parsing and interactive 3D rendering. This driver process specifically includes three key technical steps: real-time spatiotemporal data parsing, interactive 3D rendering, and visualization engine execution. Real-time spatiotemporal data parsing means that the engine can read and understand carbon flow data with timestamps and geographic coordinates, and dynamically update the data status as the timeline progresses, providing a "flowing" data source for visualization. Interactive 3D rendering means that the engine uses computer graphics technology to transform the parsed data into graphics with visual elements such as 3D stereoscopic effects, color mapping, and dynamic particle flow, and responds to user operation commands such as zooming, panning, rotating, and filtering to achieve dynamic control of perspective and content.

[0063] In this embodiment, the visualization of the carbon emission attribution mapping map based on visualization rendering specifically includes: The carbon emission attribution mapping map is converted into a three-dimensional carbon flow network spatiotemporal evolution model based on an open-source 3D geographic information rendering library. The spatiotemporal evolution model of the three-dimensional carbon flow network is visualized using a visualization rendering driver.

[0064] In practice, the process begins by extracting geographic information, source-sink relationships, and temporal data from the carbon emission attribution mapping map. This involves using a 3D geographic information rendering library to write scripts that create a 3D scene base, load a simplified 3D base map of the park's buildings, and then place 3D icons representing emission sources and sinks in the scene based on geographic coordinates. Based on the time series and source-sink relationships, 3D lines or particle flows with flowing animation effects connecting the sources and sinks are dynamically generated to simulate carbon flow paths. This completed interactive 3D scene set, including the base, icons, and dynamic paths, serves as a 3D carbon flow network spatiotemporal evolution model. Then, the model is displayed through visualization rendering. This involves starting the graphics rendering engine, loading and parsing the 3D carbon flow network spatiotemporal evolution model, and drawing the model data onto the screen in real time according to preset rules (such as using color to represent intensity and particle velocity to represent flow rate), generating an interactive 3D interface.

[0065] It should be noted that the three-dimensional carbon flow network spatiotemporal evolution model is a collection of objects that transforms carbon emission data into a three-dimensional visualization scene. It is used to map the carbon source convergence relationship and spatiotemporal changes into an intuitive virtual environment. The visualization is the final visualization effect that is output by the rendering engine as a graphical interface and supports user interaction. It is an interactive system that can intuitively reflect the dynamics of carbon flow.

[0066] In summary, the technical solution adopted in this application can realize dynamic carbon flow tracing and visualization based on data collaborative correction, improve the spatiotemporal resolution of carbon emission control in the park, and thus reduce carbon accounting errors in the park.

[0067] Example 2: This application provides a visualization analysis system for carbon emissions in industrial parks, referring to... Figure 4 As shown, this figure is a module structure diagram of a park carbon emission visualization analysis system provided in this application. The analysis system includes: The multi-source data sensing module 100 is used to acquire energy metering data and satellite remote sensing images of the park; The carbon flow mapping module 200 is used to perform spectral feature analysis on the satellite remote sensing image, thereby identifying emission source intensity data of the emission anomaly area, and constructing a spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationship within the park. The flow field inversion module 300 is used to use the emission source strength data as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix, and to perform flow field inversion on the energy metering data and the carbon flow tracking matrix to obtain a dynamic carbon flow field. The attribution and tracing module 400 is used to perform carbon flow network tracing based on the flow path of the dynamic carbon flow field in the spatial topology to obtain a carbon emission attribution mapping map. The visualization rendering module 500 is used to visualize the carbon emission attribution mapping map based on the visualization rendering driver.

[0068] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0069] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0070] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for visualizing and analyzing carbon emissions in industrial parks, characterized in that, The analytical method includes the following steps: Acquire energy metering data and satellite remote sensing imagery of the park; The spectral features of the satellite remote sensing images are analyzed to identify emission source intensity data in areas of abnormal emissions, and a spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationship within the park are constructed. The emission source strength data is used as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix. The energy metering data and the carbon flow tracking matrix are then used to perform flow field inversion to obtain a dynamic carbon flow field. Based on the flow path of the dynamic carbon flow field in the spatial topology, carbon flow network source tracing is performed to obtain a carbon emission attribution mapping map. The carbon emission attribution mapping map is visualized using a visualization rendering driver.

2. The method for visualizing and analyzing carbon emissions in a park as described in claim 1, characterized in that, Acquiring energy metering data and satellite remote sensing imagery of the park specifically includes: Energy metering data for the park is obtained through metering instruments; Satellite remote sensing images of the park were obtained through a satellite data platform.

3. The method for visual analysis of carbon emissions in industrial parks as described in claim 1, characterized in that, The specific steps for analyzing the spectral features of the satellite remote sensing images to identify emission source intensity data for areas of abnormal emissions include: Radiometric calibration and atmospheric correction were performed on the satellite remote sensing images to obtain surface reflectance image data; Spectral feature matching is performed on the surface reflectance image data to obtain enhanced value data; Morphological filtering is performed on the enhanced value data to obtain a spatial distribution mask of the emission anomaly area; Wind speed field data is acquired, and the enhancement value data within the spatially distributed mask is integrally calculated based on the wind speed field data to obtain the emission source intensity data of the emission anomaly area.

4. The method for visual analysis of carbon emissions in industrial parks as described in claim 1, characterized in that, The construction of the spatial topology and initial carbon flow model characterizing the energy conversion and transmission relationships within the park specifically includes: A spatial topology structure with equipment as nodes and energy flow as edges is constructed by analyzing the spatial location and connectivity of energy equipment in the park. The node attribute dataset is determined based on the energy conversion type of the nodes in the spatial topology. Based on the node attribute dataset and in accordance with the carbon conservation constraint, a set of node carbon potential balance equations is constructed, thereby generating an initial carbon flow model.

5. The method for visualizing and analyzing carbon emissions in a park as described in claim 1, characterized in that, Using the emission source strength data as a spatial constraint to perform inverse correction on the initial carbon flow model, the resulting carbon flow tracking matrix specifically includes: The emission source intensity data is mapped to a spatial topology that characterizes the energy conversion and transmission relationships within the park to obtain a set of node constraints. Run the initial carbon flow model and calculate the residual sequence between the theoretical carbon flow distribution output by the initial carbon flow model and the node constraint set; The initial carbon flow model is assimilated and updated based on the residual sequence using an ensemble Kalman filter algorithm to obtain a reconstructed set of nodal carbon potential balance equations. The carbon potential balance equations of the reconstructed node are solved to obtain the carbon flow tracking matrix.

6. The method for visualizing and analyzing carbon emissions in a park as described in claim 1, characterized in that, The flow field is inverted by performing flow field inversion on the energy metering data and the carbon flow tracing matrix to obtain the dynamic carbon flow field, specifically including: By performing a time-series correlation between the energy metering data and the carbon flow tracing matrix, a system of linear equations is obtained. Boundary condition constraints are added to the linear equation system based on carbon conservation constraints, thereby constructing a carbon flow inversion model; The carbon flow inversion model is solved using a linear programming algorithm to obtain the carbon flow allocation value of each node at each time step. The spatiotemporal kriging interpolation algorithm is used to perform spatial continuity and temporal smoothing on all carbon flow assignment values ​​to obtain a dynamic carbon flow field.

7. The method for visualizing and analyzing carbon emissions in a park as described in claim 1, characterized in that, Based on the flow path of the dynamic carbon flow field in the spatial topology, carbon flow network tracing is performed to obtain a carbon emission attribution mapping map, specifically including: A carbon flow intensity distribution matrix is ​​constructed based on the spatial topology and the dynamic carbon flow field. The graph theory-based flow network decomposition algorithm determines the carbon flow transfer ratio from each source node to the downstream node in the carbon flow intensity distribution matrix, thereby obtaining the carbon flow transfer ratio matrix. The carbon contribution dataset is determined based on the carbon flow transfer ratio matrix and the dynamic carbon flow field. Spatial overlay and aggregation analysis were performed on the carbon contribution dataset to obtain a carbon emission attribution mapping map.

8. The method for visual analysis of carbon emissions in industrial parks as described in claim 1, characterized in that, The visualization of the carbon emission attribution mapping map based on visualization rendering specifically includes: The carbon emission attribution mapping map is converted into a three-dimensional carbon flow network spatiotemporal evolution model based on an open-source 3D geographic information rendering library. The spatiotemporal evolution model of the three-dimensional carbon flow network is visualized using a visualization rendering driver.

9. The method for visual analysis of carbon emissions in industrial parks as described in claim 1, characterized in that, The visualization rendering driver refers to the execution process of a visualization engine that performs real-time analysis of spatiotemporal data and interactive 3D rendering.

10. A park carbon emission visualization analysis system, used to execute a park carbon emission visualization analysis method as described in any one of claims 1 to 9, characterized in that, The analysis system includes: The multi-source data sensing module is used to acquire energy metering data and satellite remote sensing images of the park; The carbon flow mapping module is used to perform spectral feature analysis on the satellite remote sensing images, thereby identifying emission source intensity data in areas of abnormal emissions, and constructing a spatial topology and initial carbon flow model that characterize the energy conversion and transmission relationships within the park. The flow field inversion module is used to use the emission source strength data as a spatial constraint to perform reverse correction on the initial carbon flow model to obtain a carbon flow tracking matrix, and to perform flow field inversion on the energy metering data and the carbon flow tracking matrix to obtain a dynamic carbon flow field. The attribution and tracing module is used to trace the carbon flow network based on the flow path of the dynamic carbon flow field in the spatial topology and obtain a carbon emission attribution mapping map. The visualization rendering module is used to visualize the carbon emission attribution mapping map based on the visualization rendering driver.

Images