Carbon data processing method and system based on edge calculation, medium and electronic equipment

Abstract

The invention discloses a carbon data processing method and system based on edge computing, a medium and an electronic device, the system comprises a plurality of target devices, a plurality of edge terminals, a control center and an informatization management system, and the method comprises the following steps: the target devices collect and transmit original data; the first edge terminal carries out data fusion, dynamic allocation of calculation tasks, dynamic segmentation of data blocks and dimensionality reduction transmission of preprocessed data in sequence; and the control center analyzes the data. The system comprises an acquisition and transmission module, a fusion module, a dynamic allocation module, a dynamic segmentation module, a dimension reduction and transmission module and an analysis module, multi-source data fusion, resource awareness scheduling and mixed blocking based on data features are completed on the edge side, lightweight work edge completion is achieved, complex calculation is dynamically unloaded to a cloud end, and the efficiency is improved. And real-time performance and computing power requirements are balanced, so that a processing scene of massive carbon data is effectively dealt with.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to carbon data processing methods, systems, media, and electronic devices based on edge computing. Background Technology

[0002] Against the backdrop of current global climate change, carbon emissions have become a focus of widespread international attention. The greenhouse effect caused by carbon emissions is leading to a continuous rise in global temperatures, posing many severe challenges to the ecological environment and sustainable social development. In the power system, traditional power generation methods such as thermal power generation also generate a large amount of carbon emissions. Through carbon data processing technology in carbon monitoring systems, the in-depth development of carbon emission reduction efforts can be effectively promoted.

[0003] Currently, carbon data processing methods mainly include centralized data processing and edge computing-assisted processing. Centralized data processing involves transmitting and aggregating carbon data collected from various monitoring points to a central server for centralized analysis and processing. This method ensures standardized data processing, but data transmission relies on the network, leading to high latency and costs. Furthermore, a failure of the central server can paralyze the entire system, resulting in poor reliability. Traditional edge computing-assisted processing involves deploying edge nodes with basic computing capabilities within the monitoring area. These devices can perform preliminary screening and simple analysis of the collected carbon data before transmitting the processed data to the central server for further processing. This can improve the real-time performance of data processing to some extent, but computational resources are often allocated irrationally. Especially when dealing with massive amounts of carbon data, it is difficult to quickly complete complex real-time analysis tasks and cannot meet the requirements for the real-time performance and accuracy of carbon emission data.

[0004] Therefore, when faced with massive amounts of carbon data, improving data processing efficiency, accuracy of analysis results, and system stability are urgent problems that need to be solved. Summary of the Invention

[0005] To overcome the above problems, the purpose of this invention is to provide a carbon data processing method based on edge computing. This method completes multi-source data fusion, resource-aware scheduling, and hybrid block partitioning based on data characteristics at the edge, enabling lightweight work to be completed at the edge and complex calculations to be dynamically offloaded to the cloud, balancing real-time performance and computing power requirements, thereby effectively addressing the processing scenarios of massive carbon data.

[0006] The technical solution adopted in this invention is:

[0007] The carbon data processing method based on edge computing includes several target devices, several edge terminals, a control center and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center through a lightweight real-time transmission protocol. A network connection supporting parallel task execution is set up between each edge terminal. The edge terminal is an intelligent gateway equipped with a CPU of 32 bits or more and an embedded real-time operating system.

[0008] The carbon data processing method includes the following steps:

[0009] S1: The target device collects the original data in real time and then transmits the original data to the first edge terminal;

[0010] S2: The first edge terminal converts the original data from hexadecimal to decimal, and then merges the original decimal device data according to the spatiotemporal correlation to construct a unified carbon footprint matrix, thus obtaining fused data;

[0011] S3: According to the preset strategy, the first edge terminal will dynamically distribute the fused data to other edge terminals to perform computing tasks in parallel;

[0012] S4: The first edge terminal dynamically divides the computational data into data blocks according to data characteristics;

[0013] S5: The first edge terminal performs dimensionality reduction processing on the data block, and then transmits the dimensionality-reduced preprocessed data to the control center;

[0014] S6: The control center collects and parses the preprocessed data from each edge terminal, and then transmits the parsed data to the information management system.

[0015] As a further description of the present invention, the lightweight real-time transmission protocol communication adopts the MQTT transmission protocol, the Modbus TCP transmission protocol, or the Modbus RTU transmission protocol.

[0016] As a further description of the present invention, the dimensionality reduction of the data block is achieved through key data points to obtain preprocessed data.

[0017] As a further description of the present invention, before step S4, the following step is also included:

[0018] Deploy a lightweight initial prediction model at the first edge terminal;

[0019] The first edge terminal identifies parameters that have a small impact on the output of the initial prediction model through L1 / L2 norm or gradient sensitivity analysis;

[0020] The first edge terminal combines the recognition results to perform iterative structured pruning on the prediction model, resulting in a compressed prediction model;

[0021] The first edge terminal performs real-time analysis and prediction processing of the fused data using a compressed prediction model.

[0022] As a further description of the present invention, before step S4, the following step is also included:

[0023] The first edge terminal statistically analyzes the maximum and minimum values ​​of all 32-bit floating-point numbers in the fused data, and then determines the target value range by combining the values ​​with the preset weighting rules.

[0024] The first edge terminal divides the target value range into 256 target intervals, either uniformly or non-uniformly.

[0025] The first edge terminal combines the target range, converts each floating-point number into a corresponding integer, and then performs the calculation.

[0026] As a further description of the present invention, the CPU of the edge terminal integrates a SIMD instruction set, and the target device and the edge terminal communicate via the Modbus transmission protocol. Step S1 further includes: the target device collects raw data in real time, and then transmits the raw data and Modbus messages to the first edge terminal.

[0027] Before step S2, the following step is also included:

[0028] The first edge terminal uses the SIMD instruction set to parse Modbus messages in batches and extract valid data fields, which include address code, function code, and data field.

[0029] Align valid data fields with memory format and rearrange them;

[0030] Step S2 further includes: the first edge terminal converts the original data from hexadecimal to decimal in batches using the SIMD instruction set, and then merges the converted decimal data according to the valid fields to obtain fused data.

[0031] The carbon data processing system based on edge computing includes several target devices, several edge terminals, a control center and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center through a lightweight real-time transmission protocol. Network connections supporting parallel task execution are set up between each edge terminal. The edge terminal is an intelligent gateway equipped with a CPU of 32 bits or more and an embedded real-time operating system.

[0032] The system also includes:

[0033] The data acquisition and transmission module is used to acquire raw data from the target device in real time and then transmit the raw data to the first edge terminal.

[0034] The fusion module is used by the first edge terminal to convert the original data from hexadecimal to decimal, and then fuse the decimal device metadata according to the spatiotemporal correlation to construct a unified carbon footprint matrix, thus obtaining fused data;

[0035] The dynamic allocation module is used to dynamically allocate the fused data from the first edge terminal to other edge terminals to perform computing tasks in parallel, according to a preset strategy.

[0036] The dynamic segmentation module is used by the first edge terminal to dynamically segment the computational data into data blocks according to data characteristics;

[0037] The dimension reduction transmission module is used by the first edge terminal to perform dimension reduction processing on the data block, and then transmit the pre-processed data after dimension reduction to the control center.

[0038] The parsing and transmission module is used to control the central hub to collect preprocessed data from each edge terminal, parse it, and then transmit the parsed data to the information management system.

[0039] As a further description of the present invention, the CPU of the edge terminal also integrates a cache memory.

[0040] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the edge computing-based carbon data processing method.

[0041] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the aforementioned edge computing-based carbon data processing method.

[0042] The beneficial effects of this invention are:

[0043] This invention relates to a carbon data processing method based on edge computing. This method completes multi-source data fusion, resource-aware scheduling, and hybrid block partitioning based on data features at the edge, enabling lightweight work to be completed at the edge and complex calculations to be dynamically offloaded to the cloud, balancing real-time performance and computing power requirements, thereby effectively addressing the processing scenarios of massive carbon data.

[0044] This invention relates to a carbon data processing method based on edge computing. This method can control the measurement error of carbon emissions within a small range, and the data calculation accuracy rate can reach 95%, ensuring that the data accurately reflects the actual carbon emissions.

[0045] This invention relates to a carbon data processing method based on edge computing. By using edge computing technology for localized processing and ensuring data security, complex computing tasks are distributed to edge nodes to achieve distributed computing, thereby improving the computing and transmission speed of massive carbon data. This ensures that the information management system can keep abreast of carbon emission dynamics, laying the foundation for rapid response and adjustment of energy use strategies, and thus improving the stability of power grid operation. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the hardware-software technology implementation architecture provided in Embodiment 1 of the present invention.

[0047] Figure 2 This is a schematic diagram of the carbon data processing method based on edge computing provided in Embodiment 1 of the present invention.

[0048] Figure 3 This is a schematic diagram of the carbon data processing method based on edge computing provided in Embodiment 2 of the present invention.

[0049] Figure 4 This is a schematic diagram of the carbon data processing method based on edge computing provided in Embodiment 3 of the present invention.

[0050] Figure 5 This is a schematic diagram of the carbon data processing system based on edge computing provided in Embodiment 4 of the present invention.

[0051] Figure 6 This is a line graph showing the difference between the emission factor method and the mass balance method results in experimental test example A of this invention.

[0052] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0053] Explanation of reference numerals in the attached figures

[0054] 4001 - Acquisition and Transmission Module;

[0055] 4002 - Fusion Module;

[0056] 4003 - Dynamic Allocation Module;

[0057] 4004 - Dynamic Segmentation Module;

[0058] 4005 - Dimensional Reduction Transmission Module;

[0059] 4006 - Parsing the transmission module;

[0060] 510-processor;

[0061] 520 - Communication Interface;

[0062] 530 - Memory;

[0063] 540 - Communication bus. Detailed Implementation

[0064] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0065] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0066] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0067] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0068] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0069] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0070] like Figures 1-7 As shown, it illustrates a specific embodiment of the present invention:

[0071] Example 1

[0072] A carbon data processing method based on edge computing, such as Figure 1 The diagram shown is a schematic of the technical implementation architecture provided in this embodiment. The architecture includes several target devices, several edge terminals, a control center, and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center via the MQTT transmission protocol. A network connection supporting parallel task execution is set up between each edge terminal. The edge terminal is an intelligent gateway equipped with a 32-bit CPU and an embedded real-time operating system.

[0073] like Figure 2 The diagram shown is a schematic of the carbon data processing method based on edge computing provided in this embodiment, including the following steps:

[0074] Step S101: The target device collects the original data in real time and then transmits the original data to the first edge terminal.

[0075] In this embodiment, the target device refers to the device that needs to collect carbon data. In the industrial scenario involved in this embodiment, the target device includes, but is not limited to: remote water meters, remote power meters, UPS, emergency energy storage devices, photovoltaic devices, air compressors, and dehumidification devices. The target device generally collects data through sensors and then transmits it. Figure 1 As shown, target devices and edge terminals are generally referred to as sensing devices. The data collected and processed by them can be transmitted to the control center on the business system platform through wired, wireless, or hybrid networking methods.

[0076] Step S102: The first edge terminal converts the original data from hexadecimal to decimal, and then merges the original decimal device data according to the spatiotemporal correlation to construct a unified carbon footprint matrix, thus obtaining fused data.

[0077] In this embodiment, the first edge terminal is a smart gateway specifically designed for this scenario. It includes a 32-bit CPU, an embedded real-time operating system, a wireless module, a Wi-Fi module, a switch module, an Ethernet interface module, a user interface module, an antenna and other interface modules, a SIM / UIM card interface, a power module, etc. Furthermore, the first edge terminal should be understood as one of several edge terminals; the other edge terminals perform the same data fusion, dynamic allocation of computing tasks, dynamic segmentation of data blocks, and dimensionality reduction and preprocessing of data transmission operations.

[0078] Among them, the maximum processing size of the 32-bit CPU is 0x7FFFFFFF, and the maximum processing size of the 64-bit system is 0x7FFFFFFFFFFFFFFF, both of which can effectively support the conversion of raw carbon data from hexadecimal to decimal. Moreover, CPUs with 32 bits or more also support parallel computing, SIMD instruction acceleration, DMA batch transfer, and floating-point precision processing. Compared with serial bit-by-bit calculation, general logic operation, byte-by-byte memory read and write, and software-simulated floating-point operation based on 8 / 16-bit CPUs, the first edge terminal used in this embodiment can better meet the needs of the scenario in terms of speed, energy efficiency, and accuracy.

[0079] Among them, the embedded real-time operating system is an operating system that runs on the gateway hardware. It directly manages the processor (such as a 32-bit CPU), memory, and peripherals (such as a wireless module), and provides real-time scheduling support for upper-layer applications (protocol parsing, data fusion). It ensures that critical tasks are executed first through priority scheduling, thereby laying the foundation for reducing the burden on the system platform server.

[0080] In this embodiment, the carbon data initially collected by the edge terminal is in hexadecimal format, such as carbon dioxide concentration and temperature. Therefore, we need to convert this raw hexadecimal data into decimal format. This improves subsequent computational efficiency and optimizes memory usage. Furthermore, it facilitates integration with decimal device metadata using a spatiotemporal association model, such as device ID and timestamps, to construct a unified carbon footprint matrix. This unifies heterogeneous data into a matrix format, enabling cross-target device correlation analysis, improving processing accuracy, and optimizing edge-side processing latency. The carbon footprint matrix can be formatted as: time × device × parameter; this embodiment does not impose a specific limitation on this format.

[0081] Step S103: According to the preset strategy, the first edge terminal dynamically distributes the fused data to other edge terminals to perform computing tasks in parallel.

[0082] In this embodiment, the aforementioned preset strategy can be a strategy based on data complexity, such as information entropy; a strategy based on network status, such as bandwidth and latency; or a strategy based on a combination of both. According to the preset strategy, the edge terminal can achieve resource awareness, complete dynamically allocated computing tasks, avoid uploading all data to the system platform or processing all data on the edge terminal, thereby improving energy efficiency and reducing response latency.

[0083] Step S104: The first edge terminal dynamically divides the computational data into data blocks according to data characteristics.

[0084] In this embodiment, data characteristics refer to elements such as mutation frequency and sensor type. During the data stability period, large blocks are processed at fixed time intervals, such as dividing the data into blocks at 10-minute window intervals. During the mutation period, fine-grained blocks are triggered. For example, when an anomaly is detected, blocks are triggered at 1-minute window intervals, thereby reducing invalid calculations.

[0085] Step S105: The first edge terminal performs dimensionality reduction processing on the data block, and then transmits the dimensionality-reduced preprocessed data to the control center.

[0086] Step S106: The control center collects and parses the preprocessed data from each edge terminal, and then transmits the parsed data to the information management system.

[0087] In this embodiment, an enterprise information management system refers to a system that integrates various tools, services, and resources to support enterprise operations and management. It is designed to help enterprises achieve process automation, data sharing, and collaboration, including but not limited to: Enterprise Resource Planning (ERP), Product Lifecycle Management (PLM), Manufacturing Execution System (MES), and big data platforms. An information management system is... Figure 1 As shown, the enterprise information management system is located on the business platform and communicates with the control center through data transmission and communication methods such as WEB API, TCP, and MQ.

[0088] In this embodiment, multi-source data fusion, resource-aware scheduling, and hybrid segmentation based on data characteristics are completed at the edge, enabling lightweight tasks to be completed at the edge and complex calculations to be dynamically offloaded to the cloud, balancing real-time performance and computing power requirements, thereby effectively addressing the processing scenarios of massive carbon data.

[0089] Example 2

[0090] In this embodiment, the system includes several target devices, several edge terminals, a control center, and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center via the Modbus TCP transmission protocol. Network connections supporting parallel task execution are established between the edge terminals. Each edge terminal is an intelligent gateway equipped with a 32-bit CPU and an embedded real-time operating system. Figure 3 As shown, this method includes:

[0091] Step S201: The target device collects the original data in real time and then transmits the original data to the first edge terminal.

[0092] In step S202, the first edge terminal converts the original data from hexadecimal to decimal, and then merges the decimal device metadata according to the spatiotemporal correlation to construct a unified carbon footprint matrix, thus obtaining fused data.

[0093] Step S203: Deploy a lightweight initial prediction model on the first edge terminal.

[0094] In step S204, the first edge terminal identifies parameters that have a small impact on the output of the initial prediction model through L1 / L2 norm or gradient sensitivity analysis.

[0095] In step S205, the first edge terminal combines the recognition results to perform iterative structured pruning on the prediction model to obtain a compressed prediction model.

[0096] In step S206, the first edge terminal performs real-time analysis and prediction processing on the fused data using a compressed prediction model.

[0097] In this embodiment, the initial prediction model refers to a lightweight neural network model, such as MobileNet or TinyLSTM, used to predict or classify fused data.

[0098] In this embodiment, neurons or channels with absolute weight values ​​below a threshold can be removed through L1 / L2 norm or gradient sensitivity analysis.

[0099] In this embodiment, structured pruning does not involve arbitrarily deleting parameters, but rather pruning according to specific structural units to maintain the integrity of the network's computational graph. For example, channel pruning directly removes the entire channel of a convolutional layer, ensuring that the input and output dimensions remain aligned; layer pruning directly deletes an entire network layer, but requires adjusting the connections between adjacent layers. Iterative pruning refers to progressively compressing the model according to a "training → pruning → fine-tuning loop," fine-tuning after pruning a small number of parameters in each round, which avoids performance crashes caused by excessive pruning at once. By compressing the prediction model, a regular structure can be established, removing parameters with minimal impact on performance and helping hardware execute the algorithm more efficiently.

[0100] Step S207: The first edge terminal counts the maximum and minimum values ​​of all 32-bit floating-point numbers in the fused data, and then determines the target value range by combining the values ​​with the preset weighting rules.

[0101] In step S208, the first edge terminal divides the target value range into 256 target intervals, either uniformly or non-uniformly.

[0102] In step S209, the first edge terminal combines the target range, converts each floating-point number into a corresponding integer, and then performs the calculation.

[0103] In this embodiment, converting 32-bit floating-point numbers to 8-bit integers or lower helps reduce memory usage and computational load, and can accelerate decimal integer operations. A specific operational example is provided below: The maximum value (25) and minimum value (-10) of all 32-bit floating-point numbers are counted. Based on the weights, a certain weight range is chosen as [-3.8, 5.2]. This range is evenly divided into 256 intervals, i.e., (5.2 - (-3.8)) / 256 ≈ 0.035. When mapping to integers, the floating-point number 2.0 falls in the 170th interval, and is converted to the integer 170. Only 8 bits are needed for storage, saving 75% of memory compared to 32-bit floating-point numbers. During inference, the integer is converted back to a floating-point number, and the integer 170 is dequantized to 170 * 0.035 + (-3.8) ≈ 2.0.

[0104] In step S210, according to a preset strategy, the first edge terminal dynamically distributes the fused data to other edge terminals to perform computing tasks in parallel.

[0105] In step S211, the first edge terminal dynamically divides the computational data into data blocks according to data characteristics.

[0106] In step S212, the first edge terminal performs dimensionality reduction processing on the data block through key data points to obtain preprocessed data.

[0107] In this embodiment, key data points refer to parameters that have core value for carbon emission calculations, such as meter power, temperature, pressure, and carbon dioxide concentration. These are specifically screened and dimensionality reduced through the following steps:

[0108] a. Based on physical correlation and variance analysis, redundant parameters are eliminated;

[0109] b. Align the data from multiple devices by timestamp, construct a spatiotemporal matrix, calculate the principal components, retain the first three dimensions, and compress the data volume.

[0110] By selecting important features or dimensions from the data, the complexity of the data can be reduced while retaining key information.

[0111] In step S213, the control center collects and parses the preprocessed data from each edge terminal, and then transmits the parsed data to the information management system.

[0112] In the embodiments of the present invention, model compression and quantization techniques are used to reduce the size of the algorithm and improve processing efficiency; combined with data collection point planning strategies, high-value feature data is filtered to achieve data simplification while ensuring the integrity of device data; at the same time, based on the edge-side real-time processing architecture, resource consumption and transmission load are effectively reduced, while taking into account both system real-time performance and analysis accuracy.

[0113] Example 3

[0114] This invention includes several target devices, several edge terminals, a control center, and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center via the Modbus RTU transmission protocol. Network connections supporting parallel task execution are established between the edge terminals. Each edge terminal is an intelligent gateway equipped with a 32-bit CPU and an embedded real-time operating system. Figure 4 As shown, this method includes:

[0115] In step S301, the target device collects raw data in real time and then transmits the raw data and Modbus messages to the first edge terminal.

[0116] In step S302, the first edge terminal parses Modbus messages in batches using the SIMD instruction set and extracts valid data fields, which include address code, function code, and data field.

[0117] Step S303: Align the valid data fields with the memory format and rearrange them.

[0118] In step S304, the first edge terminal converts the original data from hexadecimal to decimal in batches using the SIMD instruction set, then merges the converted decimal data according to the valid fields, and then integrates the decimal device metadata according to the spatiotemporal correlation to construct a unified carbon footprint matrix, thus obtaining fused data.

[0119] Step S305: Deploy a lightweight initial prediction model on the first edge terminal.

[0120] In step S306, the first edge terminal identifies parameters that have a small impact on the output of the initial prediction model through L1 / L2 norm or gradient sensitivity analysis.

[0121] In step S307, the first edge terminal combines the recognition results to perform iterative structured pruning on the prediction model to obtain a compressed prediction model.

[0122] In step S308, the first edge terminal performs real-time analysis and prediction processing on the fused data using a compressed prediction model.

[0123] Step S309: The first edge terminal counts the maximum and minimum values ​​of all 32-bit floating-point numbers in the fused data, and then determines the target value range by combining the values ​​with the preset weighting rules.

[0124] In step S310, the first edge terminal divides the target value range into 256 target intervals, either uniformly or non-uniformly.

[0125] In step S311, the first edge terminal combines the target range, converts each floating-point number into a corresponding integer, and then performs the calculation.

[0126] In step S312, according to the preset strategy, the first edge terminal dynamically distributes the fused data to other edge terminals to perform computing tasks in parallel.

[0127] In step S313, the first edge terminal dynamically divides the computational data into data blocks according to data characteristics.

[0128] In step S314, the first edge terminal performs dimensionality reduction processing on the data block through key data points to obtain preprocessed data.

[0129] In step S315, the control center collects and parses the preprocessed data from each edge terminal, and then transmits the parsed data to the information management system.

[0130] In this embodiment, the SIMD instruction set is a parallel computing technique that allows a single instruction to process multiple data elements simultaneously, thereby improving computational efficiency. The SIMD instruction set utilizes spatial parallelism to process multiple data elements concurrently, reducing the number of instruction executions and data accesses.

[0131] In this embodiment, the Modbus message structure is fixed, namely address code 1B, function code 1B, data field N bytes, and CRC 2B. Multiple frames of data are loaded and parsed at once through SIMD instructions, and hardware CRC instructions are used to accelerate verification, allowing for batch parsing of Modbus frames. At the same time, SIMD supports multi-core parallel task allocation, which can dynamically adjust the task allocation of each core when facing massive carbon data traffic.

[0132] Furthermore, incorporating the SIMD instruction set offers numerous benefits for both hardware support and software optimization. Regarding hardware support, writing SIMD instructions leverages the parallel computing capabilities of SIMD registers, improving data processing speed and efficiency. Developing specific algorithms or functions utilizes the acceleration capabilities of dedicated instructions, such as CRC32 and NEON vector instructions, further enhancing computational speed and efficiency. For software optimization, SIMD instructions require data to be aligned to the register width. Enforcing 16-byte alignment in code avoids performance penalties during SIMD loading. Dividing data into SIMD register-sized blocks (e.g., 16 bytes each) adapts to single-instruction, multiple-data operations. Allocating SIMD-intensive tasks to idle cores avoids resource contention and facilitates multi-threaded scheduling.

[0133] In embodiments of this invention, Modbus protocol fields are parsed in batches using the SIMD instruction set, and tasks are dynamically allocated to each core for parallel processing using a multi-core architecture, enabling the processing of multiple frames of data in a single cycle and improving parsing speed. Simultaneously, hardware CRC instructions are used to optimize checksum calculations, reducing end-to-end latency and achieving "protocol features - hardware acceleration - resource scheduling," further optimizing the processing of massive amounts of carbon data.

[0134] Example 4

[0135] like Figure 5 As shown, this embodiment of the invention provides a carbon data processing system based on edge computing, including several target devices, several edge terminals, a control center and an information management system. Each group of target devices communicates with one edge terminal. The edge terminals communicate with the control center through a lightweight real-time transmission protocol. A network connection supporting parallel task execution is set up between each edge terminal. The edge terminal is an intelligent gateway equipped with a 32-bit CPU and an embedded real-time operating system.

[0136] The system also includes:

[0137] The data acquisition and transmission module 4001 is used to acquire raw data from the target device in real time and then transmit the raw data to the first edge terminal.

[0138] The fusion module 4002 is used by the first edge terminal to convert the original data from hexadecimal to decimal, and then fuse the decimal device metadata according to the spatiotemporal correlation to construct a unified carbon footprint matrix to obtain fused data;

[0139] The dynamic allocation module 4003 is used to dynamically allocate the fused data from the first edge terminal to other edge terminals to perform computing tasks in parallel according to a preset strategy.

[0140] The dynamic segmentation module 4004 is used by the first edge terminal to dynamically segment the computational data into data blocks according to data characteristics.

[0141] The dimension reduction transmission module 4005 is used by the first edge terminal to perform dimension reduction processing on the data block, and then transmit the pre-processed data after dimension reduction to the control center.

[0142] The parsing and transmission module 4006 is used to control the central hub to collect preprocessed data from each edge terminal, parse it, and then transmit the parsed data to the information management system.

[0143] As a preferred option, the CPU of the edge terminal also integrates a high-speed cache memory.

[0144] Below, we provide experimental test case A and experimental test case B to further illustrate this embodiment, as follows:

[0145] Table 1. Experimental Environment Configuration Parameters

[0146]

[0147] Experimental Test Example A

[0148] This experiment aims to verify the accuracy and stability of the technical solution in calculating carbon emission factors. By comparing the calculation results of the traditional mass balance method (benchmark method) and the method of this embodiment, the optimization effect of the technical solution is evaluated, as follows:

[0149] 1. Experimental Design

[0150] Test scenario:

[0151] Three typical energy consumption scenarios were selected, covering different emission intensities and data scales, to verify the universality of the technical solution:

[0152] Test Example 1 (High Emission Scenario): Gasoline combustion (482 tons, carbon content 44 GJ / ton), simulating a high carbon emission scenario in industry.

[0153] Test Example 2 (Low Sensitivity Scenario): Diesel consumption (370 tons, lower heating value 2 GJ / ton), to verify the accuracy of the calculation of small values.

[0154] Test Example 3 (Large-scale data scenario): Coal-fired power generation (32322MWh, oxidation rate 95%), testing the massive data processing capability.

[0155] Comparison with benchmark:

[0156] The industry-standard quality balance method is used as a benchmark to ensure the verifiability of the results.

[0157] 2. Implementation of the technical solution

[0158] All experimental examples followed the methods and steps described in the above embodiments, using Java to write the same running code and run it on a computer. Java was used in this experiment mainly because Java has good integration and is relatively simple to implement.

[0159] 3. Results Comparison and Verification

[0160] (1) Calculation of difference rate

[0161] formula:

[0162] Difference rate = |Emission factor method result - Mass balance method result| ÷ Mass balance method result × 100%

[0163] (2) Experimental results (as shown in Table 2 and Figure 6 (as shown)

[0164] Table 2. Verification results of the accuracy of carbon emission factor calculation

[0165]

[0166] like Figure 6 As shown, in the three scenarios, the three broken lines are highly aligned, indicating that the calculation results of the technical solution are highly consistent with the benchmark method, demonstrating that the emission factor and the mass balance method have the same trend; the difference rate is less than 5%, which shows the efficiency and stability of this embodiment.

[0167] 4. Conclusion

[0168] According to the experimental results, the difference rate of the three scenarios was far below the 5% threshold (maximum 0.68%), proving that the technical solution of the present invention has high accuracy and stability in carbon emission accounting.

[0169] Experimental Test Example B

[0170] Based on the existing carbon emission factor difference rate test, the verification of transmission speed and edge computing efficiency has been added to comprehensively evaluate the overall performance of the technical solution in terms of real-time performance and resource optimization.

[0171] 1. Experimental group design

[0172] The test scheme of this technical solution group: The edge terminal processes the data through the gateway, outputs structured JSON data, and transmits it to the control center.

[0173] Control group test scheme: Raw data is transmitted directly, and protocol parsing and calculation are completed in the cloud.

[0174] Test metrics: transmission time, computation time.

[0175] 2. Implementation of the technical solution

[0176] All experimental examples followed the methods and steps described in the above embodiments, using Java to write the same running code and run it on a computer. Java was used in this experiment mainly because Java has good integration and is relatively simple to implement.

[0177] 3. Test Results

[0178]

[0179] (1) Transmission time optimization: The average transmission time was reduced from 2000ms to 500ms, and gateway preprocessing reduced invalid data transmission, resulting in a 75% reduction in transmission time.

[0180] (2) Improved computation efficiency: The average computation time was reduced from 0.45s to 0.35s, and the data cleaning and structuring were completed at the edge, reducing the computation time by 22%.

[0181] Special Note:

[0182] The original data quantity refers to the hexadecimal data before it has been processed by the gateway. It is usually in two sets, such as "0103 04 00 12 D6 87 44 34" and "01 03 04 00 12 D6 87 44 34". After being processed by the gateway, it becomes a single set of hexadecimal data, such as "01 03 04 00 12 D6 87 44 34", which is then converted to decimal to obtain the converted data, such as "12345.67".

[0183] 4. Conclusion

[0184] Data transmission and computation are both controlled within a very short time frame. This improvement ensures that the system can keep abreast of carbon emission dynamics, especially in high-concurrency or large-scale data scenarios, where these differences can accumulate to a significant level, enabling rapid response and adjustment of energy use strategies.

[0185] In this embodiment of the invention, a certain computing power is integrated into the edge terminal to perform real-time analysis and preprocessing of the collected data. When faced with massive amounts of carbon data, it can quickly complete complex real-time analysis tasks and meet the requirements for the real-time performance and accuracy of carbon emission data.

[0186] Example 5

[0187] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 7 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540. The processor 510, communication interface 520, and memory 530 communicate with each other via the communication bus 540. The processor 510 can call logical instructions in the memory 530 to execute the following methods: the target device collects raw data in real time and then transmits the raw data to a first edge terminal; the first edge terminal converts the raw data from hexadecimal to decimal, and then fuses the decimal device metadata according to spatiotemporal correlation to construct a unified carbon footprint matrix, obtaining fused data; according to a preset strategy, the first edge terminal dynamically distributes the fused data to other edge terminals for parallel computation tasks; the first edge terminal dynamically divides the computation data into data blocks according to data characteristics; the first edge terminal performs dimensionality reduction processing on the data blocks and then transmits the dimensionality-reduced preprocessed data to the control center; the control center collects and parses the preprocessed data from each edge terminal and then transmits the parsed data to the information management system.

[0188] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, 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 described in the various embodiments of the present invention. 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.

[0189] Example 6

[0190] In this embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, it implements the multi-version control method under the API framework described in the above embodiment.

[0191] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0192] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0193] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including—but not limited to—wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0194] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, Ruby, and Go, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0195] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

[0196] Many other changes and modifications can be made without departing from the concept and scope of this invention. It should be understood that this invention is not limited to the specific embodiments, and the scope of this invention is defined by the appended claims.

Claims

1. A method for processing carbon data based on edge computing, characterized in that, It comprises a plurality of target devices, a plurality of edge terminals, a control center and an information management system, each group of target devices communicates with an edge terminal, the edge terminal communicates with the control center through a lightweight real-time transmission protocol, network connection supporting parallel execution of tasks is arranged between each edge terminal, and the edge terminal is an intelligent gateway provided with a 32-bit or higher CPU and an embedded real-time operating system; The carbon data processing method comprises the following steps: S1: The target device collects original data in real time, and then transmits the original data to the first edge terminal; S2: The first edge terminal converts the original data from hexadecimal to decimal, fuses the device original data in decimal according to the space-time correlation relationship to construct a unified carbon footprint matrix, and obtains fused data; S3: According to a preset strategy, the first edge terminal dynamically distributes the fused data to other edge terminals to perform a calculation task in parallel; S4: The first edge terminal dynamically divides the calculation data into data blocks according to data characteristics; S5: The first edge terminal performs dimension reduction processing on the data blocks, and then transmits the preprocessed data after dimension reduction to the control center; S6: The control center collects and analyzes the preprocessed data of each edge terminal, and then transmits the analyzed data to the information management system. 2.The edge computing based carbon data processing method according to claim 1, characterized in that, The lightweight real-time transmission protocol communication adopts an MQTT transmission protocol, a Modbus TCP transmission protocol or a Modbus RTU transmission protocol. 3.The edge computing based carbon data processing method according to claim 1, characterized in that, In the step S5, the dimension reduction processing on the data blocks is realized by key data points, and preprocessed data is obtained.

4. The edge computing based carbon data processing method according to any one of claims 1 to 3, characterized in that, Before the step S4, the following steps are further included: A lightweight initial prediction model is deployed on the first edge terminal; The first edge terminal identifies parameters that have a smaller impact on the output of the initial prediction model through L1 / L2 norm or gradient sensitivity analysis; The first edge terminal iteratively performs structured pruning on the prediction model according to the identification result to obtain a compressed prediction model; The first edge terminal performs real-time analysis and prediction processing on the fused data through the compressed prediction model.

5. The edge-computing-based carbon data processing method according to any one of claims 1 to 3, characterized in that, Before the step S4, the following steps are further included: The first edge terminal determines a target value range by combining a preset weight rule and the maximum and minimum values of all 32-bit floating-point numbers in the fused data; The first edge terminal uniformly or non-uniformly divides the target value range into 256 target intervals; The first edge terminal converts each floating-point number into a corresponding integer according to the target interval, and then performs calculation.

6. The edge-computing-based carbon data processing method according to any one of claims 1 to 3, characterized in that, The CPU of the edge terminal integrates a SIMD instruction set, the target device and the edge terminal communicate through a Modbus transmission protocol, and the step S1 further comprises: the target device collects original data in real time, and then transmits the original data and a Modbus message to the first edge terminal; Before the step S2, the following steps are further included: The first edge terminal batch analyzes the Modbus message through the SIMD instruction set to extract an effective data field, and the effective data field comprises an address code, a function code and a data field; The effective data field is aligned with the memory format and is rearranged; The step S2 further comprises: the first edge terminal converting the original data from a hexadecimal batch to a decimal by a SIMD instruction set, and merging the converted decimal data according to valid fields to obtain fusion data.

7. A carbon data processing system based on edge computing, characterized by, The system comprises a plurality of target devices, a plurality of edge terminals, a control hub and an information management system, each group of target devices communicates with an edge terminal, the edge terminals communicate with the control hub through a lightweight real-time transport protocol, network connections supporting parallel execution of tasks are arranged between the edge terminals, and the edge terminals are intelligent gateways provided with a CPU of 32 bits or more and an embedded real-time operating system; The system further comprises: The acquisition and transmission module is configured to collect original data in real time from the target device and transmit the original data to the first edge terminal; The fusion module is configured to convert the original data from a hexadecimal to a decimal by the first edge terminal, and fuse the decimal device metadata according to a space-time correlation relationship to construct a unified carbon footprint matrix, thereby obtaining fusion data. The dynamic allocation module is configured to dynamically allocate the fusion data to other edge terminals for parallel execution of a calculation task according to a preset strategy by the first edge terminal. The dynamic segmentation module is configured to dynamically segment the calculation data into data blocks according to data characteristics by the first edge terminal. The dimension reduction transmission module is configured to perform dimension reduction processing on the data blocks by the first edge terminal, and transmit the dimension-reduced preprocessed data to the control hub. The analysis transmission module is configured to collect the preprocessed data of the edge terminals by the control hub, analyze the preprocessed data, and transmit the analyzed data to the information management system.

8. The edge-computing-based carbon data processing system according to claim 7, characterized in that, The CPU of the edge terminal is further integrated with a high-speed cache memory.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the carbon data processing method based on edge computing according to any one of claims 1 to 6.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the carbon data processing method based on edge computing according to any one of claims 1 to 6.

Images