An online analysis method of carbon emissions based on multi-source metering data

Abstract

The present invention discloses an online analysis method for carbon emissions based on multi-source metering data. By establishing a mathematical model for multi-source data emissions of emission units, data rationality verification is carried out between related monitoring points of similar emission sources and different types of emission data sources, and a mathematical model for reasonable data judgment is established. Classification and aggregation are performed according to the nature dimensions of the targets involved in carbon emission measurement, including the data relationship between monitoring point data of the same type and the data relationship between different types of emission sources. The present invention provides an online analysis method for carbon emissions based on multi-source metering data. Through the relationship between monitoring point data and probability methodology, different data sources are verified, abnormal data and disconnected data are fitted and calculated, and corrected. Combined with the actual production of the enterprise, the metering data model is supplemented and improved, so that within a certain period of time, the data of the monitoring system truly reflects the carbon emissions of the enterprise.

Description

Technical Field

[0001] The present invention relates to the field of carbon neutrality technology, and in particular to an online analysis method for carbon emissions based on multi-source metering data. Background Art

[0002] Enterprises need to upgrade their industries and products, carry out energy-saving technology transformation, use clean energy, etc., in order to effectively reduce the carbon dioxide emissions of industrial enterprises.

[0003] However, the current carbon dioxide emissions of industrial enterprises are mainly calculated through accounting methods (generally for one year), and some industrial enterprises have built carbon emission information systems, but the monitoring of their own emission sources is incomplete. During the data collection process, secondary instruments are replaced, secondary instruments fail, some emission source data communication lines have abnormal data or data disconnection due to signal interference, or production structure adjustments lead to abnormal online analysis data of carbon emissions, which affects industrial enterprises' energy conservation and carbon reduction forecasting and early warning, data accounting, etc. The existing technology lacks a technical solution to this problem. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide an online analysis method for carbon emissions based on multi-source metering data. By establishing a mathematical model of multi-source data emissions of emission units, the abnormal data and disconnected data are fitted, calculated and corrected through the relationship between metering point data and combined with probability methodology.

[0005] The present invention is achieved in that:

[0006] An online carbon emissions analysis method based on multi-source metering data is used to analyze the carbon emissions of at least one emission unit, each emission unit includes multiple types of emission sources, and each type of emission source includes one or more monitoring points. The method includes:

[0007] Step 10: learning and establishing a mathematical model based on the emission source historical data of the emission unit, wherein the mathematical model includes an emission calculation model for the emission unit, a historical data relationship model for different types of emission sources, a calculation model for different monitoring points of the same emission source type, and a mathematical relationship model for activity levels between different monitoring points of the same emission source type;

[0008] Step 20, inputting the metrological performance of the metrological equipment for measuring each data source, wherein the metrological performance includes the measurement standard deviation σ of the metrological equipment;

[0009] Step 30: According to the historical data relationship model of different types of emission sources, the existing emission source data is used as an independent variable, and the emission source that needs to be predicted or verified is used as a dependent variable to obtain a predicted value or a verified value; according to the mathematical relationship model of activity levels between different monitoring points of the same emission source type, the emission amount of the metered monitoring point that has been obtained is determined as an independent variable, and the emission data of the monitoring point that needs to be predicted or verified is used as a dependent variable to obtain a predicted value or a verified value;

[0010] Data verification process: perform data fitting verification according to the set data verification time limit and verification frequency; determine the independent variable based on the emission source data of the acquired data and the accuracy of the metering equipment, and then calculate the verification value of the dependent variable based on the historical data relationship model of different types of emission sources or the mathematical relationship model of activity levels between different monitoring points of the same emission source type. When the verification value is within the deviation threshold range, it is determined that the verification value is consistent with the actual measurement value. If it exceeds the deviation threshold, feedback is generated to generate warning information;

[0011] Data prediction process: When disconnected data or abnormal data is generated, if it is data at the emission source level, the predicted value is calculated based on the historical data relationship model of different types of emission sources and the existing emission source data; if it is data at the monitoring point level, the predicted value is calculated based on the mathematical relationship model of activity levels between different monitoring points of the same emission source type and the existing monitoring point data, and the uncertainty of the predicted value is given according to the model;

[0012] Step 40: Analyze the warning information recorded in the data verification process. If it is a problem in the data collection or transmission process, perform data fitting and correction according to the correction model, and mark the deviation uncertainty in the data fitting. If it is a change in production status, it is necessary to analyze and confirm according to the actual production status adjustment;

[0013] Step 50: Calculate the carbon emissions of the industrial enterprise based on the measured values ​​of the monitoring points and the prediction results when the measured values ​​of the monitoring points are missing, and provide relevant information on the prediction and early warning.

[0014] Furthermore, the mathematical model specifically includes:

[0015] The calculation model of the emission unit emission is: ;

[0016] Among them, E 总 represents the total emissions per emission unit, Indicates i Types of emission source data, n is the number of emission source types of the emission unit, including direct emissions, indirect emissions and energy indirect emissions;

[0017] The historical data relationship model of different types of emission sources is: ;

[0018] Among them, f(E yi ) is the data relationship function between different types of emission sources obtained by historical data analysis;

[0019] The calculation model for different monitoring points of the same emission source type is: ; E dj =AD j ×EF j；

[0020] Among them, AD j represents the activity level data of the jth monitoring point under the same emission source, EF j represents the emission factor of the jth monitoring point;

[0021] The mathematical relationship model of activity levels between different monitoring points of the same emission source type is: ;

[0022] Among them, f( AD j ) is a mathematical relationship function for obtaining the activity levels of different monitoring points of the same emission source type based on historical data.

[0023] Furthermore, in step 30, the measuring device with high accuracy is determined as the independent variable, and the measuring device with low accuracy is determined as the dependent variable.

[0024] Furthermore, in step 30, the deviation threshold is ±3σ.

[0025] The present invention has the following advantages:

[0026] By establishing a mathematical model for measuring carbon emission multi-source data, the data of different emission sources or the data between different monitoring points of the same emission source type are verified according to metrological methods, and predictions are made through metrological methods. For enterprises with data interruption and data anomalies, the measurement data model is supplemented and improved through mathematical model correction and combined with the actual production of the enterprise, so that within a certain period of time, the data of the monitoring system truly reflects the carbon emissions of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0028] Figure 1 1 is a flowchart of the execution of the method according to the embodiment of the present invention. DETAILED DESCRIPTION

[0029] The method of the present invention establishes a mathematical model for multi-source data emissions of emission units, conducts data rationality verification between related monitoring points of similar emission sources and different types of emission data sources, and establishes a mathematical model for reasonable data judgment. Classification and aggregation are performed according to the nature dimensions of the targets involved in carbon emission measurement, including the data relationship between measurement points of the same type of emission sources and the data relationship between different types of emission sources. Through the relationship between monitoring point data combined with probability methodology, different data sources are verified, abnormal data and disconnected data are fitted and calculated and corrected, and combined with the actual production of the enterprise, the measurement data model is supplemented and improved, so that within a certain period of time, the data of the monitoring system truly reflects the carbon emissions of the enterprise.

[0030] The general idea of ​​the present invention is as follows:

[0031] Step 1: Emission monitoring data integrity analysis. Learn and establish a mathematical model based on the historical data of emission sources of emission units. The integrity model includes whether the boundary covers all emission source types that are not less than 1%.

[0032] 1. For those that can be monitored online technically, all of them should be monitored;

[0033] 2. For emission sources that cannot be monitored online technically, their emission data should account for no more than 1% of the total emissions.

[0034] 3. Calculation model of emission unit based on historical data:

[0035] Model 1: Emission calculation model for emission units ;

[0036] Among them, E 总 represents the total emissions per emission unit, E yi Indicates i The data are emission source data, where n is the number of emission source types of the emission unit, including direct emissions, indirect emissions, and energy indirect emissions.

[0037] 4. The mathematical relationship model between different emission sources is obtained by analyzing historical data:

[0038] Model 2: Relationship model of historical data of different emission sources ;

[0039] Among them, f(E yi ) is the data relationship function between different emission sources obtained by historical data analysis.

[0040] Step 2: Establish a data model between different monitoring points of the same emission source type. Correction and prediction of emission monitoring point level data, and fit the data using activity data.

[0041] 1. Carbon emission calculation model for different monitoring points of the same emission source type

[0042] Model 3: Calculation model of the same emission source with different monitoring points ;

[0043] E dj =AD j ×EF j；

[0044] Among them, AD j represents the activity level data of the jth monitoring point under the same emission source, EF j represents the emission factor of the jth monitoring point.

[0045] 2. Mathematical relationship model of activity levels of different monitoring points of the same emission source type obtained from historical data

[0046] Model 4: Mathematical relationship model of activity levels between different monitoring points of the same emission source type ;

[0047] Among them, f( AD j ) is a mathematical relationship function for obtaining the activity levels of different monitoring points of the same emission source type based on historical data.

[0048] Step 3: Input the metrological performance of the monitoring equipment related to the metering data of each emission source. According to the type of carbon emission-related data source of the emission unit, determine the metrological performance of the relevant metering equipment, the standard deviation of measurement uncertainty, etc.

[0049] Step 4: Determine the data prediction value and verification value. According to Model 2, the existing emission source data is used as the independent variable, and the emission source to be predicted or verified is used as the dependent variable to obtain the prediction or verification value. E yi According to Model 4, the emission data of the metered monitoring points that have been obtained are determined as the independent variable, and the emission data of the monitoring points that need to be predicted or verified are used as the dependent variable to obtain the prediction or verification data of different monitoring points of the same emission source type. AD jn .

[0050] Step 5: Data prediction and verification.

[0051] A. When data verification is required, set the data verification time limit and verification frequency, and perform data fitting verification.

[0052] 1. Data verification of different types of emission sources. First, set the accuracy level of the emission source metering equipment of Model 2 and the acquired data, calculate the dependent variable based on the independent variable, and according to the normal distribution function, take the probability of the metering equipment exceeding the uncertainty confidence interval less than 1% as the deviation of the measurement result, that is, the measurement result E yi ±3σ is substituted into Model 2, and the calculated result is used as the threshold judgment of the verification result of the evaluation dependent variable. If the result is within the threshold range, it is judged that the verification value is consistent with the actual measurement value result. If it exceeds the threshold, feedback is generated to generate warning information.

[0053] 2. Data verification of different monitoring points of the same emission source type. First, set the mathematical relationship model 4 and the accuracy level of the measuring equipment, calculate the dependent variable based on the independent variable, and according to the normal distribution function, take the probability of the measuring equipment exceeding the uncertainty confidence interval less than 1% as the deviation of the measurement result, that is, the measurement result AD jn-1 ±3σ is substituted into mathematical relationship model 4, and the calculated result is used as the threshold judgment of the prediction result of the evaluation dependent variable. If the result is within the threshold range, it is judged that the verification value is consistent with the actual measurement value result. If it exceeds the threshold, feedback is generated to generate warning information.

[0054] B. When data forecasting is needed, perform data forecasting by setting the forecasting time.

[0055] 1. Data prediction of different types of emission sources. First, the accuracy level of emission source metering equipment in Model 2 and the acquired data is used to calculate the dependent variable, that is, the measurement result, based on the independent variable. E yi Substitute into model 2, and use the calculated results as the predicted emission source data. The uncertainty of the predicted results is ( E yi ±3σ) into the results of Model 2.

[0056] 2. Data prediction for different monitoring points of the same emission source type. First, set the accuracy level of model 4 and the measuring equipment, and calculate the dependent variable, i.e. the measurement result, based on the independent variable. AD jn-1 Substitute the calculated results into model 4 as the predicted emission source data, and the uncertainty of the predicted results is ( AD jn-1 ±3σ) into the results of Model 4.

[0057] Note: When the data is measured as a normal distribution, its probability density function p ( x )for ;

[0058] when When , the confidence interval probability is 99.73%, so 3 times the measurement deviation of the independent variable measuring instrument is selected as the deviation threshold.

[0059] Step 6: Confirmation of data verification. Classify and analyze the recorded warning information. Through automatic analysis and identification, if relevant data is missing, it is judged that there is a problem in the data collection or transmission process. Then, the data is fitted according to the correction model and the prediction correction is made. The deviation uncertainty in the data fitting is marked according to step 5. If the system has received the data, but the deviation from the model is large, it is confirmed through manual analysis that the production status has changed. Model 2 and Model 4 are readjusted according to the actual production status to confirm the data calculation results.

[0060] Step 7: Confirmation of data prediction results. If there are production changes, the corresponding Model 2 or Model 4 needs to be corrected, and then the fitting calculation is performed. The judgment can be made based on the relationship between the relevant measurement data and other equipment data. The prediction result deviation is the f( E yi ±3σ) or f( AD jn-1 ±3σ).

[0061] Step 8: Calculate the carbon emissions of industrial enterprises and provide relevant forecast and warning information, including the proportion of warnings, forecast data deviation, etc.

[0062] like Figure 1 As shown, a specific embodiment of the present invention discloses an online analysis method of carbon emissions based on multi-source metering data, which is used to analyze the carbon emissions of at least one emission unit, each emission unit includes multiple types of emission sources, and each type of emission source includes one or more monitoring points. The method steps are as follows:

[0063] Step 10: learning and establishing a mathematical model based on the emission source historical data of the emission unit, wherein the mathematical model includes an emission calculation model for the emission unit, a historical data relationship model for different emission sources, a calculation model for different monitoring points of the same emission source type, and a mathematical relationship model for activity levels between different monitoring points of the same emission source type;

[0064] The specific models are as follows:

[0065] The calculation model of the emission unit emission is: ;

[0066] Among them, E 总 represents the total emissions per emission unit, E yi Indicates i Types of emission source data, n is the number of emission source types of the emission unit, including direct emissions, indirect emissions and energy indirect emissions, etc.;

[0067] The historical data relationship model of different types of emission sources is: ;

[0068] Among them, f(E yi ) is the data relationship function between different emission sources obtained by historical data analysis;

[0069] The calculation model for different monitoring points of the same emission source type is: ; E dj =AD j ×EF j；

[0070] Among them, AD j represents the activity level data of the jth monitoring point under the same emission source, EF j represents the emission factor of the jth monitoring point;

[0071] The mathematical relationship model of activity levels between different monitoring points of the same emission source type is: ;

[0072] Among them, f( AD j ) is a mathematical relationship function for obtaining the activity levels of different monitoring points of the same emission source type based on historical data.

[0073] Step 20, inputting the metrological performance of the metrological equipment for measuring each data source, wherein the metrological performance includes a measurement standard deviation σ;

[0074] Step 30: According to the historical data relationship model of different types of emission sources, the existing emission source data is used as an independent variable, and the emission source that needs to be predicted or verified is used as a dependent variable to obtain a predicted value or a verified value; according to the mathematical relationship model of activity levels between different monitoring points of the same emission source type, the emission amount of the metered monitoring point that has been obtained is determined as an independent variable, and the emission data of the monitoring point that needs to be predicted or verified is used as a dependent variable to obtain a predicted value or a verified value;

[0075] Data verification process: perform data fitting verification according to the set data verification time limit and verification frequency; determine the independent variable based on the emission source data of the acquired data and the accuracy of the metering equipment (determine the metering equipment with higher accuracy as the independent variable), and then calculate the verification value of the dependent variable based on the historical data relationship model of different types of emission sources or the mathematical relationship model of activity levels between different monitoring points of the same emission source type. When the verification value is within the deviation threshold (±3σ), it is determined that the verification value is consistent with the actual measurement value. If it exceeds the deviation threshold, feedback is generated to generate warning information;

[0076] Data prediction process: When disconnected data or abnormal data is generated, if it is data at the emission source level, the predicted value is calculated based on the historical data relationship model of different types of emission sources and the existing emission source data; if it is data at the monitoring point level, the predicted value is calculated based on the mathematical relationship model of activity levels between different monitoring points of the same emission source type and the existing monitoring point data, and the uncertainty of the predicted value is given according to the model;

[0077] Step 40: Analyze the warning information recorded in the data verification process. If it is a problem in the data collection or transmission process, perform data fitting and correction according to the correction model, and mark the deviation uncertainty in the data fitting. If it is a change in production status, it is necessary to analyze and confirm according to the actual production status adjustment;

[0078] Step 50: Calculate the carbon emissions of the industrial enterprise based on the measured values ​​of the monitoring points and the prediction results when the measured values ​​of the monitoring points are missing, and provide relevant information on the prediction and early warning.

[0079] The present invention has the following advantages:

[0080] By establishing a mathematical model for measuring carbon emission multi-source data, the data of different types of emission sources or different monitoring points of the same emission source type are verified according to metrological methods, and predictions are made through metrological methods. For enterprises with data interruption and data anomalies, the measurement data model is supplemented and improved through mathematical model correction and combined with the actual production of the enterprise, so that within a certain period of time, the data of the monitoring system truly reflects the carbon emissions of the enterprise.

[0081] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for online analysis of carbon emissions based on multi-source metering data, characterized in that: For analyzing carbon emissions of at least one emission unit, each emission unit includes multiple types of emission sources, each type of emission source includes one or more monitoring points, the method includes: Step 10: learning and establishing a mathematical model based on the emission source historical data of the emission unit, wherein the mathematical model includes an emission calculation model for the emission unit, a historical data relationship model for different types of emission sources, a calculation model for different monitoring points of the same emission source type, and a mathematical relationship model for activity levels between different monitoring points of the same emission source type; The calculation model of the emission unit emission is as follows: Among them, E 总 Indicates the total emission of emission units, E yi represents the data of the ith emission source, and n is the number of emission source types of the emission unit, including direct emissions, indirect emissions, and energy indirect emissions; The historical data relationship model of different types of emission sources is: AND yi =f(E y1 ,AND y2 ...,AND yi-1 ) Among them, f(·) is the data relationship function between different types of emission sources obtained by historical data analysis; The calculation model for different monitoring points of the same emission source type is: AND dj =AD j ×EF j Among them, AD j represents the activity level data of the jth monitoring point under the same emission source, EF j represents the emission factor of the jth monitoring point; The mathematical relationship model of activity levels between different monitoring points of the same emission source type is: AD jn =f1(AD j1 ,AD j2 ...,AD jn-1 ) Among them, f1(·) is the mathematical relationship function of the activity level of different monitoring points of the same emission source type obtained from historical data; Step 20, inputting the metrological performance of the metrological equipment for measuring each data source, wherein the metrological performance includes the measurement standard deviation σ of the metrological equipment; Step 30: Data verification process and data prediction process, specifically as follows: Data verification process: Data fitting verification is performed according to the set data verification time limit and verification frequency; for data verification of different types of emission sources, the independent variable is determined based on the acquired emission source data and the accuracy of its metering equipment, and then the verification value of the dependent variable is calculated based on the historical data relationship model of different types of emission sources; for data verification of different monitoring points of the same emission source type, the independent variable is determined based on the activity level data of the existing monitoring points and the accuracy of their metering equipment, and then the verification value of the dependent variable is calculated based on the mathematical relationship model of the activity levels between different monitoring points of the same emission source type; when the verification value is within the deviation threshold range, it is determined that the verification value is consistent with the actual measurement value. If it exceeds the deviation threshold, feedback is generated to generate warning information; Data prediction process: When disconnected data or abnormal data is generated, if it is data at the emission source level, the predicted value is calculated based on the historical data relationship model of different types of emission sources and the existing emission source data; if it is data at the monitoring point level, the predicted value is calculated based on the mathematical relationship model of activity levels between different monitoring points of the same emission source type and the existing monitoring point activity level data; the uncertainty of the predicted value is given according to the corresponding model; Step 40: Analyze the warning information recorded in the data verification process. If it is a problem in the data collection or transmission process, perform data fitting and correction according to the correction model, and mark the deviation uncertainty in the data fitting. If it is a change in production status, it is necessary to analyze and confirm according to the actual production status adjustment; Step 50: Calculate the carbon emissions of industrial enterprises based on the activity level data of the monitoring points and the prediction results when the activity level data of the monitoring points are missing, and provide relevant information on the prediction and early warning.

2. The method according to claim 1, characterized in that: In step 30, the deviation threshold is ±3σ.

Images