Edge computing gateway device for comprehensive monitoring of multimodal energy and carbon emissions

Abstract

The present invention discloses an edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions, which relates to the technical field related to carbon emission monitoring. The device includes: a carbon emission data acquisition module for acquiring multiple carbon emission sources and multiple carbon emission monitoring indicator sets in an industrial park; a priority evaluation module for performing priority evaluation and acquiring multiple monitoring indicator sequences; a carbon emission source analysis module for analyzing and acquiring multiple predicted emission trends and multiple emission fluctuations; and a carbon emission monitoring module for configuring an optimized data collection strategy for an edge computing gateway and executing carbon emission monitoring in a preset time zone. The device solves the technical problems existing in the prior art, such as limited monitoring range, unclear indicator priority, lagging data processing, and rigid collection strategy, which lead to poor response speed of carbon emission monitoring, and achieves the technical effects of comprehensive and accurate monitoring, optimized resource allocation, real-time trend prediction, dynamic adjustment of strategies, and improved response speed of carbon emission monitoring.

Description

Technical Field

[0001] The present application relates to the technical field related to carbon emission monitoring, and specifically to an edge computing gateway device for comprehensive monitoring of multi-modal energy carbon emissions. Background Art

[0002] Industrial parks, as concentrated areas of industrial production activities, are key targets for carbon emissions control. Accurate monitoring and effective management of their carbon emissions are crucial to achieving carbon reduction targets. However, existing industrial park carbon emissions monitoring systems have significant deficiencies in data processing and analysis capabilities, making it difficult to meet the complex needs of multimodal data processing. On the one hand, traditional carbon emissions monitoring cannot fully reflect the true carbon emissions status of industrial parks. At the same time, there is a lack of a scientific priority evaluation mechanism for different carbon emissions monitoring indicators, resulting in an inability to rationally allocate resources during the data collection process, which not only wastes resources but also affects monitoring efficiency and the accuracy of monitoring results. On the other hand, existing carbon emissions monitoring relies on cloud computing, which has large delays in data transmission, making it impossible to timely predict and analyze carbon emission trends, making it difficult to quickly respond to sudden carbon emission fluctuations. There is a lack of accurate analysis of emission trends and fluctuations of carbon emission sources, making it impossible to flexibly adjust data collection strategies according to actual conditions.

[0003] Therefore, at the current stage, relevant technologies have technical problems such as limited monitoring scope, unclear indicator priority, delayed data processing, and rigid collection strategies, which lead to poor response speed of carbon emission monitoring. Summary of the Invention

[0004] This application solves the technical problems in the existing technology, such as limited monitoring scope, unclear indicator priority, delayed data processing, and rigid collection strategy, which lead to poor response speed of carbon emission monitoring, by providing an edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions. It achieves the technical effects of comprehensive and accurate monitoring, optimized resource allocation, real-time trend prediction, dynamic strategy adjustment and improved response speed of carbon emission monitoring.

[0005] The present application provides an edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions, the device comprising: a carbon emission data acquisition module for acquiring multiple carbon emission sources and multiple carbon emission monitoring indicator sets of an industrial park; a priority evaluation module for sequentially performing priority evaluations on multiple carbon emission monitoring indicators in the multiple carbon emission monitoring indicator sets to acquire multiple monitoring indicator sequences; a carbon emission source analysis module for analyzing and acquiring multiple predicted emission trends and multiple emission fluctuations of the multiple carbon emission sources within a preset time zone based on the carbon emission monitoring logs of the industrial park; a carbon emission monitoring module for configuring an optimized data collection strategy of the edge computing gateway according to the multiple monitoring indicator sequences, multiple predicted emission trends and multiple emission fluctuations, and executing carbon emission monitoring in the preset time zone, wherein the data collection strategy includes monitoring indicator type, data collection frequency and key feature screening threshold.

[0006] In a possible implementation, the edge computing gateway device for multimodal energy carbon emission comprehensive monitoring also performs the following processing: randomly selecting a first carbon emission monitoring indicator set of a first carbon emission source to obtain multiple first carbon emission monitoring indicators; obtaining multiple first unit monitoring data volumes of the multiple first carbon emission monitoring indicators; performing indicator correlation and data credibility evaluation on the multiple first carbon emission monitoring indicators according to the carbon emission monitoring log of the first carbon emission source, and determining multiple first indicator correlations and multiple first data credibility; performing priority evaluation on the multiple first carbon emission monitoring indicators according to the multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility, constructing a first monitoring indicator sequence, and analyzing in sequence to obtain multiple monitoring indicator sequences.

[0007] In a possible implementation, the edge computing gateway device for multimodal energy carbon emission comprehensive monitoring also performs the following processing: according to the carbon emission monitoring log of the first carbon emission source and the performance of the edge computing gateway, the data volume weight, correlation weight and credibility weight are configured based on the coefficient of variation method; after dimensionless processing of the multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility, multiple first priorities are obtained based on the data volume weight, correlation weight and credibility weight, wherein the priority is negatively correlated with the unit monitoring data volume and positively correlated with the indicator correlation and data credibility; based on the multiple first priorities, the multiple first carbon emission monitoring indicators are arranged from large to small according to the first priority to obtain a first monitoring indicator sequence.

[0008] In a possible implementation, the edge computing gateway device for multimodal energy carbon emission comprehensive monitoring also performs the following processing: randomly selecting a first carbon emission source to obtain a first carbon emission monitoring log of the first carbon emission source; searching the first carbon emission monitoring log for the same time period within a preset time range based on the preset time zone to obtain multiple sample first carbon emission monitoring sequences; analyzing the multiple sample first carbon emission monitoring sequences to obtain a first predicted emission trend and a first emission volatility, and sequentially analyzing to obtain multiple predicted emission trends and multiple emission volatility of the multiple carbon emission sources.

[0009] In a possible implementation, the edge computing gateway device for multimodal energy carbon emission comprehensive monitoring also performs the following processing: in a two-dimensional coordinate system, construct multiple sample first carbon emission curves based on the multiple sample first carbon emission monitoring sequences, and perform curve fitting to determine a first fitted carbon emission curve; calculate a first increase ratio based on the first fitted carbon emission curve, and set it as a first predicted emission trend; perform mean calculation on the multiple sample first carbon emission monitoring sequences respectively to obtain the mean of the multiple sample first carbon emissions, and perform mean calculation again to obtain the first predicted carbon emissions; perform emission fluctuation analysis based on the multiple sample first carbon emission means to calculate the first emission fluctuation, wherein the emission fluctuation is the ratio of the standard deviation of the mean of the multiple sample first carbon emissions to the first predicted carbon emissions.

[0010] In a possible implementation, the edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions also performs the following processing: mapping and obtaining the first edge computing gateway of the first carbon emission source; determining the first fluctuation scale based on the first predicted emission trend, the first predicted carbon emission amount and the first emission fluctuation analysis; adjusting the first initial data collection plan based on the first fluctuation scale to obtain a first optimized data collection plan; sequentially analyzing and obtaining multiple optimized data collection plans of multiple edge computing gateways to construct an optimized data collection strategy.

[0011] In a possible implementation, the edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions also performs the following processing: calculating the ratio of the first predicted carbon emissions to the historical average of the first carbon emissions of the first carbon emission source, setting it as the first emission scale; calculating the ratio of the first predicted emission trend to the historical average of the first emission trend, and obtaining a trend adjustment coefficient; calculating the ratio of the first emission volatility to the historical average of the first emission volatility, and obtaining an emission fluctuation adjustment coefficient; using the trend adjustment coefficient and the emission fluctuation adjustment coefficient, double compensation is performed on the first emission scale, and the first fluctuation scale is output.

[0012] In a possible implementation, the edge computing gateway device for comprehensive monitoring of multimodal energy and carbon emissions also performs the following processing: configuring a first initial data collection plan, wherein the first initial data collection plan includes a first monitoring indicator selection quantity, a first data collection frequency, and a first key feature screening threshold; adjusting the first monitoring indicator selection quantity according to the first fluctuation scale to determine a first optimized selection quantity, and selecting the first optimized selection quantity of monitoring indicators in the first monitoring indicator sequence to obtain a first monitoring indicator type; adjusting the first data collection frequency according to the first fluctuation scale to determine a first optimized collection frequency; adjusting the first key feature screening threshold according to the first fluctuation scale to determine a first optimized feature screening threshold; constructing a first optimized data collection plan based on the first monitoring indicator type, the first optimized collection frequency, and the first optimized feature screening threshold.

[0013] In a possible implementation, the edge computing gateway device for comprehensive monitoring of multimodal energy carbon emissions also performs the following processing: setting the inverse of the first fluctuation scale as the first threshold adjustment coefficient, adjusting the first key feature screening threshold, and obtaining the first optimized feature screening threshold.

[0014] In a possible implementation, the edge computing gateway device for multimodal energy carbon emission comprehensive monitoring also performs the following processing: when performing carbon emission monitoring according to the first monitoring indicator type and the first optimized collection frequency, it is determined whether the monitoring data deviation under the adjacent monitoring nodes is greater than the first optimized feature screening threshold; if it is greater, the monitoring data under the adjacent monitoring nodes is retained as key data; if it is less than, the monitoring data under the adjacent monitoring nodes is discarded, and the judgment and screening are performed iteratively.

[0015] The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring proposed in this application is intended to include a carbon emission data acquisition module for acquiring multiple carbon emission sources and multiple carbon emission monitoring indicator sets in an industrial park; a priority evaluation module for performing priority evaluation and acquiring multiple monitoring indicator sequences; a carbon emission source analysis module for analyzing and acquiring multiple predicted emission trends and multiple emission fluctuations; and a carbon emission monitoring module for configuring the optimized data collection strategy of the edge computing gateway and executing carbon emission monitoring in a preset time zone. This solves the technical problems in the prior art of limited monitoring scope, unclear indicator priority, lagging data processing, and rigid collection strategies, which lead to poor response speed in carbon emission monitoring, and achieves the technical effects of comprehensive and accurate monitoring, optimized resource allocation, real-time trend prediction, dynamic strategy adjustment, and improved response speed in carbon emission monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the accompanying drawings of the embodiments of the present disclosure are briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the apparatus according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in precise order. Instead, various steps may be processed in reverse order or simultaneously as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0017] Figure 1 Schematic diagram of the structure of an edge computing gateway device for comprehensive monitoring of multimodal energy and carbon emissions provided in an embodiment of the present application.

[0018] Figure 2 Schematic diagram of the execution process of the priority evaluation module in the edge computing gateway device for comprehensive monitoring of multimodal energy and carbon emissions provided in an embodiment of the present application.

[0019] Explanation of the reference numerals: carbon emission data acquisition module 10 , priority evaluation module 20 , carbon emission source analysis module 30 , carbon emission monitoring module 40 . DETAILED DESCRIPTION

[0020] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0021] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0022] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, and the terms “first\second” involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, device, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0023] The embodiment of the present application provides an edge computing gateway device for comprehensive monitoring of multi-modal energy and carbon emissions, such as Figure 1 As shown, the device includes:

[0024] The carbon emission data acquisition module 10 is used to acquire multiple carbon emission sources and multiple carbon emission monitoring indicator sets of the industrial park.

[0025] Preferably, multiple carbon emission sources and multiple carbon emission monitoring indicator sets are obtained from the industrial park. By comprehensively identifying various types of carbon emission sources, all-dimensional emission entities in the park are covered. Each carbon emission source may include multiple emission monitoring indicators. The more indicators, the higher the accuracy of carbon emission monitoring and analysis. Specifically, carbon emission sources refer to specific objects in the industrial park that produce carbon dioxide or other greenhouse gas emissions, including production equipment, energy facilities (heating stations, substations, self-contained power plants, etc. in the park), building systems (heating, air conditioning, lighting, etc. of office buildings, factory buildings, etc.), transportation vehicles (fuel or gas-powered vehicles such as freight vehicles, forklifts, construction machinery, etc. in the park), industrial production processes (chemical synthesis, steel smelting, cement production, etc.) and waste treatment (sewage treatment plants, solid waste landfills, etc.), etc. Among them, production equipment includes boilers, kilns, reactors and other industrial equipment that directly burn fossil fuels (coal, natural gas, oil) to produce carbon emissions.

[0026] Preferably, the carbon emission monitoring indicators include multi-dimensional and multi-modal monitoring parameters of different carbon emission sources, which are used to quantify carbon emission levels and evaluate emission intensity. Specifically, they include direct emission indicators obtained through metering equipment, which reflect the amount of greenhouse gases produced by carbon emission sources, such as the mass or volume emissions of gases such as CO2, CH4, N2O, and carbon emissions per unit product or unit output value; indirect emission indicators, which reflect energy consumption or process parameters related to carbon emission sources, such as energy consumption indicators such as electricity consumption, water consumption, and steam consumption, process parameters such as reaction temperature, pressure, and raw material ratio, and equipment operating status such as equipment load rate, start-stop frequency, and energy efficiency level; environmental management indicators, which reflect the environmental impact of carbon emissions, such as CO2 concentration and PM2.5 concentration in the atmosphere around the park, and carbon reduction target indicators such as carbon emission reduction rate per unit output value and clean energy substitution ratio. Taking the boiler in the industrial park as an example, its carbon emissions are determined by multiple factors such as its combustion efficiency, fuel type, and type of emission gas. Monitoring indicators may include CO2 concentration, NO x Emissions, PM2.5 concentration, boiler exhaust gas temperature, energy (natural gas or coal) consumption, boiler exhaust gas flow and pressure, etc.

[0027] The priority evaluation module 20 is configured to sequentially evaluate the priorities of the plurality of carbon emission monitoring indicators in the plurality of carbon emission monitoring indicator sets to obtain a plurality of monitoring indicator sequences.

[0028] Furthermore, the specific configuration of the priority evaluation module 20 also includes randomly selecting a first carbon emission monitoring indicator set of the first carbon emission source to obtain multiple first carbon emission monitoring indicators; obtaining multiple first unit monitoring data volumes of the multiple first carbon emission monitoring indicators; performing indicator correlation and data credibility evaluation on the multiple first carbon emission monitoring indicators according to the carbon emission monitoring log of the first carbon emission source, and determining multiple first indicator correlations and multiple first data credibility; performing priority evaluation on the multiple first carbon emission monitoring indicators according to the multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility, constructing a first monitoring indicator sequence, and analyzing in sequence to obtain multiple monitoring indicator sequences.

[0029] Preferably, the first carbon emission source is a specific emission source (such as a boiler, a production line, etc.) randomly selected from multiple carbon emission sources in the industrial park, and the first carbon emission monitoring indicator set is a collection of all monitoring indicators configured for the emission source. The first carbon emission monitoring indicator set is composed of multiple first carbon emission monitoring indicators. A single emission source is used as the minimum analysis unit, and the priority of its monitoring indicators is evaluated one by one to finally form an indicator priority sequence for the entire park; then the historical data flow statistics are directly obtained through the sensor or device interface, that is, the first unit monitoring data volume corresponding to each first carbon emission monitoring indicator, which represents the amount of data generated by a single indicator per unit time. For example, if the CO2 emission indicator generates 100 monitoring data per hour, then its corresponding unit monitoring data volume is 100 pieces / hour, wherein the unit monitoring data volume directly affects the data transmission cost, storage pressure and computing load of the edge computing gateway.

[0030] Preferably, the carbon emission monitoring log of the first carbon emission source is obtained, which includes historical monitoring data and related information of the first carbon emission source (such as a boiler), such as time series data of each indicator, equipment operation records (such as start and stop time, load status), sensor calibration records, data transmission status (such as whether there is a network disconnection resulting in data loss) and manually marked abnormal events (such as equipment failure, process adjustment, etc.); then, based on the carbon emission monitoring log, the indicator correlation evaluation is performed on multiple first carbon emission monitoring indicators, that is, the correlation coefficient (such as Pearson coefficient) of each carbon emission monitoring indicator and the carbon emission amount is calculated. The closer the absolute value of the coefficient is to 1, the higher the correlation, and then the correlation is represented by a value from 0 to 1 (such as 0.8 represents strong correlation, and 0.3 represents weak correlation). The higher the value, the more important the indicator is to carbon emission analysis.

[0031] Preferably, based on the carbon emission monitoring log, a credibility assessment is performed on multiple first carbon emission monitoring indicators respectively, including assessing the accuracy (the degree of closeness of the carbon emission monitoring indicator data to the true value), integrity (the degree of continuity of the carbon emission monitoring indicator data in the time series, whether there are any missing or interrupted) and timeliness (the degree of delay from data collection to data transmission to the edge computing gateway) of each carbon emission monitoring indicator. Specifically, the accuracy assessment refers to comparing the monitoring data through manual sampling and detection, calculating the error rate, or analyzing whether the data of different monitoring points of the same indicator match; the integrity assessment includes counting the proportion of missing time of indicator data over a period of time (such as the past 30 days), and the number and duration of data collection interruptions caused by equipment failure, network interruption, etc.; the timeliness assessment refers to calculating the time difference between the sensor collection time and the gateway reception time, or monitoring whether the carbon emission monitoring indicator is stably collected at a preset frequency (such as once per minute), and whether there is a frequency fluctuation (such as occasional omission resulting in a longer interval), and then using a value from 0 to 1 to represent the credibility (such as 0.9 for high credibility and 0.5 for medium credibility). The higher the value, the more reliable the data quality.

[0032] Preferably, priority evaluation is performed on multiple first carbon emission monitoring indicators based on multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility. Specifically, weights of multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility are set according to business needs. A negative value for unit data volume indicates that the larger the data volume, the higher the negative impact on resource consumption and the lower the priority may be. Then, weighted calculation is performed and sorted from high to low according to the priority score to construct the first monitoring indicator sequence of the carbon emission source. Finally, all carbon emission sources in the industrial park are evaluated and calculated to obtain the indicator priority sequence corresponding to each emission source, and finally multiple monitoring indicator sequences are formed.

[0033] Furthermore, the specific configuration of the priority evaluation module 20 also includes configuring the data volume weight, correlation weight and credibility weight based on the coefficient of variation method according to the carbon emission monitoring log of the first carbon emission source and the performance of the edge computing gateway; after dimensionless processing of the multiple first unit monitoring data volumes, multiple first indicator correlations and multiple first data credibility, multiple first priorities are obtained based on the data volume weight, correlation weight and credibility weight, wherein the priority is negatively correlated with the unit monitoring data volume, and positively correlated with the indicator correlation and data credibility; based on the multiple first priorities, the multiple first carbon emission monitoring indicators are arranged from large to small according to the first priority to obtain a first monitoring indicator sequence.

[0034] Preferably, the coefficient of variation method determines the weight by the degree of discreteness of the data itself, and is suitable for measuring the relative importance of different indicators. Among them, the data volume weight reflects the pressure of the unit monitoring data volume on the edge computing gateway resources (such as storage and transmission bandwidth). The larger the data volume, the greater the impact on the gateway performance, and the priority needs to be adjusted through the weight; the correlation weight measures the degree of correlation between the indicator and the core characteristics of carbon emissions (for example, the direct correlation between energy consumption data and carbon emissions is usually higher than that of ambient temperature and humidity data). The higher the correlation, the greater the contribution to monitoring carbon emissions, and the higher the weight should be; the credibility weight evaluates the reliability of the indicator data (such as sensor accuracy and historical data integrity). Data with low credibility may introduce noise, and its weight needs to be reduced to avoid interfering with the monitoring results.

[0035] Preferably, according to the carbon emission monitoring log of the first carbon emission source and the performance of the edge computing gateway, the data volume weight, correlation weight and credibility weight are configured based on the coefficient of variation method. Specifically, the carbon emission monitoring log is analyzed, and the historical data fluctuations of each indicator (such as data missing rate, outlier frequency) are counted. Combined with the performance parameters of the edge computing gateway (such as maximum data processing capacity, storage capacity), the data volume weight, correlation weight and credibility weight are calculated by the coefficient of variation method. For example, if the data volume of a carbon emission monitoring indicator fluctuates greatly and occupies more gateway resources, its data volume weight may be higher; if a certain indicator has a strong correlation with the carbon emission trend (such as fuel consumption), its correlation weight will be higher.

[0036] Preferably, the multiple first unit monitoring data volumes, multiple first indicator correlations, and multiple first data credibility levels are dimensionlessly processed. That is, they are converted into directly comparable dimensionless values ​​through a standardization method (such as normalization or Z-score standardization) to prevent dimensional differences from affecting the calculation results. Multiple first priorities are then weighted and calculated based on the data volume weight, correlation weight, and credibility weight. The priority is negatively correlated with the unit monitoring data volume and positively correlated with the indicator correlation and data credibility. In other words, the larger the unit monitoring data volume, the lower the priority (because high data volume may increase the gateway burden, requiring a trade-off in resource allocation); the higher the indicator correlation and the higher the data credibility, the higher the priority (these indicators are more critical to the effectiveness and accuracy of carbon emission monitoring). Finally, all first carbon emission monitoring indicators are arranged from high to low according to the calculated first priorities to form a first monitoring indicator sequence, thereby achieving differentiated management of carbon emission monitoring indicators and ensuring that the edge computing gateway prioritizes the most valuable data for carbon emission monitoring within limited resources, thereby improving monitoring efficiency and accuracy.

[0037] The carbon emission source analysis module 30 is used to analyze and obtain multiple predicted emission trends and multiple emission fluctuations of the multiple carbon emission sources within a preset time zone based on the carbon emission monitoring log of the industrial park.

[0038] Furthermore, the specific configuration of the carbon emission source analysis module 30 also includes randomly selecting a first carbon emission source to obtain a first carbon emission monitoring log of the first carbon emission source; searching the first carbon emission monitoring log for the same time period within a preset time range based on the preset time zone to obtain multiple sample first carbon emission monitoring sequences; obtaining a first predicted emission trend and a first emission volatility based on analysis of the multiple sample first carbon emission monitoring sequences, and successively analyzing to obtain multiple predicted emission trends and multiple emission volatility of the multiple carbon emission sources.

[0039] Preferably, from multiple carbon emission sources in the industrial park (such as factory workshops, energy equipment, vehicles, etc.), one is randomly selected as the first carbon emission source (such as the boiler system of a factory), and the first carbon emission monitoring log of the carbon emission source is obtained, that is, a record set of its historical carbon emission data, including timestamp, carbon emissions, and related influencing factors (such as production load, energy consumption, equipment operating status, etc.); then the time zone is set according to the region where the industrial park is located or business needs, and a preset time range (such as the past 1 year, 1 quarter, 1 month, etc.) is configured to limit historical data retrieval, and then within the preset time range, carbon emission data of repeated time periods are extracted according to a fixed cycle (such as 8:00-10:00 every day, weekdays every week, etc.) to form multiple sample first carbon emission monitoring sequences.

[0040] Preferably, based on multiple sample first carbon emission monitoring sequences, time series analysis (such as moving average and exponential smoothing) is used to fit the long-term trend of carbon emissions over time, thereby obtaining a first predicted emission trend and a first emission volatility. Specifically, time series modeling is performed on each sample sequence (e.g., emissions during the same time period each day), and trend terms (such as increasing, decreasing, or stable) are extracted. The trend characteristics of all samples are then combined to obtain an overall predicted trend for the carbon emission source during the same time period. For example, if most sample sequences show a linear increase in carbon emissions during the morning period with increasing production load, the predicted trend is "increasing emissions within the time period." Statistical indicators such as standard deviation, coefficient of variation, and range are then calculated for the emission data of the multiple sample sequences to obtain a first emission volatility. This volatility reflects the degree of dispersion or stability of carbon emissions during the same time period, reflecting the volatility of the emission process (such as sudden peaks and abnormal fluctuations). A larger value indicates more severe emission fluctuations, while a smaller value indicates more stable emissions. All carbon emission sources within the industrial park (e.g., the first, second, and third carbon emission sources) are analyzed and calculated sequentially to obtain multiple predicted emission trends and multiple emission volatility measures. This will then form a carbon emission characteristic map of the entire industrial park, identify sources of high-trend growth (such as production lines with continuously rising emissions) and sources of high volatility (such as equipment that operates intermittently), and facilitate carbon emission control and energy optimization scheduling.

[0041] Furthermore, the specific configuration of the carbon emission source analysis module 30 also includes, in a two-dimensional coordinate system, constructing multiple sample first carbon emission curves based on the multiple sample first carbon emission monitoring sequences, and performing curve fitting to determine a first fitted carbon emission curve; calculating a first increase ratio based on the first fitted carbon emission curve, and setting it as a first predicted emission trend; performing mean calculation on the multiple sample first carbon emission monitoring sequences respectively to obtain the mean of the multiple sample first carbon emissions, and performing mean calculation again to obtain the first predicted carbon emissions; performing emission fluctuation analysis based on the multiple sample first carbon emission means to calculate the first emission fluctuation, wherein the emission fluctuation is the ratio of the standard deviation of the mean of the multiple sample first carbon emissions to the first predicted carbon emissions.

[0042] Preferably, a two-dimensional coordinate system is constructed with the horizontal axis as time (such as in minutes, hours or days) and the vertical axis as carbon emissions, and each sample sequence (such as emission data within the same period of a certain day) is plotted as a line graph, with the horizontal axis corresponding to the time point and the vertical axis corresponding to the emission at that moment. Multiple sample first carbon emission curves are obtained, and curve fitting is performed on the multiple sample curves through mathematical methods (such as the least squares method) to obtain a smooth curve (i.e., the first fitted carbon emission curve) that can reflect the overall trend, eliminating the interference of random fluctuations, and used to characterize the typical emission trend of the carbon emission source within a preset time period.

[0043] Preferably, a first increase ratio is calculated based on the first fitted carbon emission curve, i.e., first increase ratio = (emissions at the end of the fitted curve − emissions at the starting point of the fitted curve) / emissions at the starting point of the fitted curve × 100%. If the result is a positive value, it indicates an increasing trend in emissions; if it is a negative value, it indicates a decreasing trend; if it is close to 0, it indicates a stable trend, and the first increase ratio is set as the first predicted emission trend. A double mean calculation is then performed, i.e., the emissions mean is calculated for each sample sequence (e.g., a time period of a particular day), and the mean of all sample means is again calculated to obtain a first predicted carbon emissions. This reflects the average emission level of the carbon emission source during the preset time period and serves as a benchmark for predicting emissions during the same time period in the future.

[0044] Preferably, an emission volatility analysis is performed based on the mean of the first carbon emissions of multiple samples. That is, the standard deviation of the mean of the first carbon emissions of multiple samples is used as the numerator and the first predicted carbon emissions is used as the denominator. The ratio of the two is calculated as the first emission volatility. The numerator (standard deviation) is used to measure the absolute dispersion of the mean of each sample and the overall mean. The larger the standard deviation, the greater the difference in emissions between different samples. The denominator (predicted carbon emissions) converts the absolute dispersion into a relative value, eliminating the impact of the magnitude of emissions on the volatility assessment. The volatility reflects the stability of carbon emissions. If the volatility is low (such as less than 10%), it indicates that the emission process is highly regular and is less affected by random factors. If the volatility is high (such as greater than 40%), it indicates that there is significant uncertainty in the emissions, which may be caused by equipment startup and shutdown, raw material changes, process adjustments, etc.

[0045] The carbon emission monitoring module 40 is used to configure the optimized data collection strategy of the edge computing gateway according to the multiple monitoring indicator sequences, multiple predicted emission trends and multiple emission fluctuations, and perform carbon emission monitoring in the preset time zone, wherein the data collection strategy includes the monitoring indicator type, data collection frequency and key feature screening threshold.

[0046] Furthermore, the specific configuration of the carbon emission monitoring module 40 also includes mapping and obtaining the first edge computing gateway of the first carbon emission source; determining the first fluctuation scale based on the first predicted emission trend, the first predicted carbon emission amount and the first emission fluctuation analysis; adjusting the first initial data collection plan according to the first fluctuation scale to obtain a first optimized data collection plan; analyzing in turn to obtain multiple optimized data collection plans of multiple edge computing gateways, and constructing an optimized data collection strategy.

[0047] Preferably, the optimized data collection strategy of the edge computing gateway is configured according to multiple monitoring indicator sequences, multiple predicted emission trends and multiple emission fluctuations, that is, the edge computing gateway constructs a dynamic data collection strategy by integrating the monitoring indicator sequences, predicted emission trends and emission fluctuations, which specifically includes the screening of monitoring indicator types, the configuration of data collection frequency and the optimization of key feature screening thresholds. Among them, the screening of monitoring indicator types includes giving priority to the collection of high-priority indicators (such as indicators with correlation > 0.8 and credibility > 0.9) to ensure that the core characteristics of carbon emissions are effectively monitored; for low-priority indicators (such as indicators with correlation < 0.3 or credibility < 0.6), the collection is suspended or triggered only when the equipment is abnormal.

[0048] Preferably, the configuration of data collection frequency includes increasing the collection frequency of key indicators if the emission trend is rising, and decreasing the collection frequency of non-key indicators if the emission trend is falling; if the emission fluctuation is greater than 30%, the collection frequency of all indicators is ≥1 time / minute; if the emission fluctuation is less than 10%, the collection frequency of conventional indicators is ≤1 time / 5 minutes. Optimization of key feature screening thresholds includes adjusting the anomaly detection threshold and setting the data compression threshold. For example, the anomaly detection threshold is dynamically adjusted based on historical emission data, with the normal fluctuation range being ±2 times the mean standard deviation. When the fluctuation increases, the threshold is relaxed (such as ±3 times the standard deviation) to reduce false alarms. When the fluctuation decreases, the threshold is tightened (such as ±1.5 times the standard deviation) to increase sensitivity. For indicators that change slowly (such as the temperature of stable operating equipment), a larger differential threshold is set to reduce the amount of data transmitted.

[0049] Preferably, a mapping relationship is established between the carbon emission source and its corresponding edge computing gateway. Each carbon emission source (such as a boiler, a production line) corresponds to one or more edge computing gateways, which are responsible for local data collection and preprocessing. For the randomly selected first carbon emission source (such as a key boiler), the mapping is used to determine the execution entity of its data collection, that is, the first edge computing gateway; then the first fluctuation scale is determined based on the first predicted emission trend, the first predicted carbon emissions and the first emission fluctuation analysis, that is, the first predicted emission trend, the first predicted carbon emissions and the first emission fluctuation are comprehensively analyzed to judge the fluctuation scale of carbon emissions. For example, if the trend is rising and the fluctuation is large, it means that the emissions are unstable and the fluctuation scale is large; if the trend is stable and the fluctuation is small, it means that the emissions are stable and the fluctuation scale is small.

[0050] Preferably, the first initial data collection plan is adjusted according to the first fluctuation scale, wherein the initial data collection plan refers to a preset collection rule, such as collecting all indicators every 30 minutes. Specifically, if the fluctuation scale is large, the collection frequency is increased (such as adjusting from 30 minutes to 10 minutes), the monitoring indicators are refined (such as adding new equipment operating status, fuel consumption and other related indicators), and the key feature screening threshold is lowered (such as retaining more detailed data to avoid filtering important fluctuation information). If the fluctuation scale is small, the collection frequency is reduced (such as adjusting from 30 minutes to 1 hour), the monitoring indicators are simplified (only carbon emissions and emission concentrations are retained), and the key feature screening threshold is increased (filtering noise data and focusing on major trends). Through dynamic adjustment, the first optimized data collection plan is obtained to ensure that it is more in line with the actual emission characteristics and balances data accuracy and computing resource efficiency.

[0051] Preferably, the edge computing gateway corresponding to each carbon emission source is repeatedly analyzed to generate multiple optimized data collection schemes (for example, gateway X is responsible for high-volatility sources and adopts high-frequency collection; gateway Y is responsible for stable sources and adopts low-frequency collection). Finally, all optimized data collection schemes are integrated into a data collection strategy, and the monitoring indicator type (a list of indicators to be collected in different scenarios), data collection frequency (dynamically adjusted frequency rules) and key feature screening thresholds (standards for data filtering and feature extraction) are clearly defined, thereby improving the efficiency and accuracy of carbon emission monitoring.

[0052] Optimally, carbon emissions monitoring is performed in preset time zones based on the determined optimized data collection strategy. This involves deploying the configured collection strategy to edge nodes, enabling localized data processing and reducing cloud reliance. For example, an edge gateway is deployed in the industrial park's power distribution room to process power consumption and carbon emissions data in real time. Time zones are divided according to business activity patterns (e.g., production and non-production time zones), with differentiated execution strategies. After each preset time zone is monitored, the edge computing gateway automatically evaluates the effectiveness of the strategy. If the actual emission fluctuation differs from the forecast by more than 20%, the weight is recalculated and the collection frequency is adjusted. If no anomalies are detected for a certain indicator for three consecutive time zones, its collection priority is lowered. This achieves the transition from extensive data collection to precise intelligent monitoring, ensuring efficient and reliable carbon emissions monitoring and management in industrial parks.

[0053] Furthermore, the specific configuration of the carbon emission monitoring module 40 also includes calculating the ratio of the first predicted carbon emissions to the historical average of the first carbon emissions of the first carbon emission source, setting it as the first emission scale; calculating the ratio of the first predicted emission trend to the historical average of the first emission trend to obtain a trend adjustment coefficient; calculating the ratio of the first emission volatility to the historical average of the first emission volatility to obtain an emission fluctuation adjustment coefficient; using the trend adjustment coefficient and the emission fluctuation adjustment coefficient to perform double compensation on the first emission scale and output the first fluctuation scale.

[0054] Preferably, the ratio of the first predicted carbon emissions to the average of the historical first carbon emissions of the first carbon emission source is calculated as the first emission scale. If the first emission scale is greater than 1, it indicates that the current predicted emissions are higher than the historical average level. If the first emission scale is equal to 1, it indicates that it is the same as the historical level. If the first emission scale is less than 1, it indicates that it is lower than the historical level. The ratio of the first predicted emission trend to the average of the historical first emission trends is calculated to obtain a trend adjustment coefficient. The first predicted emission trend is the increase ratio obtained by curve fitting, and the average of the historical first emission trend is the average of the emission trends over a certain period of time in the past. The ratio of the first emission volatility to the average of the historical first emission volatility is calculated to obtain an emission volatility adjustment coefficient. The first emission volatility represents the degree of fluctuation of the current emissions, and the average of the historical first emission volatility represents the average of the historical volatility. Finally, the trend adjustment coefficient and the emission fluctuation adjustment coefficient are used to perform double compensation on the first emission scale, that is, the first fluctuation scale = trend adjustment coefficient × emission fluctuation adjustment coefficient × first emission scale. Among them, if the current emission trend is significantly higher than the historical average (such as accelerated emissions due to increased production), the assessment value of the fluctuation scale will be magnified, indicating the need for strengthened monitoring; if the current fluctuation amplitude increases (such as frequent equipment start-up and shutdown), the fluctuation scale will be further magnified, reflecting the increased uncertainty of the emission process.

[0055] Furthermore, the specific configuration of the carbon emission monitoring module 40 also includes configuring a first initial data collection scheme, wherein the first initial data collection scheme includes a first monitoring indicator selection quantity, a first data collection frequency and a first key feature screening threshold; adjusting the first monitoring indicator selection quantity according to the first fluctuation scale to determine a first optimized selection quantity, and selecting the first optimized selection quantity of monitoring indicators in the first monitoring indicator sequence to obtain a first monitoring indicator type; adjusting the first data collection frequency according to the first fluctuation scale to determine a first optimized collection frequency; adjusting the first key feature screening threshold according to the first fluctuation scale to determine a first optimized feature screening threshold; and constructing a first optimized data collection scheme based on the first monitoring indicator type, the first optimized collection frequency and the first optimized feature screening threshold.

[0056] Preferably, the first initial data collection scheme is configured with the first number of monitoring indicators selected, the first data collection frequency and the first key feature screening threshold, wherein the first number of monitoring indicators selected is the total number of preset collected indicators (such as 20 indicators are initially selected, including emissions, energy consumption, equipment status, etc.), the first data collection frequency is a fixed collection interval (such as once per hour) or cycle (such as 8 times per day), and the first key feature screening threshold is a preset standard for data filtering or feature extraction, such as only retaining data points with a fluctuation range of >5%, or compressing indicator values ​​with a change of <2%.

[0057] Preferably, the number of selected first monitoring indicators is adjusted according to the first fluctuation scale, specifically, by the formula , where k is the adjustment coefficient (set according to the business scenario). When the fluctuation scale is greater than 1, the number of optimizations increases, and new highly correlated indicators (such as fuel composition, process temperature, etc.) are added. For example, if the fluctuation scale is 2 and the initial number is 10, the number of indicators selected for the first optimization is 10×(1+0.5×2)=20. When the fluctuation scale is ≤1, the number of optimizations decreases, and non-critical indicators may be reduced after rounding. Then, the first N indicators (where N is a positive integer, indicating the number of optimizations) are selected from the monitoring indicator sequence (sorted by priority) to obtain the first monitoring indicator type, ensuring that high-value indicators are collected first.

[0058] Preferably, the first data collection frequency is adjusted according to the first fluctuation scale to determine the first optimized collection frequency. Specifically, the larger the fluctuation scale, the larger the first optimized collection frequency. If the fluctuation scale exceeds the threshold, a high-frequency collection mode is triggered, such as real-time collection. Conversely, the smaller the fluctuation scale, the smaller the first optimized collection frequency. If the fluctuation scale is lower than the threshold, the energy-saving mode is entered, such as collecting data once every 4 hours. The first key feature screening threshold is adjusted according to the first fluctuation scale to determine the first optimized feature screening threshold. Specifically, the larger the fluctuation scale, the smaller the first optimized feature screening threshold, that is, smaller fluctuation data is retained, improving monitoring accuracy; the smaller the fluctuation scale, the larger the first optimized feature screening threshold, that is, more noise data is filtered, reducing invalid storage. Finally, the first monitoring indicator type, the first optimized collection frequency, and the first optimized feature screening threshold are combined to construct a first optimized data collection scheme, that is, a personalized data collection strategy for the first carbon emission source.

[0059] Furthermore, the specific configuration of the carbon emission monitoring module 40 also includes setting the inverse of the first fluctuation scale as a first threshold adjustment coefficient, adjusting the first key feature screening threshold, and obtaining a first optimized feature screening threshold.

[0060] Preferably, the inverse of the first fluctuation scale is set as the first threshold adjustment coefficient, that is, the larger the fluctuation scale, the smaller the threshold adjustment coefficient, and the looser the optimized threshold (allowing larger data fluctuations); the smaller the fluctuation scale, the larger the threshold adjustment coefficient, and the stricter the optimized threshold (limiting the data fluctuation range); then, the first key feature screening threshold is adjusted to obtain the first optimized feature screening threshold. Specifically, the expression of the first optimized feature screening threshold is: ;

[0061] Among them, based on the real-time fluctuation characteristics of carbon emissions, the data screening standards are adaptively adjusted to improve the efficiency and reliability of the edge computing gateway while ensuring monitoring accuracy.

[0062] Furthermore, the specific configuration of the carbon emission monitoring module 40 also includes, when performing carbon emission monitoring according to the first monitoring indicator type and the first optimized collection frequency, judging whether the monitoring data deviation under adjacent monitoring nodes is greater than the first optimized feature screening threshold; if it is greater, retaining the monitoring data under the adjacent monitoring nodes as key data; if it is less than, discarding the monitoring data under the adjacent monitoring nodes, and iteratively performing judgment and screening.

[0063] Preferably, when carbon emissions monitoring is performed according to the first monitoring indicator type and the first optimized collection frequency, it is determined whether the monitoring data deviation under adjacent monitoring nodes is greater than the first optimized feature screening threshold, wherein adjacent monitoring nodes refer to two consecutive data collection points in the time series, the monitoring data deviation refers to the numerical difference of the same indicator between adjacent nodes, and the first optimized feature screening threshold is a threshold that is dynamically adjusted according to the fluctuation scale; specifically, if the monitoring data deviation under adjacent monitoring nodes is greater than the first optimized feature screening threshold, the monitoring data under the adjacent monitoring nodes are retained as key data; if the monitoring data deviation under the adjacent monitoring nodes is less than the first optimized feature screening threshold, the monitoring data under the adjacent monitoring nodes are discarded; and judgment and screening are performed iteratively, that is, all adjacent monitoring nodes are judged, and data are screened pair by pair, and finally a key data sequence containing only significant fluctuations is retained; thereby enabling the edge computing gateway to capture the most valuable emission fluctuation information with the smallest amount of data to ensure efficient and reliable carbon emission monitoring and management in the industrial park.

[0064] Although the present application makes various references to certain modules in the apparatus according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0065] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

Claims

1. An edge computing gateway device for comprehensive monitoring of multi-modal energy and carbon emissions, characterized by: The edge computing gateway device includes: A carbon emission data acquisition module is used to obtain multiple carbon emission sources and multiple carbon emission monitoring indicator sets of the industrial park; a priority evaluation module, configured to sequentially evaluate the priorities of the plurality of carbon emission monitoring indicators in the plurality of carbon emission monitoring indicator sets to obtain a plurality of monitoring indicator sequences; A carbon emission source analysis module is used to analyze and obtain multiple predicted emission trends and multiple emission fluctuations of the multiple carbon emission sources within a preset time zone based on the carbon emission monitoring log of the industrial park; A carbon emissions monitoring module, configured to configure an optimized data collection strategy for an edge computing gateway based on the multiple monitoring indicator sequences, multiple predicted emission trends, and multiple emission fluctuations, and perform carbon emissions monitoring in the preset time zone, wherein the data collection strategy includes monitoring indicator type, data collection frequency, and key feature screening threshold; The steps performed by the carbon emission monitoring module include: Calculate the ratio of the first predicted carbon emissions to the average of the first historical carbon emissions of the first carbon emission source, and set it as the first emission scale; Calculating a ratio of a first predicted emission trend to an average of a first historical emission trend to obtain a trend adjustment coefficient, wherein the first predicted emission trend is a first increase ratio calculated based on the first fitted carbon emission curve; Calculating a ratio of a first emission volatility to an average of historical first emission volatility to obtain an emission volatility adjustment coefficient, wherein the first emission volatility is calculated by performing an emission volatility analysis based on an average of first carbon emissions of multiple samples; Using the trend adjustment coefficient and the emission fluctuation adjustment coefficient, double compensation is performed on the first emission scale to output a first fluctuation scale; The steps performed by the carbon emission monitoring module also include adjusting the number of first monitoring indicator selections, the first data collection frequency, and the first key feature screening threshold according to the first fluctuation scale.

2. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 1 is characterized in that: The steps performed by the priority evaluation module include: Randomly selecting a first carbon emission monitoring indicator set of a first carbon emission source to obtain a plurality of first carbon emission monitoring indicators; Obtaining a plurality of first unit monitoring data amounts of the plurality of first carbon emission monitoring indicators; According to the carbon emission monitoring log of the first carbon emission source, respectively perform indicator correlation and data credibility assessment on the plurality of first carbon emission monitoring indicators to determine the plurality of first indicator correlations and the plurality of first data credibility; The plurality of first carbon emission monitoring indicators are prioritized according to the plurality of first unit monitoring data volumes, the plurality of first indicator correlations and the plurality of first data credibility, a first monitoring indicator sequence is constructed, and a plurality of monitoring indicator sequences are obtained by sequential analysis.

3. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 2 is characterized in that: The steps performed by the priority evaluation module include: According to the carbon emission monitoring log of the first carbon emission source and the performance of the edge computing gateway, the data volume weight, correlation weight and credibility weight are configured based on the coefficient of variation method; After dimensionless processing of the plurality of first unit monitoring data volumes, the plurality of first indicator correlations, and the plurality of first data credibility, a plurality of first priorities are obtained by weighting based on the data volume weight, the correlation weight, and the credibility weight, wherein the priority is negatively correlated with the unit monitoring data volume and positively correlated with the indicator correlation and the data credibility; Based on the multiple first priorities, the multiple first carbon emission monitoring indicators are arranged from large to small according to the first priorities to obtain a first monitoring indicator sequence.

4. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 1 is characterized in that: The steps performed by the carbon emission source analysis module include: Randomly selecting a first carbon emission source, and obtaining a first carbon emission monitoring log of the first carbon emission source; Searching the first carbon emission monitoring log for the same period within a preset time range based on the preset time zone to obtain a plurality of sample first carbon emission monitoring sequences; The first predicted emission trend and the first emission fluctuation are obtained by analyzing the first carbon emission monitoring sequence of the multiple samples, and multiple predicted emission trends and multiple emission fluctuations of the multiple carbon emission sources are obtained by analyzing in sequence.

5. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 4 is characterized in that: The steps performed by the carbon emission source analysis module include: In a two-dimensional coordinate system, constructing a plurality of sample first carbon emission curves according to the plurality of sample first carbon emission monitoring sequences, and performing curve fitting to determine a first fitted carbon emission curve; A first increase ratio is calculated according to the first fitted carbon emission curve and is set as the first predicted emission trend; performing mean calculation on the plurality of sample first carbon emission monitoring sequences respectively to obtain the mean of the plurality of sample first carbon emission, and performing mean calculation again to obtain the first predicted carbon emission; An emission fluctuation analysis is performed based on the mean values ​​of the first carbon emissions of the multiple samples to calculate the first emission fluctuation, wherein the emission fluctuation is the ratio of the standard deviation of the mean values ​​of the first carbon emissions of the multiple samples to the first predicted carbon emissions.

6. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 5 is characterized in that: The steps performed by the carbon emission monitoring module include: Mapping and obtaining a first edge computing gateway of a first carbon emission source; Determining a first fluctuation scale based on the first predicted emission trend, the first predicted carbon emission amount, and the first emission fluctuation analysis; Adjusting the first initial data collection plan according to the first fluctuation scale to obtain a first optimized data collection plan; Analyze multiple edge computing gateways in turn to obtain multiple optimized data collection solutions and build an optimized data collection strategy.

7. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 1 is characterized in that: The steps performed by the carbon emission monitoring module include: Configuring a first initial data collection plan, wherein the first initial data collection plan includes the number of first monitoring indicators to be selected, the first data collection frequency, and the first key feature screening threshold; Adjusting the number of selected first monitoring indicators according to the first fluctuation scale to determine a first optimized number of selected indicators, and selecting the first optimized number of selected monitoring indicators from the first monitoring indicator sequence to obtain a first monitoring indicator type; Adjusting the first data collection frequency according to the first fluctuation scale to determine a first optimized collection frequency; Adjusting the first key feature screening threshold according to the first fluctuation scale to determine a first optimized feature screening threshold; A first optimized data collection plan is constructed based on the first monitoring indicator type, the first optimized collection frequency, and the first optimized feature screening threshold.

8. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 7 is characterized in that: The inverse of the first fluctuation scale is set as a first threshold adjustment coefficient, and the first key feature screening threshold is adjusted to obtain a first optimized feature screening threshold.

9. The edge computing gateway device for multi-modal energy carbon emission comprehensive monitoring according to claim 8, characterized in that: When carbon emissions monitoring is performed according to the first monitoring indicator type and the first optimized collection frequency, it is determined whether the monitoring data deviation under adjacent monitoring nodes is greater than the first optimized feature screening threshold. If it is greater, the monitoring data under the adjacent monitoring nodes are retained as key data. If it is less than, the monitoring data under the adjacent monitoring nodes are discarded, and judgment and screening are performed iteratively.

Images