Real-time monitoring and tracing method for industrial carbon emission based on multi-source heterogeneous data fusion

Abstract

The application discloses a method for real-time monitoring and tracing of industrial carbon emission based on multi-source heterogeneous data fusion, and relates to the technical field of industrial internet and carbon monitoring, wherein the method comprises the following steps: collecting multi-source data of the park through the Internet of Things gateway, and completing cleaning and time alignment preprocessing; according to the dynamic update factor library of energy index and power grid emission coefficient, combining with material balance calculation, the four-level carbon emission of equipment, production line, workshop and park is calculated step by step, and a carbon flow model is built; when the carbon emission is abnormal, the pipe network and production topology are analyzed based on the graph neural network, the node characteristics are aggregated, and the abnormal cause point is accurately located. The application realizes real-time monitoring and accurate tracing of industrial park carbon emission, solves the problems of data lag and low calculation accuracy of traditional methods, and improves the response speed and tracing accuracy of abnormal emission.

Description

Technical Field

[0001] This invention relates to the fields of industrial internet and carbon monitoring technology, and in particular to a method for real-time monitoring and tracing of industrial carbon emissions based on the fusion of multi-source heterogeneous data. Background Technology

[0002] Industrial parks, as the core carriers of industrial carbon emissions, have carbon management capabilities that are directly related to the achievement of the national "dual-carbon" strategic goals. Among related technologies, a carbon management system covering data collection, energy consumption statistics, and emission accounting has been constructed through the collaborative operation of IoT gateways, DCS systems, and smart meters.

[0003] However, existing carbon management methods for industrial parks directly rely on energy consumption statistics reports submitted by enterprises on a regular basis. This does not enable real-time monitoring and dynamic tracing based on multi-source heterogeneous data. As a result, data may be significantly delayed, difficult to verify, or have insufficient accuracy. This affects the manager's ability to respond quickly to sudden emission anomalies and makes it difficult to accurately trace the total macro emissions to specific equipment or production lines. Summary of the Invention

[0004] The main objective of this invention is to provide a method for real-time monitoring and tracing of industrial carbon emissions based on the fusion of multi-source heterogeneous data.

[0005] Another objective of this invention is to propose a real-time monitoring and tracing device for industrial carbon emissions based on the fusion of multi-source heterogeneous data.

[0006] The third objective of this invention is to provide an electronic device.

[0007] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0008] To achieve the above objectives, a first aspect of the present invention proposes a method for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion, comprising:

[0009] S1 collects multi-source heterogeneous data from the industrial park in real time through an IoT gateway, and cleans and aligns the multi-source heterogeneous data to obtain pre-processed monitoring data. S2, based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, dynamically updates the emission factor library to obtain real-time emission factors that match the current operating conditions; S3 utilizes pre-processed monitoring data and real-time emission factors, combined with material balance algorithms, to calculate the carbon emissions of various equipment, production lines, workshops, and the park within the industrial park level by level, and construct a four-level carbon flow model. S4. When abnormal fluctuations in carbon emission data are detected, the topological relationship between the park's energy pipeline network and production process is analyzed based on graph neural network, and the root cause node causing abnormal carbon emissions is located by aggregating node features.

[0010] Optionally, the multi-source heterogeneous data can be cleaned and time-series aligned, including: A sliding window algorithm is used to remove abnormal noise in the multi-source heterogeneous data, and the data from different acquisition frequencies are timestamped to obtain preprocessed monitoring data.

[0011] Optionally, the emission factor database is dynamically updated based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, including: Marginal emission factors published in real time by the grid:

[0012] in, Let be the marginal emission factor of the power grid at time t. The carbon content of the fuel. For fuel consumption rate, The total output energy is represented by α, which is the weighting coefficient of the energy structure. By combining data from an online calorific value analyzer, the carbon content per unit calorific value of fuel is calculated in real time. Based on the energy structure weighting coefficient, the marginal emission factor of the power grid and the fuel emission factor are weighted and fused to obtain a real-time emission factor that matches the current operating conditions.

[0013] Optionally, the carbon emissions of each piece of equipment, production line, workshop, and the entire industrial park are calculated at each level, including: By using real-time collected data on feed rate, product output, and auxiliary material consumption, combined with the real-time emission factor, the instantaneous carbon emissions of a single piece of equipment are calculated using a material balance algorithm. The data are then aggregated level by level according to equipment, production line, workshop, and park to construct a four-level carbon flow model. Instantaneous carbon emissions of a single device The calculation formula is:

[0014] in, For the first The amount of raw materials input, Its carbon content coefficient, For the first Yield of the product or by-product For the carbon sequestration content of the product, Emissions generated from process energy consumption.

[0015] Optionally, the topological relationships between the park's energy pipeline network and production processes can be analyzed based on graph neural networks, including: The park's energy pipeline network and production process are modeled as a directed graph G=(V,E). By aggregating node features, the root cause nodes leading to the surge in carbon emissions can be quickly located.

[0016] To achieve the above objectives, a second aspect of the present invention provides a real-time monitoring and tracing device for industrial carbon emissions based on multi-source heterogeneous data fusion, comprising: The preprocessing module is used to collect multi-source heterogeneous data from the industrial park in real time through the IoT gateway, and to clean and time-series align the multi-source heterogeneous data to obtain preprocessed monitoring data. The update module is used to dynamically update the emission factor library based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, so as to obtain real-time emission factors that match the current operating conditions. The accounting module is used to calculate the carbon emissions of each piece of equipment, production line, workshop and park in the industrial park by using preprocessed monitoring data and real-time emission factors, combined with the material balance algorithm, and to build a four-level carbon flow model. The analysis module is used to analyze the topological relationship of the park's energy pipeline network and production process based on graph neural networks when abnormal fluctuations in carbon emission data are detected. It locates the root cause node causing the abnormal carbon emission by aggregating node features.

[0017] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0018] To achieve the above objectives, a third aspect of this application provides an electronic device, including a processor and a memory; wherein the processor runs a program corresponding to the executable program code stored in the memory, in order to implement the real-time monitoring and tracing method for industrial carbon emissions based on multi-source heterogeneous data fusion as described in the first aspect embodiment.

[0019] To achieve the above objectives, the fourth aspect of this application proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion as described in the first aspect embodiment.

[0020] The embodiments of the present invention have the following beneficial effects: This invention enables real-time monitoring and precise source tracing of carbon emissions in industrial parks, solving the problems of data lag and low calculation accuracy in traditional methods, and improving the response speed and source tracing accuracy of abnormal emissions. Attached Figure Description

[0021] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart illustrating a method for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion, provided as an embodiment of the present invention; Figure 2 This is a structural diagram of an industrial carbon emission real-time monitoring and tracing device based on multi-source heterogeneous data fusion, provided as an embodiment of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] The following description, with reference to the accompanying drawings, describes a method and apparatus for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion, according to embodiments of the present invention.

[0025] Example 1 This invention provides a method for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion, such as... Figure 1 As shown, the method includes the following steps: S1 collects multi-source heterogeneous data from the industrial park in real time through an IoT gateway, and cleans and aligns the multi-source heterogeneous data to obtain pre-processed monitoring data.

[0026] In order to eliminate noise and time series deviations in multi-source data and ensure the accuracy of carbon accounting, this application collects various heterogeneous data from the park and obtains standardized monitoring data through sliding window cleaning and time series alignment.

[0027] In this embodiment, a global data collection network is built based on IoT gateways deployed in various production sections and energy consumption points in the park, breaking the limitations of single device data sources and comprehensively capturing various heterogeneous information related to park operation.

[0028] In this embodiment, on the one hand, it connects to the factory's DCS distributed control system to synchronize the production load data of each production line unit in real time. It continuously collects sub-item energy consumption data such as water, electricity, and gas by using smart meters and fluid flow meters deployed on site. On the other hand, it connects to the satellite remote sensing platform to obtain carbon dioxide column concentration parameters above the park after inversion calculation. Combined with the site spectral monitoring data brought back by on-site drone inspection operations, it completes the aggregation of multi-source information from three levels: production conditions, energy consumption, and regional atmospheric carbon concentration.

[0029] In the embodiments of this application, the factory configurations and sampling cycles of various monitoring devices are inconsistent, resulting in the collection of raw data containing random noise and abrupt bad values. Furthermore, the high-frequency second-level electricity consumption data and the low-frequency hour-level production data cannot be matched on the time axis. If directly substituted into the calculation, it is very easy to cause the carbon emission accounting results to be distorted.

[0030] In this embodiment, a sliding window algorithm is introduced to complete the outlier filtering operation. By relying on the data distribution pattern within the window, abrupt abnormal noise and invalid mutation data are identified and eliminated, thereby optimizing the sample quality from the perspective of data content.

[0031] After removing outlier data, this application uses a unified time base to perform timestamp calibration and alignment on cross-frequency monitoring data, eliminating time series deviations from different data sources and forming standardized preprocessed monitoring data with consistent time series caliber and compliant content.

[0032] In this embodiment, by collecting heterogeneous data from multiple sources, including production, energy consumption, and environment in the industrial park, and by using sliding window cleaning and timestamp alignment, standardized monitoring data is obtained, laying the foundation for subsequent dynamic updates to the emission factor database.

[0033] S2 dynamically updates the emission factor library based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, obtaining real-time emission factors that match the current operating conditions.

[0034] To overcome the problems of poor adaptability and insufficient calculation accuracy of traditional fixed emission factors, this application combines grid factors and measured fuel parameters to generate real-time emission factors and dynamically updates the factor library to support refined carbon accounting.

[0035] This application departs from the conventional carbon emission accounting model that uses fixed annual average emission coefficients, and establishes a dynamic emission factor update architecture that changes in real time with the energy consumption structure, continuously correcting various parameters in the factor library based on the real-time operating conditions of the industrial park.

[0036] In this embodiment, the application continuously captures the measured energy quality parameters output by the on-site online calorific value analyzer, and simultaneously connects to the power grid data interface to retrieve the officially disclosed marginal emission factor of the power grid at time t in real time. Based on these two types of real-time information, the locally stored emission factor database is continuously refreshed, so that the parameters in the database are adapted to the park's real-time energy consumption structure and production rhythm.

[0037] In this embodiment of the application, the energy consumption of the park is divided into two categories: purchased electricity and self-supplied fossil fuel consumption. The emission coefficients are calculated separately, and then the two types of factors are merged and valued by weighted calculation.

[0038] The formula for calculating the dynamic emission factor is:

[0039] in, Let be the marginal emission factor of the power grid at time t. The carbon content of the fuel. For fuel consumption rate, Let α be the total energy output and α be the weighting coefficient of the energy structure.

[0040] In this embodiment, by using an online calorific value device for continuous sampling and analysis, this application can dynamically correct the carbon content per unit calorific value according to changes in the quality of fuel entering the plant, thus avoiding the problem of emission coefficient distortion caused by fuel quality fluctuations.

[0041] In this embodiment, by accessing real-time grid factors and measured fuel parameters, the emission factor library is dynamically updated through weighted formula calculation to obtain real-time emission factors adapted to the operating conditions, laying the foundation for the subsequent construction of a four-level carbon flow model.

[0042] S3 utilizes pre-processed monitoring data and real-time emission factors, combined with material balance algorithms, to calculate the carbon emissions of various equipment, production lines, workshops, and the park within the industrial park level by level, and construct a four-level carbon flow model.

[0043] In order to refine the hierarchical statistics of carbon emissions in the park and clarify the carbon flow patterns, this application relies on material balance and real-time parameter measurement of single equipment emissions to build a four-level carbon flow accounting model from equipment to park.

[0044] In this application embodiment, the application abandons the extensive overall estimation method and uses a mature and reliable material balance method as the underlying logic of the calculation. It relies on the normalized monitoring data and the dynamic emission factor that matches the real-time operating conditions to realize the layer-by-layer carbon emission statistics from the smallest production unit to the entire park.

[0045] In this embodiment, the application retrieves real-time data on the input of various raw materials, output of main and by-products, and consumption of various auxiliary materials, and combines them with the real-time emission factors generated in the preceding steps to prioritize the accurate calculation of instantaneous carbon emissions from the perspective of a single piece of equipment.

[0046] Instantaneous carbon emissions of a single device The calculation formula is:

[0047] in, Let i be the amount of the i-th raw material input. Its carbon content coefficient, For the yield of product j or by-product, For the carbon sequestration content of the product, Emissions generated from process energy consumption.

[0048] After obtaining the carbon emission results of a single piece of equipment, this application follows a hierarchical structure of collecting data from equipment to production lines, from production lines to workshops, and from workshops to the entire park, and then superimposes the statistical data level by level to summarize the total carbon emissions of each production unit.

[0049] In this embodiment, a four-level carbon flow model is built based on the hierarchical accounting results, which can not only quantify the actual carbon emission values ​​of each level of entity, but also intuitively show the flow path and emission distribution pattern of carbon elements in the industrial park's production chain, thereby achieving full-range and refined real-time carbon emission accounting for the industrial park.

[0050] In this embodiment, the carbon emissions of a single device are calculated by combining material balance algorithm with real-time data, and a four-level carbon flow model is built by summarizing the data level by level, which lays the foundation for subsequent identification of the root cause node causing abnormal carbon emissions.

[0051] S4. When abnormal fluctuations in carbon emission data are detected, the topological relationship between the park's energy pipeline network and production process is analyzed based on graph neural network, and the root cause node causing abnormal carbon emissions is located by aggregating node features.

[0052] In order to quickly identify the causes of abnormal carbon emissions in the park, this application relies on carbon flow time series data to identify anomalies and uses graph neural network topology modeling and feature aggregation to accurately locate the root cause nodes of the anomalies.

[0053] In this embodiment, the stratified carbon emission data output in real time by the four-level carbon flow model is continuously retrieved, and trend comparison and judgment are carried out based on continuous time series data to identify abnormal emission phenomena such as sudden increases in data and irregular large fluctuations in real time, and to promptly lock in abnormal carbon emission events.

[0054] In this embodiment of the application, once an abnormal fluctuation in carbon emissions is determined, the application uses a graph neural network algorithm to complete the topology analysis of the entire park's production and energy supply chain, and traces the source of the anomaly downwards based on the topology association logic.

[0055] In this embodiment of the application, the actual energy transmission pipeline layout and the entire production chain of the park are abstracted and transformed, and a directed graph G=(V,E) topology model is built. V contains all equipment nodes such as boilers, production units, and transmission stations, and the edge set E is used to represent the association logic between nodes in terms of material transportation, energy supply, and process connection.

[0056] In this embodiment, the unique node feature aggregation advantage of graph neural networks is fully utilized to extract multi-dimensional features such as real-time carbon emission indicators and equipment operating parameters of each node and perform fusion calculations. Based on feature differences, abnormal nodes with operating parameters deviating from the rated range and emission indicators rising abnormally are screened out.

[0057] In the embodiments of this application, by combining the transmission law of the topology link and the node characteristics after aggregation, this application can accurately locate the source fault points of various types of faults such as boiler combustion condition deterioration, pipeline loss exceeding the standard, and individual equipment failure exceeding the emission standard, and accurately complete the root cause location of abnormal carbon emissions.

[0058] In this embodiment, emission anomalies are identified by real-time monitoring of carbon flow time-series data. A graph neural network is used to construct a topology map of the park and aggregate node features to accurately locate the root cause nodes of carbon emission anomalies, supporting the park's energy conservation and carbon reduction rectification and optimization.

[0059] In one specific embodiment of the present invention, when implemented in a chemical industrial park, the system collects DCS data from each enterprise via the OPC protocol and updates the emission factor every 5 minutes. When satellite remote sensing detects an abnormal increase in local CO2 concentration in the park, the system automatically triggers a comparison of ground monitoring data, calculates the specific emission source using the above formula, highlights the abnormal equipment in the 3D visualization interface, and notifies maintenance personnel to verify.

[0060] Example 2 This invention provides a real-time monitoring and tracing device 10 for industrial carbon emissions based on multi-source heterogeneous data fusion, such as... Figure 2 As shown, the device includes: The preprocessing module 100 is used to collect multi-source heterogeneous data from the industrial park in real time through the Internet of Things gateway, and to clean and time-series align the multi-source heterogeneous data to obtain preprocessed monitoring data. The update module 200 is used to dynamically update the emission factor library based on the real-time acquired energy quality parameters and the marginal emission factors published by the power grid, so as to obtain real-time emission factors that match the current operating conditions. The accounting module 300 is used to calculate the carbon emissions of each piece of equipment, production line, workshop and park in the industrial park by using pre-processed monitoring data and real-time emission factors, combined with the material balance algorithm, and to build a four-level carbon flow model. The analysis module 400 is used to analyze the topological relationship of the park's energy pipeline network and production process based on graph neural network when abnormal fluctuations in carbon emission data are detected, and to locate the root cause node causing the abnormal carbon emission by aggregating node features.

[0061] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0062] Example 3 To implement the methods of the above embodiments, the present invention also provides an electronic device, which includes a memory and a processor; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the various steps of the methods described above.

[0063] Example 4 To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing embodiments.

[0064] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0065] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0066] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for real-time monitoring and tracing of industrial carbon emissions based on multi-source heterogeneous data fusion, characterized in that, include: S1 collects multi-source heterogeneous data from the industrial park in real time through an IoT gateway, and cleans and aligns the multi-source heterogeneous data to obtain pre-processed monitoring data. S2, based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, dynamically updates the emission factor library to obtain real-time emission factors that match the current operating conditions; S3 utilizes pre-processed monitoring data and real-time emission factors, combined with material balance algorithms, to calculate the carbon emissions of various equipment, production lines, workshops, and the park within the industrial park level by level, and construct a four-level carbon flow model. S4. When abnormal fluctuations in carbon emission data are detected, the topological relationship between the park's energy pipeline network and production process is analyzed based on graph neural network, and the root cause node causing abnormal carbon emissions is located by aggregating node features.

2. The method according to claim 1, characterized in that, The cleaning and time-series alignment of multi-source heterogeneous data includes: A sliding window algorithm is used to remove abnormal noise in the multi-source heterogeneous data, and the data from different acquisition frequencies are timestamped to obtain preprocessed monitoring data.

3. The method according to claim 1, characterized in that, The dynamic updating of the emission factor database based on real-time acquired energy quality parameters and marginal emission factors published by the power grid includes: Marginal emission factors published in real time by the grid: in, Let be the marginal emission factor of the power grid at time t. The carbon content of the fuel. For fuel consumption rate, The total output energy is represented by α, which is the weighting coefficient of the energy structure. By combining data from an online calorific value analyzer, the carbon content per unit calorific value of fuel is calculated in real time. Based on the energy structure weighting coefficient, the marginal emission factor of the power grid and the fuel emission factor are weighted and fused to obtain a real-time emission factor that matches the current operating conditions.

4. The method according to claim 1, characterized in that, The carbon emissions of each piece of equipment, production line, workshop, and the entire industrial park are calculated at each level, including: By using real-time collected data on feed rate, product output, and auxiliary material consumption, combined with the real-time emission factor, the instantaneous carbon emissions of a single piece of equipment are calculated using a material balance algorithm. The data are then aggregated level by level according to equipment, production line, workshop, and park to construct a four-level carbon flow model. Instantaneous carbon emissions of a single device The calculation formula is: in, For the first The amount of raw materials input, Its carbon content coefficient, For the first Yield of the product or by-product For the carbon sequestration content of the product, Emissions generated from process energy consumption.

5. The method according to claim 1, characterized in that, The analysis of the topological relationships between the park's energy pipeline network and production processes based on graph neural networks includes: The park's energy pipeline network and production process are modeled as a directed graph G=(V,E). By aggregating node features, the root cause nodes leading to the surge in carbon emissions can be quickly located.

6. A real-time monitoring and tracing device for industrial carbon emissions based on multi-source heterogeneous data fusion, characterized in that, include: The preprocessing module is used to collect multi-source heterogeneous data from the industrial park in real time through the IoT gateway, and to clean and time-series align the multi-source heterogeneous data to obtain preprocessed monitoring data. The update module is used to dynamically update the emission factor library based on real-time acquired energy quality parameters and marginal emission factors published by the power grid, so as to obtain real-time emission factors that match the current operating conditions. The accounting module is used to calculate the carbon emissions of each piece of equipment, production line, workshop and park in the industrial park by using pre-processed monitoring data and real-time emission factors, combined with the material balance algorithm, and to build a four-level carbon flow model. The analysis module is used to analyze the topological relationship of the park's energy pipeline network and production process based on graph neural networks when abnormal fluctuations in carbon emission data are detected. It locates the root cause node causing the abnormal carbon emission by aggregating node features.

7. An electronic device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the method as described in any one of claims 1-5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-5.

Images