A Method and System for Monitoring and Analyzing Carbon Emission Factors in Distribution Networks that Integrates Edge Computing

Abstract

This application provides a method and system for monitoring and analyzing carbon emission factors in distribution networks using edge computing, relating to the field of carbon emission monitoring and analysis technology for distribution networks. The method includes a first edge device acquiring operational data and forming a collaborative edge device group, distributing the data and computing tasks to a second edge device within the collaborative edge device group. The second edge device calculates an initial value of the carbon emission factor based on the received data and corrects it through feature exchange between nodes to obtain a target value of the carbon emission factor. The first edge device aggregates multiple rounds of target values ​​and uploads them to the cloud. The cloud combines the network topology and regional division to perform global consistency correction on the multi-round results, ultimately determining the carbon emission factor for the target time period and region, and generating a monitoring and analysis report, providing a reliable basis for regional management and decision-making. This application aims to improve the problems of low accuracy, poor real-time performance, insufficient inter-node collaboration, and the impact of local deviations on global consistency in existing distribution network carbon emission factor monitoring.

Description

Technical Field

[0001] This application relates to the field of power grid carbon emission monitoring and analysis technology, and in particular to a method and system for monitoring and analyzing carbon emission factors in distribution networks that integrates edge computing. Background Technology

[0002] The carbon emission factor can be used to characterize the amount of greenhouse gas emissions per unit of electricity, and its magnitude is closely related to the energy structure, load characteristics, and regional differences of the power system. In practical applications, carbon emission levels in different regions and time periods can be quantified by analyzing distribution network operation data.

[0003] However, existing methods mostly rely on statistical averages or simple load-emission mapping, which are difficult to reflect the dynamic changes of the power system in real time, resulting in insufficient accuracy in carbon emission factor calculation. Secondly, centralized calculation methods require uploading a large amount of distributed operation data to a central server for processing, which has transmission delays and computational bottlenecks, and cannot meet the needs of real-time monitoring. Summary of the Invention

[0004] This application provides a method and system for monitoring and analyzing carbon emission factors in power distribution networks that integrates edge computing. The embodiments of this application adopt the following technical solutions:

[0005] In a first aspect, embodiments of this application provide a method for monitoring and analyzing carbon emission factors in a power distribution network that integrates edge computing, the method comprising:

[0006] The first edge device acquires power distribution network operation data and determines a collaborative edge device group, which includes the first edge device and multiple second edge devices;

[0007] The first edge device acquires the operating status of multiple second edge devices, and sends power distribution network operation data and calculation tasks to the multiple second edge devices based on their operating status;

[0008] The second edge device calculates the initial value of the carbon emission factor based on the power distribution network operation data and calculation tasks, and updates the initial value of the carbon emission factor based on the node characteristic value exchange results with other second edge devices to obtain the target value of the carbon emission factor.

[0009] The first edge device obtains the multi-round carbon emission factor target value calculation results of each second edge device and sends the multi-round carbon emission factor target value calculation results to the cloud device;

[0010] The cloud-based equipment performs global consistency correction based on multiple rounds of carbon emission factor target values ​​to determine the final carbon emission factor values ​​of the distribution network in the target time period and target area, and generates a carbon emission factor monitoring and analysis report based on the final carbon emission factor values.

[0011] In one alternative implementation, determining the collaborative edge device group includes:

[0012] The first edge device sends a network formation request message to other edge devices within the preset area;

[0013] After receiving the networking request message, other edge devices will feed back their own operating status information, node characteristic information, and carbon emission factor calculation accuracy to the first edge device.

[0014] The first edge device constructs a multi-objective optimization model based on the operating status information, node characteristic information, and carbon emission factor calculation accuracy fed back by the second edge device;

[0015] Based on the multi-objective optimization model, the objective optimization solution is determined, and based on the objective optimization solution, edge devices that meet the conditions are dynamically selected from multiple second edge devices to generate a collaborative edge device group.

[0016] In one optional implementation, the first edge device constructs a multi-objective optimization model based on the operating status information, node characteristic information, and carbon emission factor calculation accuracy fed back by the second edge device, including:

[0017] The first optimization sub-model is constructed with the goal of calculating the accuracy of carbon emission factors. The optimization direction of the first optimization sub-model is to minimize the calculation error.

[0018] A second optimization sub-model is constructed with energy consumption expenditure as the objective. The optimization direction of the second optimization sub-model is to minimize energy consumption expenditure.

[0019] A third optimization sub-model is constructed with the task completion delay as the objective. The optimization direction of the third optimization sub-model is to minimize the task completion delay.

[0020] In one optional implementation, power distribution network operation data and calculation tasks are sent to multiple second edge devices based on their operating status, including:

[0021] The first edge device divides the power distribution network operation data into lightweight data segments, and the effective data of the lightweight data segments is smaller than the effective data of the power distribution network operation data.

[0022] Lightweight data fragments and corresponding computational tasks are sent to the second edge device.

[0023] In one alternative implementation, the first edge device divides the distribution network operation data into lightweight data segments, including:

[0024] Perform an upsampling operation on the power distribution network operation data to generate intermediate feature data;

[0025] Perform deconvolution operations on distribution network operation data and intermediate feature data to generate lightweight data fragments.

[0026] In one optional implementation, the second edge device calculates an initial value for the carbon emission factor based on power distribution network operation data and computational tasks, including:

[0027] The second edge device determines the corresponding computational weights based on the feature information of the lightweight data fragments;

[0028] Preprocessing operations are performed on lightweight data fragments based on their calculated weights;

[0029] The preprocessed lightweight data fragments are input into the preset physical mechanism model and data-driven model to generate the first carbon emission factor result and the second carbon emission factor result, respectively. The physical mechanism model is used to characterize the energy flow and power loss mechanism in the operation of the distribution network, and the data-driven model is used to characterize the nonlinear mapping relationship between historical operating data and carbon emission factors.

[0030] A weighted fusion operation is performed on the results of the first carbon emission factor and the second carbon emission factor to obtain the initial value of the carbon emission factor.

[0031] In one optional implementation, the initial value of the carbon emission factor is updated based on the node feature value exchange results with other second edge devices to obtain the target value of the carbon emission factor, including:

[0032] The second edge device sends the characteristic value of this node and the initial value of the carbon emission factor to other second edge devices in the collaborative edge device group;

[0033] The second edge device receives node feature values ​​and initial carbon emission factor values ​​from other second edge devices, and calculates a similarity coefficient based on the node topology.

[0034] The initial value of the carbon emission factor is weighted and corrected based on the similarity coefficient to obtain the target value of the carbon emission factor.

[0035] In one optional implementation, the cloud device performs global consistency correction based on multiple rounds of carbon emission factor target values ​​to determine the final carbon emission factor value of the distribution network in the target time period and target area, including:

[0036] The cloud-based device receives multiple rounds of carbon emission factor target values ​​and, based on the distribution network topology and regional division, aggregates and aligns the carbon emission factor target values ​​by region and time to generate an initial global carbon emission dataset.

[0037] Based on the initial global carbon emission dataset, cloud devices correct the target values ​​of carbon emission factors for each region, so that the carbon emission factors of adjacent regions and time windows before and after remain consistent.

[0038] The cloud-based device outputs the carbon emission factor after consistency correction as the final value of the carbon emission factor according to the target time period and target region.

[0039] Secondly, embodiments of this application provide a power distribution network carbon emission factor monitoring and analysis system integrating edge computing, the system comprising:

[0040] The determination module is used for the first edge device to acquire power distribution network operation data and determine the collaborative edge device group, which includes the first edge device and multiple second edge devices;

[0041] The sending module is used for the first edge device to obtain the operating status of multiple second edge devices, and to send power distribution network operation data and calculation tasks to the multiple second edge devices according to the operating status of the multiple second edge devices;

[0042] The calculation module is used by the second edge device to calculate the initial value of the carbon emission factor based on the power distribution network operation data and calculation tasks, and to update the initial value of the carbon emission factor based on the results of exchanging node characteristic values ​​with other second edge devices, so as to obtain the target value of the carbon emission factor.

[0043] The feedback module is used by the first edge device to obtain the multi-round carbon emission factor target value calculation results of each second edge device and send the multi-round carbon emission factor target value calculation results to the cloud device;

[0044] The analysis module is used by cloud devices to perform global consistency correction based on multiple rounds of carbon emission factor target values, determine the final carbon emission factor values ​​of the distribution network in the target time period and target area, and generate a carbon emission factor monitoring and analysis report based on the final carbon emission factor values.

[0045] In one alternative implementation, the determining module includes:

[0046] The sending submodule is used by the first edge device to send a network request message to other edge devices within a preset area;

[0047] The feedback submodule is used by other edge devices to feed back their own operating status information and node characteristic information to the first edge device after receiving the networking request message;

[0048] The model building submodule is used by the first edge device to build a multi-objective optimization model based on the operating status information and node feature information fed back by the second edge device.

[0049] The determination submodule is used to determine the target optimization solution based on the multi-objective optimization model, and dynamically select edge devices that meet the conditions from multiple second edge devices based on the target optimization solution to generate a collaborative edge device group.

[0050] This application provides a method for monitoring and analyzing carbon emission factors in power distribution networks that integrates edge computing. By constructing a system where a first edge device and multiple second edge devices work collaboratively, and dynamically distributing operational data and computational tasks, each edge node can quickly calculate the initial value of the carbon emission factor locally. Multiple rounds of information exchange between nodes correct node characteristics and calculation results, significantly improving the accuracy and reliability of local calculations. Simultaneously, it fully utilizes edge computing resources, reducing the computational and transmission pressure on the central server and improving the overall real-time performance and response speed of the system. Secondly, the multi-round distributed computation combined with feature value exchange between nodes effectively suppresses the impact of single-node calculation errors or local anomalies on the global results, achieving a balance between local accuracy and global consistency in the carbon emission factor, thus improving the stability and reliability of the distributed monitoring system. Furthermore, after receiving the aggregated data, the cloud device performs global consistency correction based on the network topology and regional division, ensuring the spatial and temporal continuity of carbon emission factors in different regions and time periods, and avoiding the accumulation of local deviations or abnormal fluctuations. Furthermore, this application enables collaborative computation and dynamic optimization among different nodes and regions, supports multi-model fusion, and combines the advantages of physical mechanism models and data-driven models. This ensures that the calculation results not only conform to actual energy flow patterns but also capture nonlinear characteristics in historical operating data, thereby improving the accuracy and scientific rigor of carbon emission factor estimation. In summary, this application not only improves the accuracy, real-time performance, and spatial resolution of carbon emission factor monitoring in distribution networks but also enhances the system's adaptability, scalability, and data reliability. Attached Figure Description

[0051] Figure 1 A flowchart illustrating the steps of a method for monitoring and analyzing carbon emission factors in a power distribution network that incorporates edge computing, as provided in this application embodiment;

[0052] Figure 2 This is a schematic diagram of the structure of a power distribution network carbon emission factor monitoring and analysis system that integrates edge computing, provided as an embodiment of this application. Detailed Implementation

[0053] The terminology used in the following embodiments is for the purpose of describing particular embodiments only and is not intended to be a limitation of this application.

[0054] Traditional methods for monitoring carbon emission factors mainly rely on statistical averages or fixed power flow models, which struggle to reflect the dynamic characteristics of power systems under different load levels, equipment operating states, and seasonal fluctuations. This leads to discrepancies between calculated results and actual emissions, resulting in a lack of high accuracy and real-time performance. Existing centralized computing schemes require uploading large amounts of distributed operational data to a central server for unified processing. Due to the large data transmission volume and limited computing resources, processing delays are common, failing to meet the requirements for real-time monitoring and rapid response. Furthermore, traditional methods often use large regions or the entire distribution network as computing units, making it difficult to account for the heterogeneous characteristics of different nodes or zones. This hinders refined monitoring of carbon emission factors at local nodes, limiting regional emission analysis capabilities. In terms of data processing, edge computing resources are underutilized, with most computing tasks relying on central nodes. This prevents distributed data from being processed quickly locally, increasing the data transmission burden and reducing overall system efficiency. Existing systems do not consider collaborative computing and dynamic task allocation strategies between nodes, making it difficult to interactively correct and integrate calculation results from different nodes. This can easily lead to local biases, affecting the consistency and reliability of global carbon emission factors.

[0055] Therefore, existing technologies suffer from problems such as low accuracy, high latency, low spatial resolution, low computational efficiency, lack of collaborative computing and multi-model fusion capabilities, and lack of uncertainty assessment in the process of monitoring carbon emission factors. There is an urgent need to propose a method for monitoring and analyzing carbon emission factors in distribution networks that can make full use of edge computing resources, combine physical mechanisms and data-driven models, and realize distributed collaborative computing and global consistency correction, so as to improve computational accuracy, reduce computational latency, and realize regional refined management and decision support.

[0056] To address the aforementioned issues, this application proposes the following inventive concept: by introducing a layered architecture design that combines edge computing and cloud collaboration during the data processing of power distribution network operations, the local computing capabilities of distributed edge devices are fully utilized, avoiding the centralized uploading of all data to a central server, thereby reducing communication burden and latency. Specifically, on the edge side, a collaborative edge device group is established, enabling a first edge device to collaborate with multiple second edge devices that have pre-defined relationships with it, achieving local distribution and parallel computing of operational data. After receiving the allocated operational data and computing tasks, the second edge devices can calculate the initial value of the carbon emission factor based on a physical mechanism model and a data-driven model, and gradually converge to obtain a more accurate target value of the carbon emission factor through inter-node feature value exchange and iterative correction. This inter-node interaction and multi-round iteration can effectively overcome the local deviation problem caused by independent calculation by a single node in traditional methods, ensuring the consistency and reliability of the results across the distributed scope.

[0057] At the global level, the first edge device collects and aggregates the target carbon emission factor values ​​from multiple second edge devices and sends the results to the cloud device. The cloud device then performs global consistency correction, comprehensively considering the coupling relationships and uncertainties between different nodes to further improve the global accuracy and reliability of the carbon emission factor results. Ultimately, the cloud can determine the final value of the carbon emission factor within the target time period and target area, and generate a monitoring and analysis report with spatial resolution and temporal granularity, providing decision support for power system dispatching and regional low-carbon management.

[0058] Reference Figure 1 The present invention provides a method for monitoring and analyzing carbon emission factors in a power distribution network that integrates edge computing, which may specifically include the following steps:

[0059] S101: The first edge device acquires power distribution network operation data and determines a collaborative edge device group, which includes the first edge device and multiple second edge devices.

[0060] In this embodiment, the first edge device can be a smart terminal deployed at key nodes of the distribution network, possessing data acquisition and edge computing capabilities, such as a distribution automation terminal (FTU), a distribution master station edge node, or a smart gateway with data processing and task scheduling functions. This device can not only acquire real-time distribution network operation data related to its connected area, such as voltage, current, active power, reactive power, and switch status, but also, based on preset network topology, geographical location association, or operational characteristics, identify and incorporate multiple closely coupled second edge devices into a collaborative edge device group, thereby forming a distributed computing unit covering a local area. Specific steps may include:

[0061] S1011: The first edge device sends a network request message to other edge devices within the preset area;

[0062] S1012: After receiving the networking request message, other edge devices will feed back their own operating status information, node characteristic information, and carbon emission factor calculation accuracy to the first edge device.

[0063] S1013: The first edge device constructs a multi-objective optimization model based on the operating status information, node characteristic information, and carbon emission factor calculation accuracy fed back by the second edge device;

[0064] S1014: Based on the multi-objective optimization model, determine the objective optimization solution, and based on the objective optimization solution, dynamically select edge devices that meet the conditions from multiple second edge devices to generate a collaborative edge device group.

[0065] In the embodiments S1011 to S1014, the purpose of generating a collaborative edge device group is to efficiently distribute and collaboratively execute complex carbon emission factor calculation tasks among multiple edge nodes. The first edge device actively discovers and establishes communication with potential collaborative computing nodes by sending a network request message to other edge devices within a preset area, thereby ensuring a basic task interaction channel in a distributed environment. Subsequently, upon receiving the network request message, other edge devices feed back their own operating status information (such as current load rate, CPU and storage utilization, communication link status, etc.), node characteristic information (such as geographical location, grid topology, typical energy consumption characteristics, etc.), and carbon emission factor calculation accuracy (such as the deviation rate or uncertainty index of historical calculation results) to the first edge device, providing comprehensive data support for subsequent node selection. Based on this, the first edge device uses the collected multi-dimensional feedback information to construct a multi-objective optimization model. The model objective can simultaneously consider factors such as calculation accuracy, communication latency, energy consumption balance, and node redundancy to avoid local optima problems caused by selecting a single indicator. By solving the optimization model, the first edge device can obtain a target optimization solution that balances accuracy and efficiency. Based on this optimization solution, the most suitable node for participating in collaborative computing is dynamically selected from multiple candidate second edge devices, and finally a collaborative edge device group is generated.

[0066] As an example, in a carbon emission factor monitoring scenario of a regional power distribution network, after the first edge device sends a network request message to surrounding nodes, it receives feedback information from multiple second edge devices. Assume the feedback is as follows:

[0067] Edge device A has high computational accuracy and small historical carbon emission factor error, but its current CPU and storage loads are nearing their limits, and further task allocation may increase processing latency. Edge device B has moderate computational accuracy and low resource utilization, allowing it to handle more tasks, but its communication link latency with the first edge device is relatively high, potentially affecting the real-time performance of collaborative computing. Edge device C has moderate computational accuracy and resource utilization, and its communication link is in good condition, making it an ideal candidate node that can stably participate in collaborative computing. Edge device D has slightly lower computational accuracy, but its node load is light and its communication latency is short, making it a suitable auxiliary computing node to handle some tasks and distribute the load, ensuring overall system efficiency. Based on this information, the first edge device uses a multi-objective optimization model to comprehensively consider computational accuracy, node load, communication latency, and redundancy, ultimately dynamically selecting the second edge devices A, C, and D to form a collaborative edge device group. This dynamic selection mechanism ensures computational accuracy while balancing node load and communication efficiency, thereby improving the real-time performance, reliability, and overall collaborative efficiency of carbon emission factor calculation.

[0068] Specifically, the first edge device constructs a multi-objective optimization model based on the operating status information, node characteristic information, and carbon emission factor calculation accuracy fed back by the second edge device, including:

[0069] S10131: Construct the first optimization sub-model with the goal of calculating the accuracy of carbon emission factors. The optimization direction of the first optimization sub-model is to minimize the calculation error.

[0070] S10132: Construct a second optimization sub-model with energy consumption as the objective. The optimization direction of the second optimization sub-model is to minimize energy consumption.

[0071] S10133: Construct a third optimization sub-model with the task completion delay as the objective. The optimization direction of the third optimization sub-model is to minimize the task completion delay.

[0072] In the implementations of S10131 to S10133, when the first edge device constructs a multi-objective optimization model, it will take the carbon emission factor calculation accuracy, energy consumption and task completion latency as core optimization indicators to ensure that the generated collaborative edge device group can take into account accuracy reliability, resource utilization and real-time response capability. Specifically, in S10131, the first optimization sub-model aims at the accuracy of carbon emission factor calculation. By analyzing the calculation deviations of candidate second edge devices in historical or trial tasks, it prioritizes devices with smaller deviations between the calculated results and the actual emission levels to minimize the overall calculation error. In S10132, the second optimization sub-model aims at energy consumption. It evaluates the power consumption, heat dissipation performance, and energy efficiency ratio of different candidate devices during task execution, and selects nodes with high energy efficiency and low energy consumption to participate, thereby reducing the overall operating cost of the system. In S10133, the third optimization sub-model aims at task completion latency. It analyzes the computing power, communication bandwidth, and network topology location of candidate devices to reduce the combined latency of calculation and transmission, thereby improving the real-time performance and response speed of task execution.

[0073] When integrating the three sub-models, the first edge device generates the final target optimization solution through a multi-objective joint optimization method. Specific implementation methods can include two approaches. The first is a weighted joint optimization strategy, where weight coefficients for computational accuracy, energy consumption, and latency are pre-defined, and the weights are dynamically adjusted to address different operating scenarios. For example, the weight for latency is increased during peak loads, while the tolerance for energy consumption is reduced in energy-constrained scenarios, thus achieving flexible scheduling. The second approach is a Pareto optimal solution-based optimization method, which searches for all non-dominant solutions in the three-objective optimization problem and uses the Pareto front to select candidate solutions that achieve the best balance between accuracy, energy consumption, and latency. The optimal device combination is then selected based on the current system operating state. Through these two optimization mechanisms, the first edge device can not only globally balance the conflicting relationships between computational accuracy, energy consumption, and latency but also dynamically adapt to optimization requirements under different operating conditions, thereby generating a collaborative edge device group that meets real-time requirements, reliability, and efficiency.

[0074] S102: The first edge device acquires the operating status of multiple second edge devices, and sends power distribution network operation data and calculation tasks to the multiple second edge devices according to the operating status of the multiple second edge devices.

[0075] In this embodiment, the reason for allocating power distribution network operation data and computing tasks to each second edge device based on their operating status is to fully improve the system's computing efficiency, real-time performance, and accuracy in a distributed edge computing environment. Since the computing resources, storage capacity, and communication link status of each node differ, without dynamic allocation, some nodes may be overloaded and experience delayed task processing, while others may be idle and waste resources. By acquiring the node's load status, communication quality, and historical computing accuracy in real time, the first edge device can prioritize allocating critical data and computing tasks to nodes with sufficient resources, stable communication, and high computing accuracy. Simultaneously, it can allocate smaller or auxiliary tasks to nodes with high load or high communication latency, thereby achieving load balancing and parallel processing. Specifically, after determining the collaborative edge device group, the first edge device acquires real-time operating status information from multiple second edge devices, including parameters such as CPU and storage utilization, current load level, communication link quality, and available computing resources for each node. Based on this information, the first edge device can rationally allocate power distribution network operation data and computing tasks, ensuring load balancing and maximizing computing efficiency among the nodes.

[0076] Sending power distribution network operation data and computational tasks to multiple second edge devices includes:

[0077] S1021: The first edge device divides the power distribution network operation data into lightweight data segments, and the effective data of the lightweight data segments is less than the effective data of the power distribution network operation data;

[0078] S1022: Send lightweight data fragments and corresponding computing tasks to the second edge device.

[0079] In the implementations of S1021 and S1022, the process of sending distribution network operation data and calculation tasks to multiple second edge devices includes two specific steps: First, the first edge device divides the original distribution network operation data into multiple lightweight data segments. Although the effective data volume of each data segment is smaller than the total effective data volume of the original data, it still contains enough information to support the calculation of the carbon emission factor target value, thereby ensuring that each node can independently complete the calculation task. Subsequently, the first edge device sends each lightweight data segment and its corresponding calculation task to the target second edge device for local processing. In this way, not only is the amount of data transmitted per transmission and the communication burden reduced, improving network efficiency, but the computing power of each node can also be fully utilized to achieve distributed parallel computing, thereby ensuring that the carbon emission factor target value can be generated quickly and accurately at the edge nodes and provide reliable data support for subsequent global consistency correction.

[0080] The first edge device divides the power distribution network operation data into lightweight data segments, including:

[0081] Perform an upsampling operation on the power distribution network operation data to generate intermediate feature data;

[0082] Perform deconvolution operations on distribution network operation data and intermediate feature data to generate lightweight data fragments.

[0083] In this embodiment, firstly, an upsampling operation is performed on the original distribution network operation data. This generates intermediate feature data by increasing data sampling points or through interpolation, allowing key information from the original data to be expressed at a higher dimension or finer granularity, thereby improving the distinguishability and expressive power of the features. Subsequently, the original distribution network operation data and the generated intermediate feature data are jointly input into a deconvolution operation. Deconvolution, while compressing and reducing dimensionality, preserves the main patterns and features in the data. This results in lightweight data fragments that significantly reduce data volume and communication load while retaining sufficient key information, enabling the second edge device to independently calculate the carbon emission factor target value. This processing flow not only achieves efficient data transmission and parallel computing between distributed nodes but also ensures the accuracy and real-time performance of the carbon emission factor calculation.

[0084] S103: The second edge device calculates the initial value of the carbon emission factor based on the power distribution network operation data and calculation tasks, and updates the initial value of the carbon emission factor based on the node characteristic value exchange results with other second edge devices to obtain the target value of the carbon emission factor.

[0085] In this embodiment, the second edge device performs multiple rounds of emission factor target value calculation. Each round's calculation result includes the initial carbon emission factor value calculated by the second edge device based on received distribution network operation data and assigned calculation tasks, utilizing its own node characteristic information (such as geographical location, load conditions, and equipment type) to reflect the emission characteristics of that node under its current operating state. Subsequently, the second edge device exchanges its node characteristic values ​​and initial calculation results with other second edge devices in the collaborative group. By comparing and fusing this information, the initial carbon emission factor value is iteratively corrected to obtain a more accurate carbon emission factor target value.

[0086] Through each round of calculation, the second edge device not only calculates the initial value of the carbon emission factor using its own node characteristic information and received operational data, but also iteratively corrects the initial value by exchanging node characteristic values ​​and calculation results with other collaborative nodes. This ensures that the final target value of the carbon emission factor more accurately reflects the emission characteristics of the node under actual operating conditions. Multi-round calculations dynamically consider the mutual influence between nodes. Even if some nodes experience temporary data anomalies or communication delays, subsequent iterations can still correct them using information from other nodes, thereby enhancing the robustness and stability of the system. Simultaneously, multi-round calculations gradually incorporate the latest real-time operational data, enabling dynamic updates of the carbon emission factor and providing a reliable foundation for global consistency correction in the cloud, ensuring the consistency and comparability of the carbon emission factor across the entire network.

[0087] The calculation of the initial value of the carbon emission factor includes:

[0088] S1031: The second edge device determines the corresponding computational weights based on the feature information of the lightweight data fragments;

[0089] S1032: Perform preprocessing operations on lightweight data fragments according to the calculated weights;

[0090] S1033: Input the preprocessed lightweight data fragments into the preset physical mechanism model and data-driven model to generate the first carbon emission factor result and the second carbon emission factor result, respectively. The physical mechanism model is used to characterize the energy flow and power loss mechanism in the operation of the distribution network, and the data-driven model is used to characterize the nonlinear mapping relationship between historical operating data and carbon emission factors.

[0091] S1034: Perform a weighted fusion operation on the first carbon emission factor result and the second carbon emission factor result to obtain the initial value of the carbon emission factor.

[0092] In the embodiments of S1031 to S1034, firstly, the second edge device determines the computational weights based on the feature information of the received lightweight data segments. These computational weights reflect the importance of each data segment in the carbon emission factor estimation; for example, load fluctuations at different nodes, geographical location characteristics, or the reliability of historical data may affect its weight, thus making subsequent processing more targeted. Subsequently, the second edge device performs preprocessing operations on the lightweight data segments according to the determined computational weights. This operation may include data normalization, noise filtering, outlier correction, and feature enhancement to ensure that the input data has consistency, comparability, and stability in the model calculation, avoiding deviations caused by fluctuations in the original data.

[0093] Next, the preprocessed lightweight data fragments are input into the preset physical mechanism model and data-driven model, respectively. The physical mechanism model is used to simulate the physical laws of energy flow, line loss, and transformer efficiency in the distribution network. By calculating parameters such as voltage, current, and power factor, it generates the first carbon emission factor result to characterize the theoretical emission characteristics. The data-driven model utilizes the nonlinear mapping relationship between historical operating data and the carbon emission factor, combined with machine learning or statistical modeling methods, to generate the second carbon emission factor result, thereby reflecting the empirical characteristics under actual operating conditions.

[0094] Finally, the second edge device performs a weighted fusion calculation on the first carbon emission factor result output by the physical mechanism model and the second carbon emission factor result output by the data-driven model. The fusion weight can be dynamically adjusted based on historical accuracy, node reliability, or real-time data quality, thereby generating an initial carbon emission factor value that takes into account both theoretical mechanisms and actual operational characteristics. Through this process, each second edge device can independently generate a high-precision initial estimation result locally, providing a reliable foundation for subsequent multi-round information exchange, iterative correction, and cloud-based global consistency correction with other nodes, achieving distributed, dynamic, and refined carbon emission factor monitoring.

[0095] Based on the node feature value exchange results with other second edge devices, the initial value of the carbon emission factor is updated to obtain the target value of the carbon emission factor, including:

[0096] S1035: The second edge device sends the characteristic value of this node and the initial value of the carbon emission factor to other second edge devices in the collaborative edge device group;

[0097] S1036: The second edge device receives node feature values ​​and initial carbon emission factor values ​​from other second edge devices, and calculates a similarity coefficient based on the node topology relationship;

[0098] S1037: The initial value of the carbon emission factor is weighted and corrected based on the similarity coefficient to obtain the target value of the carbon emission factor.

[0099] In the embodiments of S1035 to S1037, after completing the initial calculation of the local carbon emission factor, the second edge device sends the characteristic value of its node and the corresponding initial carbon emission factor value to other second edge devices in the collaborative edge device group to achieve information sharing among nodes. This sharing not only enables each node to obtain the operating characteristics of adjacent or related nodes, but also provides a reference basis for subsequent joint correction.

[0100] Next, the second edge device receives node characteristic values ​​and initial carbon emission factor values ​​from other edge devices and calculates a similarity coefficient based on the node topology of the distribution network. This similarity coefficient measures the degree of similarity in operating characteristics between different nodes, such as the electrical connection strength, power flow correlation, or correlation of historical operating data between adjacent nodes. By calculating the similarity coefficient, the reference value of each external node in correcting the emission factor results of this node can be quantified.

[0101] Finally, the second edge device uses the calculated similarity coefficient to weight and correct the initial value of the carbon emission factor for this node, thereby generating the target value of the carbon emission factor. In this process, data from nodes with higher similarity receive greater weight, while data from nodes that differ significantly from this node are only used as secondary references. This weighted correction method effectively reduces bias caused by anomalies or local fluctuations in single-node data, ensuring that the final target value of the carbon emission factor retains the operational characteristics of the current node while also incorporating the calculation results from collaborating nodes, thus improving the robustness and consistency of the overall result. This update mechanism will further converge to a reasonable value after multiple iterations, laying the foundation for global consistency correction.

[0102] As an example, assume a collaborative edge device group includes three second edge devices: node A, node B, and node C. After calculating its local initial carbon emission factor value, each node first sends its own node characteristic information and corresponding initial value to other nodes in the group, enabling real-time information sharing among nodes. For example, node A sends its characteristic value and initial value to nodes B and C, and nodes B and C exchange information in the same way, allowing each node to access data from other nodes within the entire collaborative group. Subsequently, each node calculates a similarity coefficient based on the received characteristic values ​​of other nodes and topological relationships, used to measure the degree of similarity in operating characteristics between different nodes. For example, node A's similarity coefficient with node B is 0.8 based on electrical connection strength, power flow correlation, and historical operating data, and its similarity coefficient with node C is 0.5. Finally, node A weights and fuses its own initial carbon emission factor result with the initial values ​​of nodes B and C according to the similarity coefficient, with weights set to 0.6 for node A, 0.3 for node B, and 0.1 for node C, thereby generating a target carbon emission factor value. Through this process, the target value retains the operational characteristics of the node itself while incorporating reference information from collaborating nodes, effectively reducing single-node bias, reflecting the mutual influence between nodes, and achieving robust and high-precision carbon emission factor estimation under distributed computing, providing a reliable foundation for subsequent cloud-based global consistency correction.

[0103] S104: The first edge device obtains the multi-round carbon emission factor target value calculation results of each second edge device and sends the multi-round carbon emission factor target value calculation results to the cloud device.

[0104] In this embodiment, after multiple second edge devices in the collaborative edge device group have completed the calculation of carbon emission factor target values ​​in multiple rounds, the first edge device aggregates the carbon emission factor target values ​​obtained by each node in each round of calculation, forming a multi-round calculation result set. This set not only includes the final target value of each node, but also records the changing trend and convergence of the target value during the multi-round iteration, thereby reflecting the dynamic characteristics and stability of the collaborative calculation between nodes. Subsequently, the first edge device sends the multi-round carbon emission factor target value calculation results to the cloud device, enabling the cloud to obtain comprehensive data of the distributed calculation across the entire network. In this way, the cloud can perform unified analysis and global consistency correction on the multi-round calculation results based on the existing global network topology information and node characteristics, thereby generating the final carbon emission factor of the distribution network in the target time period and target area. This process ensures that the distributed collaborative calculation results of the edge nodes can be effectively integrated, improving the overall calculation accuracy and system response speed, while providing a reliable data foundation for subsequent carbon emission monitoring, analysis, and decision-making.

[0105] S105: The cloud-based equipment performs global consistency correction based on multiple rounds of carbon emission factor target values ​​to determine the final carbon emission factor values ​​of the distribution network in the target time period and target area.

[0106] In this implementation, to ensure the consistency and reliability of the carbon emission factor results obtained from distributed computing across the entire network, the following measures are taken: During multi-node collaborative computing, each second edge device performs calculations based on local data and interactions with neighboring nodes, which may result in local errors, data noise, or incomplete synchronization between nodes. If the target values ​​of each node are directly used as the carbon emission factor for the entire network, the differences between different nodes may not fully reflect the actual operating status, leading to inaccurate overall estimation.

[0107] By performing global consistency correction in the cloud, the target values ​​of each node can be uniformly adjusted, abnormal fluctuations can be eliminated, local differences can be smoothed, and by combining network topology, power flow relationships and historical operating data, it can be ensured that the differences in carbon emission factors between nodes mainly reflect actual operating characteristics rather than calculation errors.

[0108] Specifically, after receiving the multi-round carbon emission factor target values ​​aggregated from the first edge device, the cloud device first integrates and preprocesses the target values ​​of each node. Subsequently, based on the global network topology, the power flow relationship between nodes, and historical operating data, the cloud device executes a global consistency correction algorithm to uniformly correct the carbon emission factor target values ​​of each node, ensuring that the differences between nodes reflect actual operational differences rather than local calculation errors. This correction process can include methods such as weighted averaging, least squares fitting, or iterative adjustment based on graph networks to ensure the rationality and stability of the overall carbon emission factor distribution. Through global consistency correction, the cloud finally determines the final carbon emission factor value of the distribution network in the target time period and target area. This final value considers both the local calculation characteristics of each node and the overall network operating characteristics, achieving a highly accurate and stable estimation of the entire network's carbon emission factor, providing a reliable basis for regional carbon emission monitoring, energy efficiency assessment, and decision support.

[0109] The cloud-based equipment performs global consistency correction based on multiple rounds of carbon emission factor target values ​​to determine the final carbon emission factor values ​​of the distribution network in the target time period and target area, including:

[0110] S1051: The cloud device receives multiple rounds of carbon emission factor target values ​​and, based on the distribution network topology and regional division, aggregates and aligns the carbon emission factor target values ​​by region and time to generate an initial global carbon emission dataset.

[0111] S1052: Based on the initial global carbon emission dataset, the cloud device corrects the target values ​​of carbon emission factors for each region to ensure that the carbon emission factors of adjacent regions and time windows before and after remain consistent.

[0112] S1053: The cloud device will output the carbon emission factor after consistency correction as the final value of the carbon emission factor according to the target time period and target area.

[0113] In the implementations of S1051 to S1053, the cloud device receives multi-round carbon emission factor target values ​​aggregated from each first edge device, and, combined with the topology information of the distribution network and the preset regional division, classifies, aggregates, and aligns the data of each node in time to form an initial global carbon emission dataset. This dataset not only integrates the distributed computing results of different nodes but also ensures the comparability of data from each node in time and space, providing a unified basis for subsequent correction. Subsequently, the cloud device corrects the carbon emission factor target values ​​of each region based on the initial dataset. The correction process includes considering the continuity of power flow and energy consumption between adjacent regions, as well as the carbon emission change trend of previous and subsequent time windows, so that the emission factors of spatially adjacent regions and the emission factors of temporally adjacent time periods are consistent, thereby eliminating discontinuities caused by local fluctuations or inter-node deviations. Finally, the cloud device outputs the carbon emission factors after consistency correction according to the target time period and target region to form the final carbon emission factor value.

[0114] This application provides a method for monitoring and analyzing carbon emission factors in power distribution networks that integrates edge computing. By constructing a system where a first edge device and multiple second edge devices work collaboratively, and dynamically distributing operational data and computational tasks, each edge node can quickly calculate the initial value of the carbon emission factor locally. Multiple rounds of information exchange between nodes correct node characteristics and calculation results, significantly improving the accuracy and reliability of local calculations. Simultaneously, it fully utilizes edge computing resources, reducing the computational and transmission pressure on the central server and improving the overall real-time performance and response speed of the system. Secondly, the multi-round distributed computation combined with feature value exchange between nodes effectively suppresses the impact of single-node calculation errors or local anomalies on the global results, achieving a balance between local accuracy and global consistency in the carbon emission factor, thus improving the stability and reliability of the distributed monitoring system. Furthermore, after receiving the aggregated data, the cloud device performs global consistency correction based on the network topology and regional division, ensuring the spatial and temporal continuity of carbon emission factors in different regions and time periods, and avoiding the accumulation of local deviations or abnormal fluctuations. Furthermore, this method enables collaborative computation and dynamic optimization among different nodes and regions, supports multi-model fusion, and combines the advantages of physical mechanism models and data-driven models. This ensures that the calculation results not only conform to actual energy flow patterns but also capture nonlinear characteristics in historical operational data, thereby improving the accuracy and scientific rigor of carbon emission factor estimation. In summary, this application not only improves the accuracy, real-time performance, and spatial resolution of carbon emission factor monitoring in distribution networks but also enhances the system's adaptability, scalability, and data reliability.

[0115] This application also provides a power distribution network carbon emission factor monitoring and analysis system that integrates edge computing, referring to... Figure 2 The diagram illustrates a functional block diagram of a power distribution network carbon emission factor monitoring and analysis system 200 integrating edge computing, which may include the following modules:

[0116] Secondly, embodiments of this application provide a power distribution network carbon emission factor monitoring and analysis system integrating edge computing, the system comprising:

[0117] The determination module 201 is used for the first edge device to acquire power distribution network operation data and determine the collaborative edge device group, which includes the first edge device and multiple second edge devices;

[0118] The sending module 202 is used for the first edge device to obtain the operating status of multiple second edge devices, and to send power distribution network operation data and calculation tasks to the multiple second edge devices according to the operating status of the multiple second edge devices;

[0119] The calculation module 203 is used by the second edge device to calculate the initial value of the carbon emission factor based on the power distribution network operation data and calculation tasks, and to update the initial value of the carbon emission factor based on the results of exchanging node characteristic values ​​with other second edge devices to obtain the target value of the carbon emission factor.

[0120] Feedback module 204 is used for the first edge device to obtain the multi-round carbon emission factor target value calculation results of each second edge device, and send the multi-round carbon emission factor target value calculation results to the cloud device;

[0121] The analysis module 205 is used by cloud devices to perform global consistency correction based on multiple rounds of carbon emission factor target values, determine the final value of carbon emission factor of the distribution network in the target time period and target area, and generate a carbon emission factor monitoring and analysis report based on the final value of carbon emission factor.

[0122] In one alternative implementation, the determining module includes:

[0123] The sending submodule is used by the first edge device to send a network request message to other edge devices within a preset area;

[0124] The feedback submodule is used by other edge devices to feed back their own operating status information and node characteristic information to the first edge device after receiving the networking request message;

[0125] The model building submodule is used by the first edge device to build a multi-objective optimization model based on the operating status information and node feature information fed back by the second edge device.

[0126] The determination submodule is used to determine the target optimization solution based on the multi-objective optimization model, and dynamically select edge devices that meet the conditions from multiple second edge devices based on the target optimization solution to generate a collaborative edge device group.

[0127] In this embodiment, the present application also provides an electronic device, which may include a memory and one or more processors. The memory and processors are coupled. The memory is used to store computer program code, which includes computer instructions. When the processor executes the computer instructions, the electronic device can perform various functions or steps in the above method embodiments.

[0128] This embodiment also provides a computer-readable storage medium storing computer instructions that, when executed on an electronic device, cause the electronic device to perform the various functions or steps described in the above method embodiments.

[0129] This embodiment also provides a computer program product that, when run on a computer, causes the computer to perform the various functions or steps in the above method embodiments.

[0130] In this embodiment, the electronic device, computer-readable storage medium, and computer program product are all used to execute the corresponding methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0131] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable system. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, a computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0132] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0133] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0134] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, units, and processes described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0135] In the embodiments provided in this application, it should be understood that the disclosed systems, methods, and approaches can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0136] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0137] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0138] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0139] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for monitoring and analyzing carbon emission factors of a power distribution network with fusion edge computing, characterized in that, The method comprises: The first edge device obtains power grid operation data and determines a cooperative edge device group comprising the first edge device and a plurality of second edge devices; The first edge device obtains the operation states of the plurality of second edge devices and sends the power grid operation data and a calculation task to the plurality of second edge devices according to the operation states of the plurality of second edge devices; The second edge device calculates an initial value of a carbon emission factor according to the power grid operation data and the calculation task and updates the initial value of the carbon emission factor according to a node characteristic value exchange result with other second edge devices to obtain a target value of the carbon emission factor; The first edge device obtains a plurality of rounds of calculation results of the target value of the carbon emission factor of each second edge device and sends the plurality of rounds of calculation results of the target value of the carbon emission factor to a cloud device; The cloud device performs global consistency correction according to the plurality of rounds of target values of the carbon emission factor, determines a final value of the carbon emission factor of the power grid in a target period and a target region, and generates a carbon emission factor monitoring and analysis report based on the final value of the carbon emission factor.

2. The method of claim 1, wherein the method further comprises: The determination of the cooperative edge device group comprises: The first edge device sends a networking request message to other edge devices in a preset region; The other edge devices feed back operation state information, node characteristic information and carbon emission factor calculation accuracy of the edge devices to the first edge device after receiving the networking request message; The first edge device constructs a multi-objective optimization model based on the operation state information, node characteristic information and carbon emission factor calculation accuracy fed back by the second edge devices; According to the multi-objective optimization model, a target optimization solution is determined, and edge devices meeting the conditions are dynamically selected from the plurality of second edge devices according to the target optimization solution to generate the cooperative edge device group.

3. The method of claim 2, wherein the method is implemented in a cloud edge computing environment. The first edge device constructs a multi-objective optimization model based on the operation state information, node characteristic information and carbon emission factor calculation accuracy fed back by the second edge devices, comprising: A first optimization sub-model is constructed with the carbon emission factor calculation accuracy as the target, and the optimization direction of the first optimization sub-model is to minimize the calculation error; A second optimization sub-model is constructed with the energy consumption overhead as the target, and the optimization direction of the second optimization sub-model is to minimize the energy consumption overhead; A third optimization sub-model is constructed with the task completion delay as the target, and the optimization direction of the third optimization sub-model is to minimize the task completion delay.

4. The method of claim 1, wherein the method further comprises: The first edge device divides the power grid operation data into lightweight data segments, and the effective data of the lightweight data segments is less than the effective data of the power grid operation data; The lightweight data segments and corresponding calculation tasks are sent to the second edge devices. The first edge device divides the power grid operation data into lightweight data segments, comprising:

5. The method of claim 4, wherein the method further comprises: An upsampling operation is performed on the power grid operation data to generate intermediate feature data; ​ Performing a deconvolution operation on the power grid operation data and the intermediate feature data to generate the lightweight data segment.

6. The method of claim 1, wherein the method further comprises: The second edge device calculates a carbon emission factor initial value according to the power grid operation data and a calculation task, including: The second edge device determines a corresponding calculation weight according to feature information of the lightweight data segment; According to the calculation weight, a preprocessing operation is performed on the lightweight data segment; The preprocessed lightweight data segment is input into a preset physical mechanism model and a data-driven model to generate a first carbon emission factor result and a second carbon emission factor result, respectively, the physical mechanism model is used to represent the energy flow and power loss mechanism in the power grid operation process, and the data-driven model is used to represent the nonlinear mapping relationship between historical operation data and the carbon emission factor; The first carbon emission factor result and the second carbon emission factor result are subjected to a weighted fusion operation to obtain the carbon emission factor initial value.

7. The method of claim 1, wherein the method further comprises: The carbon emission factor target value is obtained by updating the carbon emission factor initial value according to the node feature value exchange result with other second edge devices, including: The second edge device sends the feature value of the node and the carbon emission factor initial value to other second edge devices in the collaborative edge device group; The second edge device receives the node feature value and the carbon emission factor initial value from other second edge devices, and calculates a similarity coefficient based on the node topology relationship; The carbon emission factor target value is obtained by weighting and correcting the carbon emission factor initial value according to the similarity coefficient.

8. The method of claim 1, wherein the method is implemented in a power distribution grid carbon emission factor monitoring and analytics system that is integrated with edge computing. The cloud device performs global consistency correction according to the multiple rounds of carbon emission factor target values to determine the final value of the carbon emission factor of the power grid in the target period and the target region, including: The cloud device receives the multiple rounds of carbon emission factor target values, and performs regional aggregation and time alignment on the carbon emission factor target values according to the power grid topology structure and the regional division to generate an initial global carbon emission data set; The cloud device corrects the carbon emission factor target value of each region based on the initial global carbon emission data set, so that the carbon emission factors of adjacent regions and adjacent time windows are consistent; The cloud device outputs the consistency corrected carbon emission factor as the final value of the carbon emission factor according to the target period and the target region.

9. A power distribution network carbon emission factor monitoring and analysis system fused with edge computing, characterized in that, The system for implementing the method of any one of claims 1-8, comprising: A determination module for a first edge device to obtain power grid operation data and determine a collaborative edge device group, the collaborative edge device group including the first edge device and a plurality of second edge devices; A sending module for the first edge device to obtain the operation states of the plurality of second edge devices and send the power grid operation data and a calculation task to the plurality of second edge devices according to the operation states of the plurality of second edge devices; A calculation module for the second edge device to calculate a carbon emission factor initial value according to the power grid operation data and the calculation task, and update the carbon emission factor initial value according to the node feature value exchange result with other second edge devices to obtain a carbon emission factor target value; The feedback module is configured to acquire, by the first edge device, a plurality of rounds of carbon emission factor target value calculation results of each second edge device, and send the plurality of rounds of carbon emission factor target value calculation results to the cloud device; The analysis module is configured to perform global consistency correction according to the plurality of rounds of carbon emission factor target values, determine a final value of a carbon emission factor of the power distribution network in a target period and a target region, and generate a carbon emission factor monitoring analysis report based on the final value of the carbon emission factor.

10. The distribution grid carbon emission factor monitoring and analytics system with converged edge computing of claim 9, wherein, The determination module comprises: The sending sub-module is configured to send, by the first edge device, a networking request message to other edge devices in a preset region; The feedback sub-module is configured to feed back, by the other edge devices, running state information and node characteristic information of the other edge devices to the first edge device after receiving the networking request message; The model construction sub-module is configured to construct, by the first edge device, a multi-objective optimization model based on the running state information and the node characteristic information fed back by the second edge devices; The determination sub-module is configured to determine a target optimization solution according to the multi-objective optimization model, and dynamically select, from the plurality of second edge devices, an edge device that meets a condition according to the target optimization solution, to generate the collaborative edge device group.

Images