Regional carbon emission model construction method and device based on energy big data
Through the construction method and system of regional carbon emission model based on energy big data, the problems of poor timeliness and high cost of carbon emission monitoring in the existing technology are solved, high-frequency carbon emission monitoring and prediction are realized, and monitoring accuracy and timeliness are improved. Provides an effective reference for energy planning.
Patent Information
- Application Number
- CN202410982969.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-12
AI Technical Summary
Existing carbon emission monitoring methods, such as carbon emission factor calculation algorithm, actual measurement method and carbon satellite observation method, cannot effectively meet the basic requirements of the "dual control" system of carbon emissions, poor timeliness or high cost, and cannot achieve high frequency, Low-cost carbon monitoring analysis.
Using regional carbon emission model construction methods and systems based on energy big data, through the combination of computer client terminals, application servers and database servers, a convenient and fast carbon emission monitoring and analysis model tool and human-computer interactive interface are built, and real-time power data is used to achieve different High-frequency carbon emission monitoring and forecasting in time, region and industry.
It greatly improves the timeliness of carbon emission monitoring, provides reference for the rational formulation of energy planning goals and progress supervision, provides a regional-level monthly carbon emission calculation solution, and improves the accuracy of model calculation.
Smart Images

Figure CN118940947A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of carbon emission monitoring and analysis, and specifically relates to a method and system for constructing a regional carbon emission model based on energy big data. Background Art
[0002] In the context of responding to global climate change, the transition to a low-carbon economy with low energy consumption, low emissions and low pollution has become the overall trend of world economic development. At present, the three carbon emission monitoring methods recognized at home and abroad, namely the carbon emission factor accounting method, the actual measurement method and the carbon satellite observation method, cannot effectively meet the basic requirements of the carbon emission "dual control" system. It is urgent to explore and build a low-cost and high-frequency carbon monitoring and analysis method.
[0003] (I) The carbon emission factor accounting method is the most commonly used, but its timeliness is poor. The basic idea of the emission factor method is to construct activity data and emission factors for each type of emission source, and then use the product of the two as the carbon emissions of the emission source. According to this method, my country has carried out national carbon emission accounting for a total of five years, namely 1994, 2005, 2010, 2012 and 2014. However, this method is limited by the long data statistical cycle and has poor timeliness. Taking energy consumption data as an example, it is generally released more than one year later.
[0004] (ii) The actual measurement method of carbon emissions requires the installation of additional metering devices, which makes it difficult to promote and apply. The actual measurement method uses metering devices to monitor the flow rate, pressure, temperature, humidity and concentration of the emission gas in real time to achieve accurate measurement of carbon emissions. This method has the advantage of high accuracy, but it is still in its infancy. On the one hand, the investment in installing additional equipment is unaffordable for companies. Surveys show that the total procurement cost of monitoring equipment is between 500,000 and 1.6 million yuan. On the other hand, the scope of application is limited, and it is not suitable for industries such as chemical and petrochemical industries with complex processes and non-fixed emission sources.
[0005] (III) The carbon satellite observation method can only monitor the concentration state, but cannot directly monitor carbon emissions. The carbon satellite method uses carbon satellite scanning to obtain regional carbon dioxide concentrations, and after data completion and other processes, it can monitor the carbon dioxide concentration in the region. This method has the advantages of high precision and support for repeated observations, but it can only monitor the carbon concentration state, and cannot accurately extract the carbon emissions caused by various activities, let alone distinguish the carbon emissions of various industries. Overall, it is in the experimental research stage.
[0006] Current research can be divided into two categories: region and industry, and the analysis cycle is mostly annual. Most regional studies use economic, population, energy intensity, and energy structure data to predict and analyze regional carbon emissions, while industry studies mostly focus on the production process of a specific industry, and calculate carbon emissions through product output, energy consumption, and production processes. There is a relative lack of research on high-frequency carbon emission models for all industries in a specific region. Summary of the invention
[0007] The purpose of the present invention is to provide a method and system for constructing a regional carbon emission model based on energy big data in response to the problems existing in the prior art. By constructing a convenient and fast carbon emission monitoring and analysis model tool and a human-computer interaction interface, based on high real-time electricity data, high-frequency monitoring and prediction of carbon emissions at different times, regions, and industries can be achieved, which greatly improves the timeliness of carbon emission monitoring and provides a reference for the reasonable formulation of energy planning goals and progress supervision.
[0008] The technical solution of the present invention is:
[0009] A device for constructing a regional carbon emission model includes a computer client terminal, a switch, an application server, and a database server. The computer client terminal is connected to the application server and the database server through an Ethernet network to realize data exchange. The computer client terminal, the application server, and the database server are connected to the switch through corresponding communication units.
[0010] Specifically, the database server includes several data storage modules.
[0011] The block includes a basic data unit, a parameter data unit, and a result data unit. The basic data unit is used to obtain and store data in the energy subject domain, the energy consumption subject domain, and the energy-carbon subject domain. The parameter data unit is used to obtain and store the time, region, and industry energy-carbon data labels set by the computer client terminal through the interactive module, the unit root, equilibrium relationship, causal relationship test method, frequency conversion method, and parameter data of the model optimization method. The result data unit is used to obtain and store various result data of the application server computing module.
[0012] Specifically, the application server includes an interactive module, a computing module and a display module.
[0013] Specifically, the Ethernet network includes a switch and a plurality of communication units; the communication unit includes two Ethernet network connectors and an unshielded twisted pair, and both ends of the unshielded twisted pair are respectively connected to an Ethernet network connector.
[0014] Specifically, the interactive module includes a module for receiving energy and carbon data based on time, region, industry, etc.
[0015] The first interactive operation unit set by the signature receives the second interactive operation unit set based on the unit root, equilibrium relationship, and causal relationship test method, receives the third interactive operation unit set based on the frequency conversion method, receives the fourth interactive operation unit set based on the model optimization method, and updates the monthly electricity consumption, energy consumption, product output, and carbon emissions of different times, regions, and industries according to the interactive operations.
[0016] Specifically, the computing module includes a data preprocessing unit for processing the first interaction of the interactive module
[0017] The operation unit sets the energy and carbon data labels such as time, region, and industry to filter out the original data required for calculation, and complete the missing data and output of abnormal data; the data verification unit is used to complete the data relationship test between electricity and energy consumption, and between electricity and product output according to the unit root, equilibrium relationship, and causal relationship test methods set by the second interactive operation unit of the interactive module, and output the calculation results of whether the test is passed or not; the frequency conversion unit, for the sequence that has passed the data test, performs monthly splitting on the sequence according to the frequency conversion method set by the third interactive operation unit of the interactive module; the factor decomposition unit, according to the factor decomposition method set by the fourth interactive operation unit of the interactive module, splits a monthly sequence into multiple ones; the total amount correction unit, adds up the monthly data after splitting, queries the original value of the annual data stored in the basic data unit according to the energy and carbon data labels such as time, region, and industry set by the first interactive operation of the interactive module, and performs total amount correction on the monthly data after splitting according to the original value; the model optimization unit, according to the model optimization method set by the fifth interactive operation unit of the interactive module, compares the model target values under different factor decompositions, sequence decompositions, and sequence types, and selects the optimal model suitable for different regions and industries.
[0018] A method for constructing a regional carbon emission model based on energy big data, using the regional carbon emission model construction device as described above, comprises the following steps:
[0019] S1: According to the energy-carbon database construction rules, the energy subject domain, energy consumption subject domain, and energy-carbon subject domain data system of the province to be detected are completed, and the basic data of historical years are entered into the basic data unit stored in the device. The basic data include: industrial product output, electricity consumption, physical energy consumption, comprehensive energy consumption, industrial production carbon emission factor, and energy consumption carbon emission factor;
[0020] S2: Calculate the energy consumption carbon emissions of historical years using the physical quantity of energy consumption and the carbon emission factor of energy consumption; calculate the industrial production carbon emissions of historical years using the industrial product output and the industrial production carbon emission factor; calculate the electricity and heat carbon emission factors of historical years using the physical quantity of energy consumption, the carbon emission factor of energy consumption, the electricity consumption, and the heat consumption; calculate the electricity exchange carbon emissions of historical years using the amount of electricity transferred in and out and the electricity carbon emission factor, and store them in the result data unit of the device;
[0021] S3: using the second interactive operation module to set the data test method as unit root ADF, cointegration EG, and Granger causality test method, using the third interactive operation module to set the frequency conversion method as cubic spline interpolation, using the fourth interactive operation module to set the factor decomposition method as X-13-ARIMA, using the fifth interactive operation module to set the model optimization method as rolling error method, and storing it in the parameter data unit of the device;
[0022] S4: Obtain the high-frequency electricity-carbon calculation model group for the region and industry, store it in the parameter data unit of the device, and store the energy consumption and carbon emissions calculated according to the latest monthly electricity consumption in the result data unit. Through the data labels such as time, region, industry, etc. set in the first interactive module, data visualization is realized in the display module.
[0023] The beneficial effects of the present invention are as follows: Applied in the field of carbon emission monitoring and analysis, the system and method provided by the present invention achieves high-frequency monitoring and prediction of carbon emissions at different times, regions, and industries based on real-time electricity data by building a convenient and fast carbon emission monitoring and analysis model tool and a human-computer interaction interface, which greatly improves the timeliness of carbon emission monitoring and provides a reference for the rational formulation of energy planning goals and progress supervision. It provides a solution for monthly carbon emission calculation at the regional level, effectively solving the problems of difficult basic data collection, long calculation cycle, and long time lag of traditional carbon emission calculation methods. At the same time, through the rolling correction and multi-analogy selection of the model, the calculation accuracy of the model is greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 It is a structural schematic diagram of the regional carbon emission model building device provided by the present invention.
[0025] Figure 2 is the change trend of the secondary energy carbon emission factor of a certain central province from 2005 to 2020 in Example 2;
[0026] Figure 3 This is a schematic diagram of the carbon emission structure of energy activities in a certain province in 2023. DETAILED DESCRIPTION
[0027] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0028] Example 1
[0029] like Figure 1 The structure diagram of a regional carbon emission model building device provided by this embodiment is shown, including a computer client terminal, a switch, an application server, and a database server. The computer client terminal is connected to the application server and the database server respectively through the Ethernet network to realize data exchange. The computer client terminal, the application server and the database server are respectively connected to the switch through corresponding communication units. This embodiment allows multiple computer client terminals to be connected in parallel at the same time. The computer client terminal is used to interact with the interactive module of the application server and set relevant parameters, store and retrieve basic data, parameter data, and result data of the data storage module of the database server, and interact with the display module to display data and results. The Ethernet network provides physical connections between the computer client terminal and the application server, and between the computer client terminal and the database server for the exchange, transmission and sharing of various data. The display module is used to display the electricity, energy consumption, product output, and carbon emissions of different times, regions, and industries according to the determined attribute dimensions.
[0030] In this embodiment, the database server includes several data storage modules.
[0031] The storage module includes a basic data unit, a parameter data unit, and a result data unit. The basic data unit is used to obtain and store data in the energy subject domain, the energy consumption subject domain, and the energy-carbon subject domain. The parameter data unit is used to obtain and store the time, region, and industry energy-carbon data labels set by the computer client terminal through the interactive module, the unit root, equilibrium relationship, causal relationship test method, frequency conversion method, and parameter data of the model optimization method. The result data unit is used to obtain and store various result data of the application server computing module. The various result data include: monthly electricity consumption, energy consumption, product output, carbon emissions, etc. at different times, regions, and industries.
[0032] In this embodiment, the application server includes an interactive module, a calculation module and a display module. The interactive module includes a first interactive operation unit for receiving energy and carbon data labels based on time, region, industry, etc., a second interactive operation unit for receiving settings based on unit root, equilibrium relationship, and causal relationship test methods, a third interactive operation unit for receiving settings based on frequency conversion methods, and a fourth interactive operation unit for receiving settings based on model optimization methods. The monthly electricity, energy consumption, product output, and carbon emissions of different times, regions, and industries are updated and displayed accordingly according to the interactive operation. According to actual needs, the interactive module is also used to click Save to trigger the interactive operation, and save the manually configured frequency conversion method and model optimization method. According to actual needs, the interactive module is also used to: click the icon switch to trigger the display of corresponding table data and graphics.
[0033] The calculation module includes a data preprocessing unit, which is used to filter out the original data required for calculation according to the energy and carbon data labels such as time, region, industry, etc. set by the first interactive operation unit of the interactive module, and complete the missing data filling and abnormal data output; a data verification unit, which is used to complete the data relationship test between electricity and energy consumption, and electricity and product output according to the unit root, equilibrium relationship, and causal relationship test methods set by the second interactive operation unit of the interactive module, and output the calculation result of whether the test is passed; a frequency conversion unit, for the sequence that has passed the data test, according to the frequency conversion method set by the third interactive operation unit of the interactive module Monthly splitting; factor decomposition unit, splits a monthly sequence into multiple ones according to the factor decomposition method set in the fourth interactive operation unit of the interactive module; total amount correction unit, adds up the split monthly data, queries the original value of the annual data stored in the basic data unit according to the energy and carbon data labels such as time, region, industry set in the first interactive operation of the interactive module, and corrects the total amount of the split monthly data according to the original value; model optimization unit, compares the model target values under different factor decompositions, sequence decompositions, and sequence types according to the model optimization method set in the fifth interactive operation unit of the interactive module, and selects the optimal model suitable for different regions and industries.
[0034] In this embodiment, the Ethernet network includes a switch and several communication units; the communication units
[0035] The device includes two Ethernet network connectors and a non-shielded twisted pair cable. Two ends of the non-shielded twisted pair cable are respectively connected to an Ethernet network connector.
[0036] In this embodiment, the switch is a Cisco SF100D-08 switch, the Ethernet network connector is an RJ45 Ethernet network connector; the unshielded twisted pair is a CAT5E unshielded twisted pair. In this embodiment, the minimum configuration requirements of the computer client terminal are as follows: operating system Centos Linux 6.8, processor 1GHz, memory 1GB, graphics card supports DirectX9128M, hard disk space 16G, equipped with Ethernet network interface, Google browser version 72 or above. Recommended configuration of application server: 8-core CPU, 16G memory, 100G hard disk, recommended configuration of database server: 4-core CPU, 256G memory, 5T hard disk.
[0037] Example 2
[0038] This embodiment provides a method for constructing a regional carbon emission model based on energy big data, using the regional carbon emission model construction device described in Example 1, including the following steps:
[0039] S1: According to the energy-carbon database construction rules, the energy subject domain, energy consumption subject domain, and energy-carbon subject domain data system of the province to be detected are completed, and the basic data of historical years are entered into the basic data unit stored in the device. The basic data include: industrial product output, electricity consumption, physical energy consumption, comprehensive energy consumption, industrial production carbon emission factor, and energy consumption carbon emission factor;
[0040] S2: Calculate the energy consumption carbon emissions of historical years using the physical quantity of energy consumption and the carbon emission factor of energy consumption; calculate the industrial production carbon emissions of historical years using the industrial product output and the industrial production carbon emission factor; calculate the electricity and heat carbon emission factors of historical years using the physical quantity of energy consumption, the carbon emission factor of energy consumption, the electricity consumption, and the heat consumption; calculate the electricity exchange carbon emissions of historical years using the amount of electricity transferred in and out and the electricity carbon emission factor, and store them in the result data unit of the device;
[0041] S3: using the second interactive operation module to set the data test method as unit root ADF, cointegration EG, and Granger causality test method, using the third interactive operation module to set the frequency conversion method as cubic spline interpolation, using the fourth interactive operation module to set the factor decomposition method as X-13-ARIMA, using the fifth interactive operation module to set the model optimization method as rolling error method, and storing it in the parameter data unit of the device;
[0042] S4: Obtain the high-frequency electricity-carbon calculation model group for the region and industry, store it in the parameter data unit of the device, and store the energy consumption and carbon emissions calculated according to the latest monthly electricity consumption in the result data unit. Through the data labels such as time, region, industry, etc. set in the first interactive module, data visualization is realized in the display module.
[0043] In the above step S1, the construction of the energy carbon database includes the following operations: according to the energy type, the energy subject domain is divided into 6 first-level subject domains, including coal, oil products, gas, heat, electricity, and other energy. The relationship between each subject domain depends on the production, transportation and consumption business of various energy industries, and the objects in the subject domain can be referenced by other domains. The first-level subject domain of coal includes 11 second-level subject domains such as raw coal, washed coal, other washed coal, coal products, coal gangue, coke, coke oven gas, blast furnace gas, converter gas, other gas, and other coking products. The first-level subject domain of oil products includes 14 second-level subject domains such as crude oil, gasoline, kerosene, diesel, fuel oil, naphtha, lubricating oil, paraffin, solvent oil, petroleum asphalt, petroleum coke, liquefied petroleum gas, refinery dry gas, and other petroleum products. The first-level subject domain of gas includes two second-level subject domains of natural gas and liquefied natural gas. Heat, electricity, and other energy sources no longer have second-level subject domains.
[0044] According to the type of energy consumption, the energy consumption subject domain is divided into 6 first-level subject domains, including the amount of energy available for consumption in the region, the input-output amount of processing and conversion, the amount of loss, the terminal consumption, the balance difference, and the total consumption. The first-level subject domain of energy available for consumption in the region includes 8 second-level subject domains, including the amount of primary energy production, the amount of transfer from other regions, the import amount, the amount of refueling of domestic aircraft and ships abroad, the amount of transfer from the region, the export amount, the amount of refueling of foreign aircraft and ships in the country, and the increase and decrease of inventory. The input-output amount of processing and conversion includes 9 second-level subject domains, including thermal power generation, heating, coal washing, coking, oil refining and coal-to-oil, gas production, natural gas liquefaction, coal product processing, and energy recovery. The terminal consumption includes 7 second-level subject domains, including agriculture, forestry, animal husbandry and fishery, industry, construction, transportation, warehousing and postal industry, wholesale, retail, accommodation and catering, others, and residents' life. The loss amount, balance difference, and total consumption no longer have second-level subject domains. Industry has 1 third-level subject domain for use as raw materials and materials.
[0045] According to the type of analysis results, the energy-carbon theme domain is divided into four first-level theme domains, including energy consumption, product output, carbon emission factor, and carbon emissions. The first-level theme domain of energy consumption includes five second-level theme domains: physical energy consumption, standard energy consumption, average low calorific value, carbon content per unit calorific value, and carbon oxidation rate. The first-level theme domain of product output includes five second-level theme domains: cement, lime, steel, calcium carbide, and others. The first-level theme domain of carbon emission factor includes three second-level theme domains: energy consumption carbon emission factor, industrial production carbon emission factor, and others. The first-level theme domain of carbon emissions includes three second-level theme domains: energy consumption carbon emissions, industrial production carbon emissions, and indirect carbon emissions contained in power exchange.
[0046] The number of primary and secondary subject domains can be further expanded according to actual conditions. Different time, region, and industry tags are set for different energy subject domains, energy consumption subject domains, and energy-carbon subject domains to achieve on-demand cross-query retrieval and rapid call of energy-carbon data indicators.
[0047] In the above step S2, the calculation of regional carbon emissions includes the following operations:
[0048] The scope of carbon emission calculation includes five major areas: energy activities, industrial processes, agricultural activities, waste treatment, and land use. Among them, carbon emissions from energy activities and industrial processes account for 99.7% of the total emissions. Carbon emissions from energy activities refer to carbon dioxide emissions caused by the combustion of various energy sources such as coal, oil, and natural gas. Carbon emissions from industrial processes refer to carbon emissions caused by chemical reactions in the processing of industrial products such as steel and cement. In addition, taking into account the emission reduction responsibilities of various regions, the indirect carbon dioxide emissions from electricity exchange are calculated.
[0049] Regional carbon emissions = carbon emissions from energy consumption + carbon emissions from industrial production + indirect carbon emissions from electricity exchange
[0050] In terms of measurement ideas, we follow the "three conversions". First, the carbon emission monitoring of energy consumption follows the principle of "electricity converted to energy + energy converted to carbon". That is, the historical energy consumption and energy carbon emission factors are used to calculate the historical carbon emissions, the historical electricity, energy consumption and carbon emissions are used to train the model, and the electricity of the current month is used to predict the carbon emissions of the current month. Second, the carbon emission monitoring of industrial production follows the principle of "electricity converted to production + production converted to carbon". That is, the historical electricity and product output are used to train the model, the monthly electricity is used to predict the monthly product output, and the product carbon emission factors are used to calculate carbon emissions. Third, the indirect carbon emissions of power exchange follow the principle of "electricity converted to carbon". That is, the indirect carbon emissions are calculated using the regional power transfer in and out, the power structure, and the carbon emission factors of the power supply in the transferred-in region.
[0051] 1. Energy consumption
[0052] Energy consumption carbon emission calculation includes two methods: production side and consumption side. Traditional carbon emission calculation is based on "producer responsibility", which means whoever produces is responsible, but it is not conducive to the reasonable allocation of the overall carbon emission reduction responsibility of the whole society. Considering that the current energy consumption statistical principle is "whoever consumes, whoever counts", two types of models, production side and consumption side, are constructed at the same time. The main difference lies in the calculation method of two secondary energy sources, electricity and heat.
[0053] The calculation scope of energy types in the production side model is four first-level subject domains: coal, oil products, gas, and other energy, and only includes primary energy. The calculation scope of energy consumption types is the sum of the second-level subject domains of thermal power generation and heating in the first-level subject domain of processing and conversion input-output, minus the third-level subject domain of raw materials and materials in the second-level subject domain of industry in the first-level subject domain of terminal consumption.
[0054]
[0055] Q 生产侧 =Q 火力发电 +Q 供热 Q 工业用作原料、材料 (2)
[0056] Among them, C E is the carbon dioxide emissions from energy consumption, Q i is the physical amount of energy consumption of the i-th type, k is the number of various energy types under the energy subject domain, NCV i is the average low calorific value of the i-th energy source, A i is the carbon content per unit calorific value of the i-th energy source, O i is the carbon oxidation rate during the combustion of the i-th energy source. NCV, A, and O should use regional data first. If they are not available, the default values in the IPCC National Greenhouse Gas Inventory Guidelines and the Provincial Greenhouse Gas Inventory Compilation Guidelines can be used for selection.
[0057] The energy type calculation scope of the consumption side model is 6 first-level subject domains: coal, oil, gas, heat, electricity, and other energy, including primary energy and secondary energy. The energy consumption type calculation scope is the first-level subject domain of terminal consumption, minus the third-level subject domain of raw materials and materials in the second-level subject domain of industry.
[0058] C 消费侧 =C 一次能源 +C 二次能源 (3)
[0059] C 二次能源 =C 电力 +C 热力 =(Q 调入电力 -Q 调出电力 )·EF 电力 +(Q 调入热力 -Q调出热力 )·EF 热力 (4)
[0061] Among them, EF is the carbon emission factor of secondary energy. With the development of the economy and society, the carbon emissions caused by secondary energy consumption show a gradually increasing trend. The regional transfer in and out of heat is relatively small, and the carbon emission factors between regions are not much different. However, the regional transfer in and out of electricity is relatively large. Most traditional carbon emission factors use the national power grid average emission factor announced by the country from time to time, which cannot reasonably reflect the changes in the power installed capacity structure in different regions and at different stages. This embodiment proposes a simple and easy method for calculating the dynamic carbon emission factor of regional secondary energy.
[0062]
[0063] Among them, EF y is the carbon emission factor of secondary energy in year y, EG y The physical amount of secondary energy consumed in year y. Energy consumption type calculation range, EG y It is the first-level theme domain of total consumption; electricity is the second-level theme domain of thermal power generation in the first-level theme domain of processing and conversion input-output; heat is the second-level theme domain of heating in the first-level theme domain of processing and conversion input-output.
[0064] Taking a central province as an example, the changes in the secondary energy carbon emission factor of the region calculated according to the above method between 2005 and 2020 are as follows: Figure 2 As shown in the figure, the results are consistent with the province’s clean energy structure transformation trend, and the pilot years are also comparable to the provincial factors announced by the state, which can truly and reasonably achieve the calculation of regional secondary energy consumption carbon emissions.
[0065] 2. Industrial production
[0066] For industries such as steel, nonferrous metals, chemicals, petrochemicals, and building materials, the carbon emissions caused by the production processes of products such as pig iron, crude steel, finished steel, primary aluminum, synthetic ammonia, cement, and flat glass are calculated by comprehensively considering the industry's carbon emission characteristics and the availability of product output data.
[0067]
[0068] Among them, C I is the carbon dioxide emissions from industrial production, P i is the output of the i-th product, EF i is the industrial process carbon emission factor of the ith product, and k is the number of various industrial products in various industries within the measurement scope.
[0069] (III) Power Exchange
[0070] Taking into account the particularity of electricity products and scientifically evaluating the contribution of non-fossil fuel electricity to mitigating carbon dioxide emissions, the indirect carbon dioxide emissions caused by the transfer of electricity in and out were calculated.
[0071] C 电力调入(出) =∑(Q 调入(出)电力 ·EF 调入(出)省份 ) (7)
[0072] Among them, C 电力调入(出) Q is the indirect carbon dioxide emissions caused by electricity transfer in and out, 调入(出)电力 EF is the amount of electricity input (output). 调入(出)省份 It is the carbon emission factor of electricity imported into (exported from) the province.
[0073] (IV) High-frequency monitoring
[0074] On the basis of completing the regional annual carbon emissions calculation, a high-frequency monitoring model for carbon emissions is further constructed. The calculation process includes data preprocessing, data verification, frequency conversion, factor decomposition, total correction, model optimization, etc. Among them, the core of the data preprocessing link is the correction of missing data values and outliers. The core of the data verification link is to use unit root test method, cointegration test method and other methods to conduct correlation analysis on electricity and energy consumption, electricity and product output to ensure the rationality of the analysis results. The core of the model optimization link is to use the model rolling forecast error as the standard, different frequency conversions, sequence types, and sequence decomposition methods as selection parameters, and screen out the long-term equilibrium model under the optimal parameters to ensure that the error rate of historical years is controlled within 3%.
[0075] (1) Data preprocessing
[0076] For data in different energy, energy consumption, and energy-carbon domains, when the regional and industry labels are the same, they are sorted in chronological order. If the situation described in formula (8) or formula (9) occurs or the value is null, it is considered an outlier and needs to be corrected according to formula (9).
[0077] F i >20·(F i-1 +F i+1 ) (8)
[0078]
[0079]
[0080] (2) Data verification
[0081] Data verification includes:
[0082] 1) Unit root test series stationarity;
[0083] 2) If stable, cointegration test whether there is a long-term equilibrium relationship;
[0084] 3) If so, causality test whether there is a causal relationship in time sequence.
[0085] Unit root test
[0086] The unit root test can be used to test the stationarity of a given time series and determine whether the series contains a unit root; if there is no unit root, the test data is stationary.
[0087] Equilibrium relationship test
[0088] First, the variables are tested for stationarity, and then a classic linear regression model is constructed to test the stationarity of the residuals.
[0089] Causality test
[0090] Granger causality test is used to test whether one set of time series is the cause of another set of time series. It can be used to determine the relationship between electricity consumption, energy consumption, and product output.
[0091] (3) Frequency conversion
[0092] After the time series data is preprocessed and passed the data test, the spline interpolation method is used to split the annual series into monthly series. Taking the cubic spline interpolation method as an example, the specific operation steps are as follows:
[0093] Step 1: Insert 11 interpolation points in two adjacent years, corresponding to the 12 months in a year.
[0094] Step 2: Construct a cubic spline function S(x), where x represents the time scale (such as month), and S(x) should meet the following two conditions:
[0095] Condition 1: At the two end points x i and x i+1 At this point, the cubic spline function value is equal to the annual data value, that is, S(x i )=Y i and S(x i+1 )=Y i+1 .
[0096] Condition 2: Continuity condition: The cubic spline function and its first and second order derivatives are continuous between segments.
[0097] Step 3: Construct and solve a set of linear equations to determine the parameters of each segment of the cubic spline interpolation function. For each subinterval, the cubic spline function is in the form of:
[0098] S(x)=a+bx+cx2 +dx 3 (11)
[0099] Where a, b, c and d are the coefficients to be determined, and the boundary conditions and continuity conditions will be converted into linear equations constraining these coefficients.
[0100] Step 4: Using the solved coefficients, calculate S(x) for the x value corresponding to each month in each year to obtain the monthly data series.
[0101] (4) Factor decomposition
[0102] For the obtained monthly series, the X-13-ARIMA method, TRAMO-SEATS and other methods are used to decompose the original monthly series into TC series, SA series, S series and I series. The monthly series decomposition formula is:
[0103]
[0104] Among them, Y t is the monthly data series, TC t It is a combination of trend and cycle terms estimated by the ARIMA model. t It is the seasonal effect obtained according to the seasonal adjustment algorithm, usually expressed in the form of relative or absolute seasonal factors. t is the residual term, representing the portion not explained by trend and seasonality.
[0105] (5) Total quantity correction
[0106] The split monthly series are corrected using known annual data.
[0107]
[0108] Among them, F is the adjustment factor, E T represents the annual data for year T, is the monthly series data of year T.
[0109] (6) Model optimization
[0110] Taking the model rolling forecast error as the selection criterion, and different frequency conversion methods, different sequence decomposition methods and different sequence types as selection parameters, the long-term equilibrium model under the optimal parameters is estimated and screened.
[0111] For example, the model with the optimal parameters selected in the carbon emission calculation of key industries is the ADL-TC (q, s) model.
[0112]
[0113] Among them, YTCt is the monthly energy consumption data, α, β i and γ j is the parameter obtained after fitting, where t represents the month, q and s are the parameters determined by optimization, and ε t is white noise.
[0114] For example, the model with the optimal parameters selected in the calculation of residents' carbon emissions is the ADL-SA (q, s) model.
[0115]
[0116] Among them, YSA ij,t is the monthly electricity data, α, β ij,k and γ ij,h is the parameter obtained after fitting, q and s are the parameters determined by optimization, and ε ij,t is white noise, where t represents month, i represents consumption category, and j represents district or county.
[0117] The present invention can be applied to the field of carbon emission monitoring and analysis, and has been applied to provincial carbon emission monitoring and analysis work, and the relevant methods and conclusions have been adopted by the provincial energy planning research results system. By constructing a convenient and fast carbon emission monitoring and analysis model tool and a human-computer interaction interface, based on high-real-time electricity data, high-frequency monitoring and prediction of carbon emissions at different times, regions, and industries can be achieved, which greatly improves the timeliness of carbon emission monitoring and provides a reference for the reasonable formulation of energy planning goals and progress supervision.
[0118] The carbon emission monitoring model after optimization of agriculture, forestry, animal husbandry and fishery in a province is constructed using the above method (where X is electricity consumption and Y is carbon emissions): 0.0358+0.9986*logX-1.978*logX(-1)+1.9792*logY(-1)+0.9901*logX(-2)-0.991*logY(-2) (16).
[0119] The results are as follows Figure 3 shown.
[0120] From the above, it can be seen that the carbon emission monitoring method and system proposed in the present invention provide a solution for monthly carbon emission calculation at the regional level, which effectively solves the problems of difficult basic data collection, long calculation cycle and long time lag of traditional carbon emission calculation methods. At the same time, through the rolling correction and multi-analogy selection of the model, the calculation accuracy of the model is greatly improved.
[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or some technical features can be replaced by equivalents without departing from the spirit of the technical solution of the present invention, which should be included in the scope of the technical solution for protection of the present invention.
Claims
1. A device for constructing a regional carbon emission model, characterized in that: It includes a computer client terminal, a switch, an application server, and a database server. The computer client terminal is connected to the application server and the database server through an Ethernet network to realize data exchange. The computer client terminal, the application server and the database server are connected to the switch through corresponding communication units.
2. The regional carbon emission model building device according to claim 1, characterized in that: The database server includes several data storage modules, and the data storage modules include basic data units, parameter data units, and result data units. The basic data units are used to obtain and store energy subject domains, energy consumption subject domains, and energy-carbon subject domain data. The parameter data units are used to obtain and store time, region, industry energy-carbon data labels, unit roots, equilibrium relationships, causal relationship test methods, frequency conversion methods, and parameter data of model optimization methods set by computer client terminals through interactive modules. The result data units are used to obtain and store various result data of the application server computing module.
3. The regional carbon emission model building device according to claim 1, characterized in that: The application server includes an interactive module, a computing module and a display module.
4. The regional carbon emission model building device according to claim 1, characterized in that: The Ethernet network includes a switch and several communication units; the communication unit includes two Ethernet network connectors and a non-shielded twisted pair, and the two ends of the non-shielded twisted pair are respectively connected to an Ethernet network connector.
5. The regional carbon emission model building device according to claim 3, characterized in that: The interactive module includes a first interactive operation unit for receiving energy and carbon data label settings based on time, region, industry, etc., a second interactive operation unit for receiving settings based on unit root, equilibrium relationship, and causal relationship test methods, a third interactive operation unit for receiving settings based on frequency conversion methods, and a fourth interactive operation unit for receiving settings based on model optimization methods. The monthly electricity consumption, energy consumption, product output, and carbon emissions of different times, regions, and industries are updated and displayed accordingly according to the interactive operations.
6. The regional carbon emission model building device according to claim 5, characterized in that: The calculation module includes a data preprocessing unit, which is used to filter out the original data required for calculation according to the energy and carbon data labels such as time, region, industry, etc. set by the first interactive operation unit of the interactive module, and complete the completion of missing data and the output of abnormal data; The data verification unit is used to complete the data relationship test between the power consumption and energy consumption, and the power consumption and product output according to the unit root, equilibrium relationship, and causal relationship test methods set by the second interactive operation unit of the interactive module, and output the calculation result of whether the test is passed; The frequency conversion unit performs monthly splitting of the sequence that has passed the data inspection according to the frequency conversion method set by the third interactive operation unit of the interactive module; The factor decomposition unit decomposes a monthly sequence into multiple factors according to the factor decomposition method set by the fourth interactive operation unit of the interactive module; The total amount correction unit adds up the split monthly data, queries the original value of the annual data stored in the basic data unit according to the energy and carbon data labels such as time, region, and industry set in the first interactive operation of the interactive module, and corrects the total amount of the split monthly data according to the original value; the model optimization unit compares the model target values under different factor decompositions, sequence decompositions, and sequence types according to the model optimization method set in the fifth interactive operation unit of the interactive module, and selects the optimal model suitable for different regions and industries.
7. A method for constructing a regional carbon emission model based on energy big data, using the regional carbon emission model construction device as described in any one of claims 1 to 6 above, characterized in that: The steps include: S1: According to the energy-carbon database construction rules, the energy subject domain, energy consumption subject domain, and energy-carbon subject domain data system of the province to be detected are completed, and the basic data of historical years are entered into the basic data unit stored in the device. The basic data include: industrial product output, electricity consumption, physical energy consumption, comprehensive energy consumption, industrial production carbon emission factor, and energy consumption carbon emission factor; S2: Calculate the energy consumption carbon emissions of historical years using the physical quantity of energy consumption and the carbon emission factor of energy consumption; calculate the industrial production carbon emissions of historical years using the industrial product output and the industrial production carbon emission factor; calculate the electricity and heat carbon emission factors of historical years using the physical quantity of energy consumption, the carbon emission factor of energy consumption, the electricity consumption, and the heat consumption; calculate the electricity exchange carbon emissions of historical years using the amount of electricity transferred in and out and the electricity carbon emission factor, and store them in the result data unit of the device; S3: using the second interactive operation module to set the data test method as unit root ADF, cointegration EG, and Granger causality test method, using the third interactive operation module to set the frequency conversion method as cubic spline interpolation, using the fourth interactive operation module to set the factor decomposition method as X-13-ARIMA, using the fifth interactive operation module to set the model optimization method as rolling error method, and storing it in the parameter data unit of the device; S4: Obtain the high-frequency electricity-carbon calculation model group for the region and industry, store it in the parameter data unit of the device, and store the energy consumption and carbon emissions calculated according to the latest monthly electricity consumption in the result data unit. Through the data labels such as time, region, industry, etc. set in the first interactive module, data visualization is realized in the display module.
