A system and method for dynamic accounting of stationary point source pollutants and carbon emission inventory

Abstract

The present application relates to a kind of fixed point source pollutants and carbon emission inventory dynamic accounting system and method, comprising the following steps: the real-time emission of fixed point source pollutants monitoring index is collected, after inputting pre-trained emission-energy activity level correlation rule model, current energy activity level is obtained;Fixed point source pollutants and carbon emission inventory are accounted based on current energy activity level, and the accuracy of fixed point source pollutants and carbon emission inventory is verified based on the measured emission of monitoring index;Fixed point source pollutants and carbon emission inventory after verification are dynamically updated, shared and stored.The present application is widely applied to the technical field of atmospheric pollution source emission inventory, which well improves the limitation that point source activity level data is difficult to obtain in real time, provides a new method for high spatio-temporal resolution fixed point source emission inventory construction, opens up a new way of thinking, and has important scientific significance and practical application value.

Description

Technical Field

[0001] This invention belongs to the field of atmospheric pollution source emission inventory accounting technology, and particularly relates to a dynamic accounting system and method for fixed point source pollutants and carbon emission inventories. Background Technology

[0002] Accurate and real-time dynamic emission inventories are the foundation for identifying pollution sources and a core support for scientifically designing pollutant reduction pathways and making refined environmental management strategic decisions. Currently, my country has formulated and implemented numerous measures for stationary source emission reduction and established a Continuous Emission Monitoring System (CEMS) for key pollution sources, achieving significant results. Regarding emission inventories, multi-scale, localized pollutant and carbon emission inventories have been developed both domestically and internationally, with varying degrees of improvement in reliability, timeliness, and resolution. However, existing gridded or point-source emission inventories that only involve online monitoring indicators (SO2, NO2, PM) still suffer from incomplete emission species coverage in accurately analyzing the contributions of stationary source emissions to complex environmental pollution situations and their spatiotemporal evolution, becoming a major bottleneck restricting refined environmental management in my country.

[0003] From the perspective of influencing factors, accurate emissions accounting requires precise activity level data and reliable emission coefficients. Previous studies used fixed emission coefficients with regional averages, neglecting the impact of factors such as industrial restructuring, technological advancements, and stricter standards on emission coefficients, leading to significant uncertainty in the results. To meet the needs of refined management, researchers have gradually applied technologies such as CEMS data, big data, satellite inversion, and computer science to emissions inventory calculations, significantly improving inventory accuracy. However, these studies are still limited to emissions of monitorable pollutants such as SO2, NO2, and PM, lacking dynamic updates for unmonitored pollutants such as VOCs, CO, and CO2.

[0004] Meanwhile, due to the difficulty in obtaining real-time activity level data at the plant level, some researchers have decomposed activity level statistics with time lag characteristics based on empirical parameters to improve the time resolution of industrial source emission inventories, but this still cannot meet the timeliness requirements of the accounting results. Summary of the Invention

[0005] To address the aforementioned issues, the purpose of this invention is to provide a scientific, accurate, and real-time dynamic accounting system and method for fixed-source pollutants and carbon emission inventories. This system can improve the data deficiencies of multi-source environmental statistics and comprehensively and accurately update the fixed-source pollutants and carbon emission inventories dynamically, effectively supporting refined environmental management.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for dynamic accounting of stationary point source pollutants and carbon emission inventories, comprising the following steps:

[0008] Real-time emissions of pollutants from fixed point sources are collected and input into a pre-trained emission-energy activity level association rule model to obtain the predicted value of the current energy activity level.

[0009] The inventory of stationary source pollutants and carbon emissions is calculated based on the current energy activity level forecast, and the accuracy of the inventory of stationary source pollutants and carbon emissions is verified based on the measured emissions of online monitoring indicators.

[0010] The verified inventory of stationary source pollutants and carbon emissions is dynamically updated, shared, and stored.

[0011] Furthermore, the real-time emissions of pollutants from fixed point sources are collected and input into a pre-trained emission-energy activity level association rule model to obtain a predicted value of the current energy activity level, specifically as follows:

[0012] Define the base year for accounting, and based on the accounting objects, define the data requirements and data sources for accounting, so as to obtain the historical dataset required for accounting of stationary point source pollutants and carbon emissions;

[0013] The acquired historical dataset was preprocessed and used to construct a rule model linking emissions and energy activity levels.

[0014] Real-time emissions of pollutants from stationary source monitoring indicators are collected and processed to obtain a real-time dataset;

[0015] By inputting the real-time dataset into the trained emission-energy activity level association rule model, the predicted value of the current energy activity level can be obtained.

[0016] Furthermore, by clearly defining the base year for accounting, and based on the accounting object, the data requirements and data sources for accounting are determined, resulting in the historical dataset required for calculating the fixed-point source pollutant and carbon emission inventory, specifically:

[0017] Based on the available point source dataset that meets the preset time resolution requirements, the base year for accounting is determined;

[0018] Collect basic data and energy activity levels for the baseline year of accounting as point source datasets;

[0019] Collect a fixed-point source CEMS dataset at an hourly scale for the baseline year of accounting;

[0020] The point source dataset and the CEMS dataset are used together as the historical dataset.

[0021] Furthermore, the preprocessing of the acquired historical dataset is used to construct a rule model linking emissions and energy activity levels, specifically as follows:

[0022] Based on the accounting requirements, the time resolution for calculating the inventory of pollutants and carbon emissions from stationary point sources should be clearly defined;

[0023] Based on a defined time resolution, the acquired historical dataset is preprocessed to obtain a training sample set and a test sample set;

[0024] Machine learning methods were used to construct a correlation rule model between emissions and energy activity levels for monitoring indicators in the baseline year. The model was trained using a training sample set and tested using a test sample set.

[0025] Furthermore, based on a determined time resolution, the acquired historical dataset is preprocessed to obtain a training sample set and a test set, specifically as follows:

[0026] Remove abnormal statistical information;

[0027] Based on real-time indicators from online monitoring of stationary pollution sources, the actual operating status of enterprises can be identified.

[0028] Based on the actual operating status of the enterprise, after removing abnormal information, the data is organized into a historical dataset with the required time resolution.

[0029] The emission data of the monitoring indicators with numerical features are logarithmically converted, and the emission data of the monitoring indicators with discrete features are one-hot encoded to obtain the model input dataset, which is then divided into a model training sample set and a test sample set according to the proportion.

[0030] Furthermore, the time resolution includes both the actual resolution and the hourly resolution;

[0031] When processing the emissions of monitoring indicators at actual resolution, the emission data of monitoring indicators at the hourly scale of the collected baseline year are tiled to obtain the characteristic data of the emission of monitoring indicators.

[0032] When processing the emissions of monitored indicators at hourly resolution, the calculation formula is as follows:

[0033] E p,i,j,h =C p,i,j,h ×V p,i ×10 -6

[0034] Among them, E p,i,j,h C represents the hourly emission rate of pollutant j monitored at the i-th emission outlet of enterprise p, in kg / h. p,i,j,h The real-time monitoring concentration of pollutant j at the i-th emission outlet of enterprise p, in mg / m³.3 V p,i Let m be the flue gas flow rate at the i-th emission outlet of enterprise p. 3 / h.

[0035] Furthermore, the calculation of the stationary source pollutant and carbon emission inventory based on the predicted value of the current energy activity level, and the verification of the stationary source pollutant and carbon emission inventory based on the measured emissions of monitoring indicators, specifically involves:

[0036] Based on the current level of energy activity and the basic parameters of stationary sources, an inventory of pollutants and carbon emissions from stationary sources is calculated, including the calculated values ​​of emissions from monitored and non-monitored indicators.

[0037] Trend analysis is performed on the calculated values ​​of emissions based on real-time emissions of monitoring indicators to verify the inventory of stationary source pollutants and carbon emissions.

[0038] Secondly, the present invention provides a dynamic accounting system for stationary point source pollutants and carbon emission inventories, comprising:

[0039] The energy activity level calculation module is used to collect real-time emissions of pollutants from fixed point sources. After inputting the data into a pre-trained emission-energy activity level association rule model, the module obtains the predicted value of the current energy activity level.

[0040] The inventory calculation and verification module is used to calculate the fixed-source pollutant and carbon emission inventory based on the predicted value of the current energy activity level, and to verify the fixed-source pollutant and carbon emission inventory based on the measured emission of monitoring indicators.

[0041] The inventory output module is used to dynamically update, share, and store the verified inventory of fixed point source pollutants and carbon emissions.

[0042] Thirdly, the present invention provides a computer-readable storage medium for storing one or more programs, said one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any of the methods.

[0043] Fourthly, the present invention provides a computing device comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods.

[0044] The present invention has the following advantages due to the adoption of the above technical solutions:

[0045] (1) The SO2, NO2 and PM emissions involved in this invention are obtained by online monitoring of different emission outlets of fixed point sources. The other input parameters are all historical dynamic data of the same period, supplemented by static parameters such as basic information of point sources. Based on historical rules and real-time online monitoring data, the calculation can quickly, accurately and effectively enhance the timeliness of the inventory, and is simple and easy to operate.

[0046] (2) The present invention uses machine learning algorithms to realize the correlation between the pollutant emission level and the fuel consumption activity level in the baseline year, and inverts the real-time fuel consumption activity level of fixed point sources, solving the problem that it is difficult to obtain high-time-resolution fuel consumption activity level data in real time during the inventory compilation process.

[0047] (3) In this invention, the machine learning algorithm performs feature tiling on the hourly pollutant emissions monitored online during modeling, which can reflect the actual emission status of point sources during operation.

[0048] (4) The present invention establishes a dynamic inventory of fixed-point source pollutants and carbon emissions based on online monitoring and machine learning algorithms, which is more timely than existing emission inventories. The method and research ideas of the present invention are universal in the compilation of dynamic inventories with high spatiotemporal resolution for other sources such as traffic sources, and have important scientific significance and practical application value.

[0049] Therefore, this invention can be widely applied in the field of atmospheric pollution source emission inventory technology. Attached Figure Description

[0050] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. In the drawings:

[0051] Figure 1 A flowchart of the dynamic accounting method for fixed-point source pollutants and carbon emission inventories provided in this embodiment of the invention;

[0052] Figure 2 This is a structural diagram of a dynamic accounting system for fixed-point source pollutants and carbon emission inventories provided in an embodiment of the present invention.

[0053] Figure 3 This is a comparison chart of predicted and actual daily-scale energy activity levels in an embodiment of the present invention.

[0054] Figure 4 This is a comparison chart of the predicted and actual values ​​of hourly-scale energy activity levels in an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0056] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0057] In some embodiments of this invention, a dynamic accounting method for fixed-source pollutants and carbon emission inventories is disclosed. First, based on measured pollutant emissions and high-resolution fixed-source activity level data, a machine learning algorithm is used to establish emission-energy activity level correlation rules for a baseline year, i.e., correlation rules between the emissions of monitored indicators such as SO2, NO2, and PM and the corresponding energy activity levels. Then, based on the emission-energy activity level correlation rules for the baseline year, the energy activity level for the corresponding period is inverted by real-time collected and acquired emissions of monitored indicators such as SO2, NO2, and PM. The real-time emissions of SO2, NO2, PM, VOCs, CO, and CO2 are further calculated, and the accuracy of the calculation results is ensured by comparing the real-time monitored SO2, NO2, and PM emissions with their corresponding calculated emissions. Ultimately, this achieves real-time dynamic updating of the fixed-source pollutant and carbon emission inventories, providing a new method and opening up new avenues for constructing high spatiotemporal resolution fixed-source emission inventories. This method has significant scientific value and important practical application value.

[0058] Correspondingly, in other embodiments of the present invention, a dynamic accounting system, device, and storage medium for fixed point source pollutants and carbon emission inventories are provided.

[0059] Example 1

[0060] like Figure 1 As shown, this embodiment provides a method for dynamic accounting of stationary point source pollutants and carbon emission inventories, specifically including the following steps:

[0061] (1) The real-time emission of pollutants from fixed point sources is collected and input into a pre-built emission-energy activity level correlation rule model to obtain the current energy activity level;

[0062] (2) The fixed-point source pollutant and carbon emission inventory is calculated based on the current energy activity level, and the fixed-point source pollutant and carbon emission inventory is verified based on the measured emission of monitoring indicators; wherein, the fixed-point source pollutant and carbon emission inventory refers to the dataset of pollutants and CO2 emitted into the atmosphere by fixed-point sources within a specific period, and the pollutants include monitoring indicators and non-monitoring indicators.

[0063] (3) The verified fixed-source pollutant and carbon emission inventory is updated quickly and dynamically, and the data is shared through an online technology platform. At the same time, the updated fixed-source pollutant and carbon emission inventory dataset is stored for the purpose of optimizing fixed-source pollutant emission reduction measures.

[0064] Preferably, step (1) above specifically includes:

[0065] (1.1) Define the base year for accounting, and define the data requirements and data sources for accounting based on the accounting objects to obtain the historical dataset required for accounting of fixed-point source pollutants and carbon emissions; among them, the fixed-point source pollutants and carbon emissions inventory refers to the dataset of pollutants and CO2 emitted into the atmosphere by fixed-point sources within a specific period, and its time resolution mainly depends on the time resolution of the energy activity level data used in modeling.

[0066] (1.2) After preprocessing the acquired historical dataset, it is used to construct an emission-energy activity level association rule model.

[0067] (1.3) Calculate the real-time emissions of pollutant monitoring indicators from fixed point sources to obtain a real-time dataset.

[0068] (1.4) Input the real-time dataset into the constructed emission-energy activity level association rule model to obtain the current energy activity level.

[0069] Preferably, the method for obtaining the historical dataset in step (1.1) above is as follows:

[0070] (1.1.1) Based on the available point source dataset that meets the preset time resolution requirements, determine the base year for accounting. The time resolution is based on days or hours. In this embodiment, the time resolution of energy consumption activity level data is based on days.

[0071] (1.1.2) Collect basic data and energy activity levels for the base year of accounting as point source datasets.

[0072] The basic data includes geographic information data of fixed point sources, commissioning time, industry category, etc.; energy activity level data includes production facility scale, raw material consumption, fuel type, fuel consumption, lower heating value, product output, boiler type, number of boilers, boiler installed capacity and other parameters, pollution control process and type, number of pollution control equipment, operation time of pollution control equipment, chimney height, power plant unit parameters (unit capacity, turbine type, pressure parameters, number), power plant unit commissioning date, etc.

[0073] (1.1.3) Collect fixed-point source CEMS datasets for the base year of accounting. The CEMS data is on an hourly scale and mainly includes: SO2 concentration, NO2 concentration, dust concentration, flue gas flow rate, flue gas temperature, flue gas humidity, flue gas oxygen content, equipment operating status, etc.

[0074] (1.1.4) The point source dataset and the CEMS dataset are used together as the historical dataset.

[0075] Preferably, in step (1.2) above, the acquired historical dataset is preprocessed and then used to construct an emission-energy activity level association rule model, specifically as follows:

[0076] (1.2.1) Based on the accounting requirements, clarify the time resolution for the accounting of stationary point source pollutants and carbon emission inventories;

[0077] (1.2.2) Based on the determined time resolution, the acquired historical dataset is preprocessed to obtain the model input sample set;

[0078] (1.2.3) Machine learning methods are used to quickly construct a correlation rule model between emissions of monitoring indicators (SO2, NO2, PM) and energy activity levels in the baseline year, and the rationality and applicability of the model are evaluated.

[0079] Preferably, in step (1.2.1) above, this embodiment uses two methods to determine the time resolution of the accounting list, namely, the method of determining the time resolution based on the actual statistical energy activity level data and the method of determining the hour resolution.

[0080] Preferably, in step (1.2.2) above, the acquired historical dataset is preprocessed, specifically as follows:

[0081] ① Combining theory and practical experience, abnormal statistical information is eliminated: For online monitoring data of emission outlets, missing or zero values ​​that last for more than 24 hours are supplemented with the monthly average value; when the missing or zero value is less than 24 hours, it is supplemented with the arithmetic mean of the valid data at the time nodes before and after the missing or zero value period; abnormal values ​​are treated as missing values.

[0082] ② Based on real-time online monitoring indicators of stationary pollution sources, identify the enterprise's operating status: normal, shutdown / shutdown, malfunction, maintenance, etc. Specifically, identifying the enterprise's actual operating status is used to obtain daily emissions or hourly resolution energy activity level decompositions that relatively accurately reflect actual operating conditions. For abnormal periods, emissions of non-monitoring indicators (VOCs, CO, CO2) are not calculated; therefore, the daily emissions are the sum of pollutants during normal operation. Hourly resolution energy activity levels can be decomposed on an hourly scale based on daily statistical values ​​and the enterprise's actual operating time.

[0083] ③ Based on the actual operating status of the enterprise and the data after removing abnormal information, organize it into a historical dataset with the required time resolution.

[0084] Actual resolution: The emission data of SO2, NO2, PM and other monitoring indicators collected at the hourly scale of the baseline year are tiled with features. That is, the relevant indicators (such as monitoring concentration and flue gas parameters) of each pollutant (SO2, NO2, PM) monitored online on the day are used as 24 feature inputs. Thus, a total of 24*3=72 features are generated for the three pollutants. Combined with actual statistical data, the association rules between the energy activity level data and the emission of monitoring indicators at the highest resolution (such as day) of actual statistics are trained.

[0085] Hourly resolution: For higher resolution requirements (e.g., hourly), the actual statistical energy activity level is decomposed into hourly scales for training association rules with the corresponding hourly-scale emissions of monitoring indicators such as SO2, NO2, and PM. The specific calculation formula for the hourly-resolution monitoring indicators (e.g., SO2, NO2, PM) emissions is as follows:

[0086] E p,i,j,h =C p,i,j,h ×V p,i ×10 -6 (1)

[0087] Among them, E p,i,j,h C represents the hourly emission rate of pollutant j monitored at the i-th emission outlet of enterprise p, in kg / h. p,i,j,h The real-time monitoring concentration of pollutant j at the i-th emission outlet of enterprise p, in mg / m³. 3 V p,i Let m be the flue gas flow rate at the i-th emission outlet of enterprise p. 3 / h.

[0088] ④ Numerical features are logarithmically converted, and discrete features are one-hot encoded. Date variables are preprocessed into months of the year, days of the week, and hours of the day, and treated as discrete features to explore the factors for daily emission weights and intra-year time parameters. Numerical features mainly include flue gas flow rate, SO2 monitoring concentration, NO2 monitoring concentration, particulate matter monitoring concentration, flue gas temperature, flue gas humidity, flue gas oxygen content, boiler parameters (such as rated evaporation capacity and lower heating value), unit capacity, and daily coal consumption. Discrete features mainly include production equipment type, fuel type, unit parameters (such as turbine type, pressure parameters, and number), and cooling method.

[0089] Preferably, in step (1.2.3) above, the actual statistical level of fuel consumption activity or the energy consumption activity level decomposed by time module is used as the output parameter, and the historical dataset, i.e., numerical features, discrete features, time parameters, and online monitoring pollutant emission data, is used as the input parameters. A machine learning algorithm is used to train the association rules between the emissions of monitoring indicators (e.g., SO2, NO2, PM) and the energy activity level in the baseline year. The importance and weight of each input feature parameter in the modeling process are analyzed. The basic principle is to consider the importance of features, the ease of obtaining their data, and the highest consistency between the inversion results and the real data. The most suitable association rule model between the emissions of monitoring indicators and the energy activity level is then selected and identified.

[0090] Preferably, in step (1.3) above, the real-time emission data of the monitoring indicators of fixed-point source pollutants are used to obtain a real-time dataset, specifically as follows:

[0091] (1.3.1) Real-time acquisition of hourly CEMS data from fixed point sources, mainly including: SO2 concentration, NO2 concentration, dust concentration, flue gas flow rate, flue gas temperature, flue gas humidity, flue gas oxygen content, equipment operating status, etc., to form a real-time dataset;

[0092] (1.3.2) Based on the hourly CEMS data of fixed point sources collected in real time, the hourly resolution emission of monitoring indicators, such as SO2, NO2 and PM emissions, is calculated.

[0093] (1.3.3) The hourly resolution emission of the monitoring indicators is logarithmically processed and used as the input characteristic parameter for real-time inventory accounting.

[0094] Preferably, step (2) above specifically includes:

[0095] (2.1) Based on the current energy activity level and the basic parameters of fixed point sources, the calculated values ​​of the monitoring indicators and non-monitoring indicators of fixed point sources are obtained.

[0096] Based on the current level of energy activity, the real-time emissions of indicators such as SO2, NO2, PM, VOCs, CO, and CO2 are calculated using the following formulas:

[0097] E p,i,j =EF j ×A p,i ×(1-η j (2)

[0098] Among them, E p,i,j A represents the emission amount of pollutant j that was not monitored at the i-th emission outlet of point source p; p,i EF represents the fuel consumption or energy activity level of the equipment corresponding to the i-th emission outlet of point source p; j For pollutant emission factors that were not monitored; η j The removal efficiency of pollution control technology for pollutant j.

[0099] (2.2) Conduct trend analysis on the measured emissions and the calculated emissions of the monitoring indicators to verify the accuracy of the pollutant and carbon emission inventory dynamically calculated based on the inversion of energy activity levels.

[0100] (2.3) Conduct a detailed literature review, collect fixed point source emission inventories from existing studies, and make a horizontal comparison with the emission inventory of this invention; perform trend verification with the emission inventory of this invention through trend analysis of regional air quality monitoring data; and use the Monte Carlo method to quantitatively assess the uncertainty of the pollutant and carbon emission inventories of this invention.

[0101] Example 2

[0102] This invention has been successfully applied to the dynamic accounting of pollutants and carbon emissions in a steel plant, employing a lightweight gradient boosting machine (LightGBM) algorithm for modeling. Based on the accounting object and requirements, the time resolution of the accounting is clearly defined; fuel consumption activity level is used as the output parameter, and other data as input parameters, with 80% of the sample data used as the training set and 20% as the test set. Basic information data, activity level data, and CEMS data are processed using logarithmic or one-hot encoding according to continuous and discrete characteristics, respectively; online monitoring data uses hourly feature tiling and is logarithmically processed; discrete features (including fuel type, production equipment type, boiler type, pollution control process and type, etc.) are one-hot encoded; continuous features (including fuel consumption, lower heating value, product output, etc.) are logarithmically processed. Additionally, weekday and month variables are added to explore the factors for daily emission weights and annual time parameters. Based on the modeling results, the importance of each input feature parameter in the modeling process is clarified. Taking into account important features, whose data are easy to obtain, and whose inversion results have the highest consistency with real data as the basic principles, the most suitable correlation rules between SO2, NO2, PM emissions and fuel consumption are selected and clarified. At the same time, the models are compared with the real values ​​to evaluate their rationality and applicability.

[0103] Example 3

[0104] This embodiment takes the dynamic accounting of pollutant and CO2 emission inventories of a power plant from March 2018 to January 2022 as an example. For the CEMS data in the power plant's historical dataset, data and abnormal data under power plant shutdown, fault, maintenance and other states are removed, and finally 33,720 valid samples are obtained. Null values ​​or zero values ​​that last for more than 24 hours are supplemented with the monthly average value; when the null value or zero value is less than 24 hours, it is supplemented with the arithmetic mean of the valid data before and after the time node of the null value or zero value period; abnormal values ​​are treated as null values; the hourly resolution NO2, SO2 and PM emissions are calculated based on the flue gas flow and NO2, SO2 and particulate matter emission concentrations monitored online.

[0105] For the actual resolution, the acquired daily-scale fuel consumption activity level data is used as the output, and the hourly-scale NO2, SO2, and PM emissions are flattened and used as input parameters. The LightGBM algorithm is used for modeling, prioritizing important features, readily available data, and the highest consistency between the inversion results and real data. The most suitable hourly resolution correlation rules between SO2, NO2, and PM emissions and fuel consumption are selected and clarified. The example test set uses comparative analysis between model output values ​​and actual values ​​(e.g., ...). Figure 3As shown), the standardized dimensionless statistics were obtained as follows: standard mean deviation (NMB, -0.035), standard mean error (NME, 0.073), mean fractional deviation (MFB, -0.028), mean fractional error (MFE, 0.075), and coefficient of determination (R²). 2 The indicators such as 0.87 are all within the reasonable range of model evaluation, further demonstrating that the established model is reasonable and applicable, and the modeling results are reliable.

[0106] For time resolution, daily-scale fuel consumption activity data is decomposed into hourly values. The decomposition coefficients are referenced from the "Technical Manual for Compiling Urban Air Pollutant Emission Inventories." The decomposed hourly-resolution fuel consumption activity data is logarithmically processed and used as output parameters. Based on the principles of considering important features, readily available data, and the highest consistency between the inversion results and real data, the most suitable hourly-resolution correlation rules between SO2, NO2, and PM emissions and fuel consumption are selected and clarified. The analysis is then based on a comparison between the predicted values ​​and actual values ​​output by the model on the test set. Figure 4 The standardized dimensionless statistics were obtained as follows: standard mean deviation (NMB, -0.004), standard mean error (NME, 0.080), mean fractional deviation (MFB, 0.004), mean fractional error (MFE, 0.081), and coefficient of determination (R²). 2 The indicators such as 0.86 are all within the reasonable range of model evaluation, further demonstrating that the established model is reasonable and applicable, and the modeling results are reliable.

[0107] The calculation results are organized to form a pollutant and carbon emission inventory based on the actual fuel consumption activity level at a time resolution (the corresponding time resolution in this embodiment is days). This embodiment also decomposes the daily-scale fuel consumption activity level into hourly levels to quickly calculate the hourly-resolution pollutant and carbon emission inventory of the power plant. Both inventories can be shared through an online technology platform to meet the different needs of governments, enterprises, and the public.

[0108] Example 4

[0109] The above-described embodiment 1 provides a method for dynamic accounting of stationary source pollutants and carbon emission inventories. Correspondingly, this embodiment provides a system for dynamic accounting of stationary source pollutants and carbon emission inventories. The system provided in this embodiment can implement the dynamic accounting method for stationary source pollutants and carbon emission inventories of embodiment 1. This system can be implemented through software, hardware, or a combination of both. For example, the system may include integrated or separate functional modules or units to execute the corresponding steps in the methods of embodiment 1. Since the system in this embodiment is basically similar to the method embodiment, the description process in this embodiment is relatively simple. Relevant details can be found in the description of embodiment 1. The system embodiment provided in this embodiment is merely illustrative.

[0110] The dynamic accounting system for fixed-point source pollutants and carbon emission inventories provided in this embodiment includes:

[0111] The energy activity level calculation module is used to collect real-time emissions of pollutants from fixed point sources. After inputting the data into a pre-built emission-energy activity level correlation rule model, the current energy activity level is obtained.

[0112] The inventory calculation and verification module is used to calculate the inventory of stationary source pollutants and carbon emissions based on the current energy activity level, and to verify the accuracy of the inventory of stationary source pollutants and carbon emissions based on the measured emissions of monitoring indicators.

[0113] The inventory output module is used to dynamically update, share, and store the verified inventory of fixed point source pollutants and carbon emissions.

[0114] Preferably, the inventory calculation and accounting module includes:

[0115] The historical data acquisition module is used to acquire historical datasets and point source datasets and perform preprocessing.

[0116] The real-time data acquisition module is used to acquire hourly CEMS data (i.e., real-time datasets) from fixed point sources in real time and to perform data preprocessing.

[0117] The model training module is used to train the emission-energy activity level association rule model using preprocessed historical datasets and point source datasets;

[0118] The calculation module calculates the real-time emissions of indicators such as SO2, NO2, PM, VOCs, CO, and CO2 based on real-time collected CEMS data and trained emission-energy activity level association rules, while also combining fixed point source information data.

[0119] Example 5

[0120] This embodiment provides a processing device corresponding to the dynamic accounting method for fixed-point source pollutants and carbon emission inventories provided in Embodiment 1. The processing device can be a processing device for a client, such as a mobile phone, laptop, tablet computer, desktop computer, etc., to execute the method of Embodiment 1.

[0121] The processing device includes a processor, a memory, a communication interface, and a bus. The processor, memory, and communication interface are connected via the bus to communicate with each other. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it executes the dynamic accounting method for fixed-point source pollutants and carbon emission inventories provided in Embodiment 1.

[0122] In some embodiments, the memory may be high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device.

[0123] In other embodiments, the processor can be a general-purpose processor of various types, such as a central processing unit (CPU) or a digital signal processor (DSP), and is not limited thereto.

[0124] Example 6

[0125] The dynamic accounting method for fixed-source pollutants and carbon emission inventories in Embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the dynamic accounting method for fixed-source pollutants and carbon emission inventories described in Embodiment 1 are loaded.

[0126] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.

[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for dynamic accounting of stationary point source pollutants and carbon emission inventories, characterized in that, Includes the following steps: Real-time emissions of pollutants from fixed point sources are collected and input into a pre-built emission-energy activity level correlation rule model to obtain the predicted value of the current energy activity level. The inventory of pollutants and carbon emissions from stationary sources is calculated based on the current energy activity level forecast, and the accuracy of the inventory of pollutants and carbon emissions from stationary sources is verified based on the measured emissions of monitoring indicators. The verified inventory of stationary source pollutants and carbon emissions is dynamically updated, shared, and stored.

2. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 1, characterized in that, The real-time emissions of pollutants from fixed point sources are collected and input into a pre-constructed emission-energy activity level correlation rule model to obtain a predicted value of the current energy activity level, specifically as follows: Define the base year for accounting, and based on the accounting objects, define the data requirements and data sources for accounting, so as to obtain the historical dataset required for accounting of fixed point source pollutants and carbon emissions; The acquired historical dataset was preprocessed and used to construct a rule model linking emissions and energy activity levels. Real-time emissions of pollutants from stationary source monitoring indicators are collected and processed to obtain a real-time dataset; By inputting the real-time dataset into the trained emission-energy activity level association rule model, the predicted value of the current energy activity level can be obtained.

3. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 2, characterized in that, The specified base year for accounting is determined, and the data requirements and data sources for accounting are defined based on the accounting objects. This yields the historical dataset required for calculating the inventory of stationary point source pollutants and carbon emissions, specifically: Based on the available point source dataset that meets the preset time resolution requirements, the base year for accounting is determined; Collect basic data and energy activity levels for the baseline year of accounting as point source datasets; Collect a fixed-point CEMS dataset at an hourly scale for the baseline year of accounting; The point source dataset and the CEMS dataset are used together as the historical dataset.

4. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 2, characterized in that, The preprocessing of the acquired historical dataset is used to construct a rule model linking emissions and energy activity levels, specifically as follows: Based on the accounting requirements, the time resolution for calculating the inventory of pollutants and carbon emissions from stationary point sources should be clearly defined; Based on a defined time resolution, the acquired historical dataset is preprocessed to obtain the model training sample set and the test sample set; Machine learning methods were used to construct a rule model for the correlation between emissions and energy activity levels of monitoring indicators in the baseline year. The model was trained using a training sample set and tested using a test sample set.

5. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 4, characterized in that, The process of preprocessing the acquired historical dataset based on a defined time resolution to obtain the model training sample set and the test sample set is as follows: Remove abnormal statistical information; Based on real-time indicators from online monitoring of stationary pollution sources, the actual operating status of enterprises can be identified. Based on the actual operating status of the enterprise and the data after removing abnormal information, a historical dataset with the required time resolution was compiled. The emission data of the monitoring indicators with numerical features are logarithmically converted, and the emission data of the monitoring indicators with discrete features are one-hot encoded to obtain the model input dataset.

6. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 5, characterized in that, The time resolution includes both the actual resolution and the hourly resolution; When processing the emissions of monitoring indicators at actual resolution, the emission data of monitoring indicators at the hourly scale of the collected baseline year are tiled to obtain the characteristic data of the emission of monitoring indicators. When processing the emissions of monitored indicators at hourly resolution, the calculation formula is as follows: AND p，i,j，h =C p，i，j,h ×V p，i ×10 -6 Among them, E p,i,j,h C represents the hourly emission rate of pollutant j monitored at the i-th emission outlet of enterprise p, in kg / h. p,i,j,h The real-time monitoring concentration of pollutant j at the i-th emission outlet of enterprise p, in mg / m³. 3 ; V p,i Let m be the flue gas flow rate at the i-th emission outlet of enterprise p. 3 / h.

7. The method for dynamic accounting of stationary point source pollutants and carbon emission inventories as described in claim 5, characterized in that, The process involves calculating the stationary source pollutant and carbon emission inventory based on the predicted value of current energy activity levels, and verifying the stationary source pollutant and carbon emission inventory based on the measured emissions of monitoring indicators. Specifically: Based on the current level of energy activity and the basic parameters of stationary sources, the inventory of pollutants and carbon emissions from stationary sources is calculated, including the calculated values ​​of emissions from monitored and non-monitored indicators. Trend analysis is performed on the calculated values ​​of emissions from monitoring indicators based on the measured emissions of these indicators, thereby verifying the accuracy of the stationary source pollutant and carbon emission inventory.

8. A dynamic accounting system for stationary point source pollutants and carbon emission inventories, characterized in that: The energy activity level calculation module is used to collect real-time emissions of pollutants from fixed point sources. After inputting the pre-built emission-energy activity level correlation rule model, the current energy activity level prediction value is obtained. The inventory calculation and verification module is used to calculate the inventory of stationary source pollutants and carbon emissions based on the predicted value of the current energy activity level, and to verify the accuracy of the inventory of stationary source pollutants and carbon emissions based on the measured emissions of monitoring indicators. The inventory output module is used to dynamically update, share, and store the verified inventory of fixed point source pollutants and carbon emissions.

9. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 7.

10. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 7.

Images