Neural network-based method and device for predicting missing data in carbon emission inventory

Abstract

The present application relates to the field of carbon emission inventory management, and discloses a method and device for predicting missing data of carbon emission inventory based on neural network, the method comprising: determining the missing time period corresponding to the missing carbon emission value to be completed in the carbon emission inventory, dividing the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information in the missing time period, and obtaining the time series activity level historical data of each missing sub-time period under similar production conditions; determining the time series activity level prediction data in each missing sub-time period according to the time series activity level historical data under similar production conditions in each missing sub-time period; determining the carbon emission prediction value in each missing sub-time period according to the emission factor data corresponding to the time series activity level prediction data in each missing sub-time period and the activity level prediction data at different time points. The present application can ensure the accuracy and scientificity of missing value completion of carbon emission inventory.

Description

Technical Field

[0001] The present application relates to the technical field of carbon emission inventory management, and more specifically, to a method and device for predicting missing data in a carbon emission inventory based on a neural network. Background Art

[0002] Carbon emission inventory is used to count greenhouse gas emissions of engineering projects, enterprises, institutions or regions, and to systematically collect, record and report direct and indirect greenhouse gas emissions within a set time period. Carbon emission inventory lists in detail the emissions of carbon dioxide (CO2) and other greenhouse gases (such as methane CH4, nitrous oxide N2O, etc.) generated by various activities. Carbon emission inventory is an important tool for assessing and managing greenhouse gas emissions. The main contents of carbon emission inventory include direct emission statistics, indirect emission statistics and other indirect emission statistics. Direct emission statistics include emissions from carbon emission sources directly owned or controlled by enterprises and institutions, such as emissions directly generated by coal-fired boilers and fuel used by vehicles of enterprises and institutions. Indirect emission statistics include emissions generated by energy purchased by enterprises and institutions, such as emissions generated by electricity purchased from the power grid. Other indirect emissions include emissions caused by the activities of enterprises and institutions but occurring outside the enterprises and institutions, such as emissions from suppliers, emissions from products during the use phase, and emissions from waste treatment. When determining the carbon emission inventory, the carbon emission inventory data is often missing due to reasons such as loss of statistical data. When formulating a carbon emission plan, the carbon emission inventory needs to be supplemented. How to scientifically and reasonably supplement the carbon emission inventory is of great significance.

[0003] Patent CN116451882B (application number: 202310718556.3) provides a method for predicting carbon emissions, which includes: obtaining the associated data of a product; when there are missing values ​​in the associated data, obtaining historical data corresponding to the missing values, including the objects for obtaining the missing values ​​and the time when the missing values ​​exist; inputting the historical data into a preset prediction model to obtain predicted data; according to the preset mapping rules, obtaining the mapping data corresponding to the associated data in the preset carbon emission factor database; according to the predicted data, the associated data and the mapping data, determining the predicted results of the carbon emissions of the product throughout its life cycle. The method in patent CN116451882B can obtain historical data including the objects for obtaining the missing values ​​and the time when the missing values ​​exist, and input the historical data into the preset prediction model to obtain the predicted data after the missing values ​​are filled in. This method for filling in the carbon emission inventory is difficult to ensure the accuracy and scientificity of the missing value filling. Summary of the invention

[0004] The purpose of this application is to provide a method and device for predicting missing data in a carbon emission inventory based on a neural network, which solves the technical problem that the existing carbon emission inventory completion method is difficult to ensure the accuracy and scientificity of missing value completion, and achieves the technical effect of ensuring the accuracy and scientificity of missing value completion in the carbon emission inventory.

[0005] An embodiment of the present application provides a method for predicting missing data in a carbon emission inventory based on a neural network, the method comprising: determining a missing time period corresponding to a missing carbon emission value that needs to be completed in the carbon emission inventory, dividing the missing time period into a plurality of missing sub-time periods according to activity level fluctuation event information within the missing time period, and obtaining time series activity level historical data under similar production conditions for each missing sub-time period; wherein the activity level fluctuation event information includes fault maintenance event information, production shutdown event information, and capacity adjustment event information; determining time series activity level prediction data for each missing sub-time period according to the time series activity level historical data under similar production conditions within each missing sub-time period through a time series activity level prediction model; determining a carbon emission prediction value for each missing sub-time period according to the time series activity level prediction data for each missing sub-time period and emission factor data corresponding to the activity level prediction data at different time points, and taking the sum of the carbon emission prediction values ​​for all missing sub-time periods as the prediction value of the carbon emission missing value.

[0006] In one possible implementation, the time series activity level historical data under similar production conditions for each missing sub-time period is obtained, including: obtaining the time series activity level historical data under similar production conditions for multiple continuous productions corresponding to the first missing sub-time period, and obtaining the time length of each time series activity level historical data; among the time series activity level historical data under similar production conditions for multiple continuous productions, determining multiple first time series activity level historical data whose time length is greater than that of the first missing sub-time period, and intercepting the middle segment of the time series activity level historical data of each first time series activity level historical data as the time series activity level historical data under similar production conditions for the first missing sub-time period.

[0007] In another possible implementation, obtaining the historical data of the time series activity level under similar production conditions for each missing sub-time period also includes: when the number of multiple first time series activity level historical data whose time length is greater than the first missing sub-time period is greater than a preset number, obtaining the first production condition corresponding to each first time series activity level historical data; wherein the first production condition includes production process arrangement characteristics, equipment status characteristics and raw material type characteristics; through a production condition comparison model, according to the first production condition corresponding to each first time series activity level historical data and the production condition corresponding to the first missing sub-time period, determining the closest first production condition corresponding to the first missing sub-time period, and using the time series activity level historical data corresponding to the closest first production condition as the time series activity level historical data under similar production conditions for the first missing sub-time period.

[0008] In another possible implementation, obtaining the time-series activity level historical data under similar production conditions for each missing sub-time period also includes: when there is no time-series activity level historical data under similar production conditions for the first missing sub-time period, obtaining a first influencing factor corresponding to the first activity level fluctuation event information before the first missing sub-time period; wherein the first influencing factor represents the degree of influence of the first activity level fluctuation event information on the time-series activity level historical data corresponding to the first missing sub-time period; obtaining the time-series activity level historical data corresponding to the preceding missing sub-time period before the first activity level fluctuation event information, and taking the product of the activity level historical data at each time point of the time-series activity level historical data corresponding to the preceding missing sub-time period and the first influencing factor as the time-series activity level historical data for the first missing sub-time period under similar production conditions.

[0009] In another possible implementation, obtaining the time-series activity level historical data for each missing sub-time period under similar production conditions also includes: when the first influencing factor changes with time, obtaining the time-series first influencing factor corresponding to the first activity level fluctuation event information before the first missing sub-time period; calculating the product of the time-series activity level historical data and the time-series first influencing factor corresponding to the preceding missing sub-time period according to the time correspondence, as the time-series activity level historical data for the first missing sub-time period under similar production conditions.

[0010] In another possible implementation, the activity level fluctuation event information also includes technology improvement event information. When the activity level fluctuation event information is technology improvement event information, the method also includes: obtaining first emission factor data corresponding to the first activity level fluctuation event information before the first missing sub-time period; wherein the first emission factor data is the emission factor data after the first activity level fluctuation event information; determining the carbon emission prediction value for each missing sub-time period based on the emission factor data corresponding to the time series activity level prediction data in each missing sub-time period and the activity level prediction data at different time points, including: determining the carbon emission prediction value for the first missing sub-time period based on the first emission factor data corresponding to the time series activity level prediction data in the first missing sub-time period and the activity level prediction data at different time points.

[0011] An embodiment of the present application also provides a device for predicting missing data in a carbon emission inventory based on a neural network, comprising a unit for executing any of the methods described above.

[0012] An embodiment of the present application also provides a carbon emission inventory missing data prediction device based on a neural network, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the methods described above when executing the computer program.

[0013] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the above items is implemented.

[0014] An embodiment of the present application further provides a computer program product, including a computer program, which implements the steps of any of the above methods when executed by a processor.

[0015] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0016] An embodiment of the present application provides a method for predicting missing data in a carbon emission inventory based on a neural network, the method comprising: determining a missing time period corresponding to a missing carbon emission value that needs to be completed in the carbon emission inventory, dividing the missing time period into a plurality of missing sub-time periods according to activity level fluctuation event information within the missing time period, and obtaining time series activity level historical data under similar production conditions for each missing sub-time period; wherein the activity level fluctuation event information includes fault maintenance event information, production shutdown event information, and capacity adjustment event information; determining time series activity level prediction data for each missing sub-time period according to the time series activity level historical data for each missing sub-time period through a time series activity level prediction model; determining a carbon emission prediction value for each missing sub-time period according to the time series activity level prediction data for each missing sub-time period and emission factor data corresponding to the activity level prediction data at different time points, and taking the sum of the carbon emission prediction values ​​for all missing sub-time periods as the prediction value of the carbon emission missing value. The method in the embodiment of the present application can perform fine processing on the missing time period corresponding to the missing value of carbon emissions in the carbon emission inventory, divide the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information within the missing time period, and obtain the time series activity level historical data of each missing sub-time period under similar production conditions, and then predict the time series activity level forecast data of each missing sub-time period based on the time series activity level historical data of each missing sub-time period, and then accurately predict the carbon emissions within the missing time period corresponding to the missing value of carbon emissions, thereby improving the prediction accuracy of the missing value of carbon emissions and improving the accuracy and scientificity of the formulation of carbon emission inventories and carbon emission plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0018] Figure 1 A schematic diagram of a flow chart of a method for predicting missing data in a carbon emission inventory based on a neural network provided in an embodiment of the present application;

[0019] Figure 2 A schematic diagram of a missing time period provided in an embodiment of the present application;

[0020] Figure 3 A schematic flow chart of another method for predicting missing data in a carbon emission inventory based on a neural network provided in an embodiment of the present application;

[0021] Figure 4A schematic diagram of the logical structure of a device for predicting missing data in a carbon emission inventory based on a neural network provided in an embodiment of the present application;

[0022] Figure 5 A schematic diagram of the logical structure of a neural network-based carbon emission inventory missing data prediction device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0024] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0026] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0027] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0028] In the prior art, when predicting and completing missing values ​​in a carbon emission inventory, historical data including the objects of missing values ​​and the time when the missing values ​​existed is obtained, and the historical data is input into a preset prediction model to obtain predicted data after completing the missing values. This method of completing the carbon emission inventory is difficult to ensure the accuracy and scientificity of the missing value completion.

[0029] Based on the above reasons, an embodiment of the present application provides a method for predicting missing data in a carbon emission inventory based on a neural network, the method comprising: determining the missing time period corresponding to the missing carbon emission values ​​that need to be completed in the carbon emission inventory, dividing the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information within the missing time period, and obtaining the time series activity level historical data of each missing sub-time period under similar production conditions; wherein the activity level fluctuation event information includes fault maintenance event information, production shutdown event information and capacity adjustment event information; through a time series activity level prediction model, according to the time series activity level historical data within each missing sub-time period, determining the time series activity level prediction data within each missing sub-time period; according to the time series activity level prediction data within each missing sub-time period and the emission factor data corresponding to the activity level prediction data at different time points, determining the carbon emission prediction value within each missing sub-time period, and taking the sum of the carbon emission prediction values ​​within all missing sub-time periods as the prediction value of the carbon emission missing value. The method in the embodiment of the present application can perform fine processing on the missing time period corresponding to the missing value of carbon emissions in the carbon emission inventory, divide the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information within the missing time period, and obtain the time series activity level historical data of each missing sub-time period under similar production conditions, and then predict the time series activity level forecast data of each missing sub-time period based on the time series activity level historical data of each missing sub-time period, and then accurately predict the carbon emissions within the missing time period corresponding to the missing value of carbon emissions, thereby improving the prediction accuracy of the missing value of carbon emissions and improving the accuracy and scientificity of the formulation of carbon emission inventories and carbon emission plans.

[0030] In some scenarios, a neural network-based carbon emission inventory missing data prediction method of an embodiment of the present application can be applied to the formulation of carbon emission inventories of enterprises and institutions, which can improve the accuracy and scientificity of the formulation of carbon emission inventories of enterprises and institutions.

[0031] The following is a detailed description of a method for predicting missing data in a carbon emission inventory based on a neural network provided in an embodiment of the present application with reference to specific examples.

[0032] Figure 1 A flowchart of a method for predicting missing data in a carbon emission inventory based on a neural network is provided in an embodiment of the present application, such as Figure 1As shown, the method includes S110 to S130, and S110 to S130 are described in detail below.

[0033] S110, determine the missing time period corresponding to the missing carbon emission value that needs to be supplemented in the carbon emission inventory, divide the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information in the missing time period, and obtain the time series activity level historical data of each missing sub-time period under similar production conditions. Among them, the activity level fluctuation event information includes fault maintenance event information, production shutdown event information and capacity adjustment event information.

[0034] When formulating a carbon emission inventory, some missing carbon emission values ​​in the carbon emission inventory often correspond to a certain time period. Therefore, we can first determine the missing carbon emission values ​​that need to be completed in the carbon emission inventory, and determine the missing time period corresponding to the missing carbon emission values, and then complete the carbon emission values ​​within the missing time period.

[0035] Exemplarily, the first missing time period corresponding to the first carbon emission missing value may be one month or one quarter.

[0036] When filling in missing values ​​of carbon emissions, the missing values ​​of carbon emissions of enterprises and institutions often correspond to missing carbon emission activity level data and emission factor data. Therefore, it is necessary to fill in the carbon emission activity level data and emission factor data corresponding to the missing values ​​of carbon emissions.

[0037] After obtaining the missing time period corresponding to the missing value of carbon emissions, in order to ensure the scientific nature of completing the carbon emission value in the missing time period, we can first collect the activity level fluctuation event information in the missing time period. The activity level fluctuation event information is the information of the event that causes the activity level fluctuation. Then, the missing time period can be segmented according to the activity level fluctuation event information, and the activity levels in the segmented missing time periods can be predicted separately to achieve accurate and scientific prediction of the activity level data.

[0038] Exemplarily, activity level fluctuation event information may include fault repair event information, production shutdown event information and capacity adjustment event information. The fault repair event information corresponds to activity level fluctuations caused by repairs after a production equipment failure, the production shutdown event information corresponds to activity level fluctuations caused by shutdown of production equipment, and the capacity adjustment event information corresponds to activity level fluctuations directly caused by capacity adjustments.

[0039] Exemplarily, when the first missing time period corresponding to the first carbon emission missing value is the first quarter, the capacity adjustment event information may be one or more fault repair event information, production shutdown event information, and capacity adjustment event information within the first quarter.

[0040] After obtaining the activity level fluctuation event information within the missing time period, the missing time period may be divided into a plurality of missing sub-time periods according to the activity level fluctuation event information within the missing time period, and the plurality of missing sub-time periods respectively correspond to different activity levels.

[0041] Figure 2 A schematic diagram of a missing time period provided in an embodiment of the present application, such as Figure 2 As shown, the missing time period may be divided into a first missing sub-time period, a second missing sub-time period and a third missing sub-time period by the first activity level fluctuation event information and the second activity level fluctuation event information.

[0042] After dividing the missing time period into multiple missing sub-time periods, in order to predict the activity level data of each missing sub-time period, the time-series activity level historical data of each missing sub-time period under similar production conditions can be obtained, and then the activity level data of each missing sub-time period can be predicted based on the time-series activity level historical data of each missing sub-time period under similar production conditions.

[0043] Exemplarily, the time-series activity level history data under similar production conditions in the first missing sub-time period may be the time-series activity level history data under similar production conditions in a previous cycle.

[0044] S120. Determine the predicted data of the time series activity level in each missing sub-time period according to the historical data of the time series activity level in each missing sub-time period through the time series activity level prediction model.

[0045] When predicting the activity level data for each missing sub-time period, the time series activity level prediction model can be used to determine the time series activity level prediction data for each missing sub-time period based on the time series activity level historical data under similar production conditions in each missing sub-time period, so as to predict the activity level data for each missing sub-time period in time series, and the predicted activity level data for each missing sub-time period is consistent with the fluctuation state of the time series activity level data in actual production, and then the activity level data for the entire missing time period can be calculated through the time series activity level prediction data for each missing sub-time period.

[0046] Exemplarily, the time series activity level prediction model may be a LSTM-based time series prediction model.

[0047] S130. Determine the predicted carbon emission value for each missing sub-time period based on the time-series activity level prediction data for each missing sub-time period and the emission factor data corresponding to the activity level prediction data at different time points, and take the sum of the predicted carbon emission values ​​for all missing sub-time periods as the predicted value of the missing carbon emission value.

[0048] After obtaining the time-series activity level forecast data for each missing sub-time period, the time-series activity level forecast data for all missing sub-time periods are consistent with the fluctuation state of the time-series activity level data in actual production, so that the fluctuation of the activity level data in the entire missing time period is consistent with the statistical law of the activity level data in real production, ensuring that the fluctuation of the activity level data in the entire missing time period is consistent with the activity level fluctuation in real production, that is, the predicted activity level data for the entire missing time period is more consistent with the activity level data in real production.

[0049] After obtaining the time-series activity level forecast data for each missing sub-time period, the carbon emission forecast value for each missing sub-time period can be determined based on the time-series activity level forecast data for each missing sub-time period and the emission factor data corresponding to the activity level forecast data at different time points, thereby realizing the prediction of the carbon emission forecast value for each missing sub-time period.

[0050] When calculating the predicted carbon emissions in the missing sub-time periods, the predicted carbon emissions in each missing sub-time period can be obtained by multiplying the time series activity level prediction data in each missing sub-time period with the emission factor data corresponding to the activity level prediction data at different time points.

[0051] After obtaining the predicted carbon emission value in each missing sub-time period, the sum of the predicted carbon emission values ​​in all missing sub-time periods may be calculated as the predicted value of the missing carbon emission value.

[0052] The beneficial effect brought about by the above-mentioned implementation method is that, according to the activity level fluctuation event information within the missing time period, the missing time period is divided into multiple missing sub-time periods, and the time series activity level prediction data of each missing sub-time period is obtained based on the time series activity level historical data of each missing sub-time period, thereby realizing the prediction of different activity levels of multiple missing sub-time periods according to the activity level fluctuation event information, and improving the accuracy and scientificity of the formulation of carbon emission inventories and carbon emission plans.

[0053] The beneficial effect brought about by the above-mentioned implementation method is that, through the time series activity level prediction model, the time series activity level prediction data is determined according to the time series activity level historical data under similar production conditions, which can simulate the fluctuation of the time series activity level historical data in actual production. When simulating each missing sub-time period, the fluctuation of the time series activity level historical data in actual production can be simulated. It can ensure that the activity level data in the predicted missing time period is consistent with the actual production status, thereby improving the prediction effect of the time series activity level data.

[0054] Figure 3A flowchart of another method for predicting missing data in a carbon emission inventory based on a neural network provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, in the above S110, the historical data of the time series activity level under similar production conditions for each missing sub-time period is obtained, including S111 to S112, and S111 to S112 are specifically described below.

[0055] S111. Obtain time series activity level historical data under similar production conditions for multiple continuous productions corresponding to the first missing sub-time period, and obtain the time length of each time series activity level historical data.

[0056] In order to scientifically determine the historical data of the time series activity level under similar production conditions for each missing sub-time period, we can first obtain the historical data of the time series activity level under similar production conditions for multiple continuous productions corresponding to the first missing sub-time period. The historical data of the time series activity level under similar production conditions for continuous production are the time series activity data without interruption in the historical data of the time series activity level under similar production conditions. This can ensure the continuity of the historical data of the time series activity level, avoid the influence of accidental factors on the historical data of the time series activity level, improve the accuracy of the historical data of the time series activity level, and improve the credibility of the prediction of the time series activity level data.

[0057] Exemplarily, the time-series activity level historical data under similar production conditions of continuous production includes time-series activity level historical data without downtime due to fault maintenance, equipment upgrade or other reasons.

[0058] After obtaining the time series activity level historical data under similar production conditions of multiple continuous productions, the time length of each time series activity level historical data can also be obtained, and then the optimal time series activity level historical data for predicting the time series activity level data can be determined according to the time length of each time series activity level historical data.

[0059] Exemplarily, the first time length of the first time series activity level historical data under similar production conditions of continuous production may be 20 days, 40 days or 60 days.

[0060] S112. Determine, from the time-series activity level historical data under similar production conditions of multiple continuous productions, multiple first time-series activity level historical data with a time length greater than the first missing sub-time period, and intercept the middle segment of the time-series activity level historical data of each first time-series activity level historical data as the time-series activity level historical data under similar production conditions of the first missing sub-time period.

[0061] After obtaining the time series activity level historical data and corresponding time lengths under similar production conditions of multiple continuous productions, multiple first time series activity level historical data with a time length greater than the first missing sub-time period can be determined from the time series activity level historical data under similar production conditions of multiple continuous productions, and further, the time series activity level historical data that is most suitable for predicting the time series activity level data can be determined from the multiple first time series activity level historical data with a time length greater than the first missing sub-time period.

[0062] After obtaining multiple first time series activity level historical data with a time length greater than the first missing sub-time period, the middle segment of the time series activity level historical data of each first time series activity level historical data can be intercepted, that is, the time series activity level historical data in the head time period and the tail time period of each first time series activity level historical data can be intercepted, as the time series activity level historical data under similar production conditions for the first missing sub-time period, thereby avoiding fluctuations in activity level data caused by factors such as unstable production conditions in the head time period and the tail time period of the first time series activity level historical data, and being able to improve the accuracy and scientificity of the time series activity level historical data used to predict activity level data.

[0063] The beneficial effect brought about by the above-mentioned implementation method is that by predicting the time series activity level data based on the historical data of the time series activity level under similar production conditions of continuous production, it is possible to avoid the historical data of the time series activity level being affected by accidental factors, thereby improving the credibility and scientificity of the prediction of the time series activity level data.

[0064] The beneficial effect brought about by the above-mentioned implementation method is that it can avoid the fluctuation of activity level data caused by factors such as unstable production conditions in the head time period and tail time period of the time series activity level historical data, and can predict the accuracy and scientificity of the time series activity level historical data of the activity level data.

[0065] In some implementations, in the above S110, the historical data of the time series activity level under similar production conditions for each missing sub-time period is obtained, and also includes S113 to S114, which are described in detail below.

[0066] S113. When the number of the plurality of first time-series activity level historical data whose time length is greater than the first missing sub-time period is greater than a preset number, obtain the first production condition corresponding to each first time-series activity level historical data.

[0067] When determining the time series activity level historical data used to predict the time series activity level data, when the number of multiple first time series activity level historical data with a time length greater than the first missing sub-time period is greater than a preset number, it means that the number of first time series activity level historical data with a time length greater than the first missing sub-time period is large, that is, there are more available data for predicting the activity level data. In order to improve the accuracy of predicting the activity level data, the first production condition corresponding to each first time series activity level historical data can be obtained, and then each first time series activity level historical data can be screened according to the first production condition corresponding to each first time series activity level historical data, so as to improve the prediction effect of the time series activity level historical data used to predict the activity level data.

[0068] Exemplarily, the first production condition may include production process arrangement characteristics, equipment status characteristics, and raw material type characteristics.

[0069] S114. Through the production condition comparison model, determine the closest first production condition corresponding to the first missing sub-time period according to the first production condition corresponding to each first time-series activity level historical data and the production condition corresponding to the first missing sub-time period, and use the time-series activity level historical data corresponding to the closest first production condition as the time-series activity level historical data of the first missing sub-time period under similar production conditions.

[0070] After obtaining the first production condition corresponding to each first time series activity level historical data, the production condition comparison model can be used to determine the closest first production condition corresponding to the first missing sub-time period according to the first production condition corresponding to each first time series activity level historical data and the production condition corresponding to the first missing sub-time period. The closest first production condition is the production condition that is closest to the production condition corresponding to the first missing sub-time period. The time series activity level historical data corresponding to the closest first production condition can make a more accurate prediction of the activity level in the first missing sub-time period.

[0071] When making activity level predictions, the time-series activity level historical data corresponding to the closest first production condition is used as the time-series activity level historical data under similar production conditions for the first missing sub-time period, so as to avoid the activity level prediction data being affected by the activity level through the time-series activity level historical data corresponding to the closest first production condition.

[0072] Exemplarily, the production condition comparison model is used to determine the first production condition that is closest to the production condition corresponding to the first missing sub-time period among multiple first production conditions. The production condition comparison model can be a classification model that compares production conditions based on production process arrangement characteristics, equipment status characteristics, and raw material type characteristics.

[0073] The beneficial effect brought about by the above-mentioned implementation method is that the most similar first production conditions are obtained by screening multiple first production conditions through production conditions, and the time-series activity level historical data corresponding to the most similar first production conditions are used as the time-series activity level historical data under similar production conditions for the first missing sub-time period. This can further improve the accuracy of activity level prediction data when there is a large amount of available data for predicting activity level data, and improve the accuracy of carbon emissions prediction.

[0074] In some implementations, in the above S110, the historical data of the time series activity level under similar production conditions for each missing sub-time period is obtained, and also includes S115 to S116, which are specifically described below.

[0075] S115. When there is no time series activity level historical data under similar production conditions in the first missing sub-time period, obtain a first impact factor corresponding to the first activity level fluctuation event information before the first missing sub-time period. The first impact factor represents the degree of impact of the first activity level fluctuation event information on the time series activity level historical data corresponding to the first missing sub-time period.

[0076] When obtaining historical activity level data, if there is no historical time series activity level data under similar production conditions in the first missing sub-time period, it means that there is currently no historical time series activity level data suitable for predicting the time series activity level data of the first missing sub-time period, and the activity level data of the first missing sub-time period can be determined by simulation calculation.

[0077] When simulating and calculating to determine the activity level data of the first missing sub-time period, the first influencing factor corresponding to the first activity level fluctuation event information before the first missing sub-time period can be obtained. The first influencing factor represents the degree of influence of the first activity level fluctuation event information on the time-series activity level historical data corresponding to the first missing sub-time period.

[0078] S116. Obtain the time series activity level historical data corresponding to the preceding missing sub-time period before the first activity level fluctuation event information, and take the product of the activity level historical data at each time point of the time series activity level historical data corresponding to the preceding missing sub-time period and the first influencing factor as the time series activity level historical data of the first missing sub-time period under similar production conditions.

[0079] When simulating and calculating to determine the activity level data of the first missing sub-time period, the time-series activity level historical data corresponding to the preceding missing sub-time period before the first activity level fluctuation event information can be obtained. The time-series activity level historical data corresponding to the preceding missing sub-time period is the activity level data before being affected by the first activity level fluctuation event. Furthermore, the activity level data of the first missing sub-time period can be predicted through the activity level data before being affected by the first activity level fluctuation event.

[0080] When determining the activity level data of the first missing sub-time period by simulation calculation, the activity level historical data of each time point of the time-series activity level historical data corresponding to the previous missing sub-time period and the first influencing factor may be multiplied together to obtain the activity level historical data of each time point of the first missing sub-time period obtained by simulation calculation. Furthermore, the activity level historical data of each time point of the first missing sub-time period may be used as the time-series activity level historical data of the first missing sub-time period under similar production conditions, to obtain the time-series activity level historical data under the closest production conditions.

[0081] Exemplarily, the first impact factor corresponding to the first activity level fluctuation event information may be obtained through actual measurement or determined based on an empirical value.

[0082] Exemplarily, when the first activity level fluctuation event information is fault maintenance event information or production shutdown event information, the first impact factor corresponding to the first activity level fluctuation event information is less than 1.

[0083] Exemplarily, when the first activity level fluctuation event information is capacity adjustment event information, the first impact factor corresponding to the first activity level fluctuation event information is greater than 1.

[0084] The beneficial effect brought about by the above-mentioned implementation method is that when the first missing sub-time period does not have the time-series activity level historical data under similar production conditions, the time-series activity level historical data of the first missing sub-time period is simulated and calculated by multiplying the activity level historical data at each time point of the time-series activity level historical data corresponding to the previous missing sub-time period and the first influencing factor, thereby ensuring the scientific nature of the simulation calculation of the activity level data of the first missing sub-time period and improving the effect of predicting the activity level data.

[0085] In some implementations, in the above S110, the historical data of the time series activity level under similar production conditions for each missing sub-time period is obtained, and also includes S117 to S118, which are described in detail below.

[0086] S117 . When the first impact factor changes with time, obtain a time-series first impact factor corresponding to the first activity level fluctuation event information before the first missing sub-time period.

[0087] When activity level fluctuation event information will have a continuous impact on the activity level, for example, when the activity level fluctuation event information is activity level fluctuation event information such as adjusting production capacity, repairing equipment, etc., the impact of the first activity level fluctuation event information on the time-series activity level historical data corresponding to the first missing sub-time period varies with time.

[0088] Exemplarily, the influence of the first activity level fluctuation event information on the time-series activity level historical data corresponding to the first missing sub-time period may fluctuate up and down relative to the median value over time.

[0089] When the first impact factor changes with time, the time series first impact factor corresponding to the first activity level fluctuation event information before the first missing sub-time period may be obtained, and the time series first impact factor includes multiple first impact factors in time series.

[0090] S118. Calculate the product of the time series activity level historical data and the first impact factor corresponding to the previous missing sub-time period according to the time correspondence, and use it as the time series activity level historical data under similar production conditions for the first missing sub-time period.

[0091] After obtaining the first time series influencing factor, the time series activity level historical data and the time series first influencing factor corresponding to the previous missing sub-time period can be used to calculate the product of the activity level historical data and the first influencing factor according to the time correspondence, and the activity level data corresponding to multiple time points of the first missing sub-time period can be obtained, which can be used as the time series activity level historical data of the first missing sub-time period under similar production conditions, thereby improving the accuracy of determining the time series activity level historical data.

[0092] The beneficial effect brought about by the above-mentioned implementation method is that when the degree of influence of the first activity level fluctuation event information on the time series activity level historical data corresponding to the first missing sub-time period changes over time, the time series first influence factor can be obtained, and the time series activity level historical data can be determined based on the time series first influence factor, thereby improving the accuracy of determining the time series activity level historical data.

[0093] In some implementations, the activity level fluctuation event information also includes technology improvement event information. When the activity level fluctuation event information is technology improvement event information, the above method further includes: obtaining first emission factor data corresponding to the first activity level fluctuation event information before the first missing sub-time period. The first emission factor data is the emission factor data after the first activity level fluctuation event information.

[0094] When predicting carbon emissions, enterprises and institutions may also undergo technological improvements in their production equipment. Technological improvements will cause changes in the emission factors of the production process. Therefore, it is necessary to predict carbon emissions when the emission factors of the production process change.

[0095] When predicting carbon emissions, the activity level fluctuation event information may be technology improvement event information. The technology improvement event information corresponds to a change in the value of the emission factor. It is necessary to predict the carbon emissions corresponding to the technology improvement event information in combination with the changed emission factor.

[0096] When the activity level fluctuation event information is technology improvement event information, and it is necessary to predict the carbon emissions when the emission factor changes, the above method also includes: obtaining first emission factor data corresponding to the first activity level fluctuation event information before the first missing sub-time period, the first emission factor data is the emission factor data after the first activity level fluctuation event information, and then the carbon emissions can be predicted in combination with the emission factor data after the first activity level fluctuation event information.

[0097] Exemplarily, the emission factor data following the first activity level fluctuation event information corresponding to the first activity level fluctuation event information may be obtained through experimental detection, or may be determined through empirical values.

[0098] Exemplarily, the emission factor data following the first activity level fluctuation event information may be the first emission factor.

[0099] In some implementations, in the above-mentioned S130, the carbon emission prediction value for each missing sub-time period is determined based on the emission factor data corresponding to the time-series activity level prediction data in each missing sub-time period and the activity level prediction data at different time points, including: determining the carbon emission prediction value for the first missing sub-time period based on the first emission factor data corresponding to the time-series activity level prediction data in the first missing sub-time period and the activity level prediction data at different time points.

[0100] After obtaining the emission factor data after the first activity level fluctuation event information, the carbon emission prediction value for the first missing sub-time period can be determined based on the first emission factor data corresponding to the time series activity level prediction data in the first missing sub-time period and the activity level prediction data at different time points, thereby achieving accurate prediction of carbon emissions based on the changed emission factor.

[0101] Exemplarily, the activity level prediction data at each time point in the time-series activity level prediction data within the first missing sub-time period and the first emission factor data corresponding to the activity level prediction data can be multiplied together to form the time-series carbon emission prediction value within the first missing sub-time period.

[0102] Exemplarily, the first emission factor data corresponding to the activity level prediction data at different time points may all be the first emission factor.

[0103] The beneficial effect brought about by the above-mentioned implementation method is that when the activity level fluctuation event information is technology improvement event information, when predicting the carbon emissions when the emission factor changes, it is possible to combine the time series activity level prediction data and the first emission factor data corresponding to the time series activity level prediction data to accurately predict the carbon emissions in the missing sub-time period, thereby improving the prediction accuracy of carbon emissions.

[0104] An embodiment of the present application also provides a device for predicting missing data in a carbon emission inventory based on a neural network, comprising a unit for executing any of the methods described above.

[0105] Figure 4 A schematic diagram of the logical structure of a device for predicting missing data in a carbon emission inventory based on a neural network according to an embodiment of the present application is shown in FIG. Figure 4 As shown, the device 1 of this embodiment includes a processing unit 11, a storage unit 12 and a transceiver unit 13. The processing unit 11 is used to process data, the storage unit 12 is used to store data, and the transceiver unit 13 is used to send and receive data. The processing unit 11, the storage unit 12 and the transceiver unit 13 cooperate with each other to implement the above method. The beneficial effects brought about by the embodiment of the present application have been described in the above method and will not be repeated here.

[0106] An embodiment of the present application also provides a carbon emission inventory missing data prediction device based on a neural network, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the methods described above when executing the computer program.

[0107] Figure 5 A schematic diagram of the physical structure of a device for predicting missing data in a carbon emission inventory based on a neural network according to an embodiment of the present application is shown in FIG. Figure 5 As shown, the device 2 of this embodiment includes: at least one processor 20 ( Figure 5 Only one processor 20 is shown in the figure), a memory 21, and a computer program 22 stored in the memory 21 and executable on the at least one processor 20. When the processor 20 executes the computer program 22, the steps in any of the above-mentioned method embodiments are implemented. The beneficial effects brought about by the embodiments of the present application have been described in the above-mentioned methods and will not be repeated here.

[0108] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0109] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0110] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0111] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0112] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0113] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0114] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0115] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0116] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0117] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for predicting missing data in carbon emission inventory based on neural network, characterized in that: The method comprises: Determine the missing time period corresponding to the missing carbon emission values ​​that need to be completed in the carbon emission inventory, divide the missing time period into multiple missing sub-time periods according to the activity level fluctuation event information within the missing time period, and obtain the time series activity level historical data of each missing sub-time period under similar production conditions; wherein the activity level fluctuation event information includes fault maintenance event information, production shutdown event information and capacity adjustment event information; Determine the forecast data of the time series activity level in each missing sub-time period according to the historical data of the time series activity level under similar production conditions in each missing sub-time period through the time series activity level forecast model; Determine the predicted carbon emission value for each missing sub-time period based on the time series activity level prediction data for each missing sub-time period and the emission factor data corresponding to the activity level prediction data at different time points, and use the sum of the predicted carbon emission values ​​for all missing sub-time periods as the predicted value of the missing carbon emission value; Obtain historical data on time-series activity levels under similar production conditions for each missing sub-period, including: Obtaining time series activity level historical data under similar production conditions for a plurality of continuous productions corresponding to the first missing sub-time period, and obtaining the time length of each time series activity level historical data; Determine, from among multiple continuous production time series activity level history data under similar production conditions, multiple first time series activity level history data with a time length greater than the first missing sub-time period, and intercept the middle segment time series activity level history data of each first time series activity level history data as the time series activity level history data under similar production conditions of the first missing sub-time period; When the number of the plurality of first time-series activity level history data whose time length is greater than the first missing sub-time period is greater than a preset number, obtaining the first production condition corresponding to each first time-series activity level history data; wherein the first production condition includes production process arrangement characteristics, equipment status characteristics and raw material type characteristics; Through the production condition comparison model, the closest first production condition corresponding to the first missing sub-time period is determined according to the first production condition corresponding to each first time-series activity level historical data and the production conditions corresponding to the first missing sub-time period, and the time-series activity level historical data corresponding to the closest first production condition is used as the time-series activity level historical data of the first missing sub-time period under similar production conditions.

2. The method according to claim 1, characterized in that Obtain historical data on time-series activity levels under similar production conditions for each missing sub-period, including: When there is no historical data of time-series activity levels under similar production conditions in the first missing sub-time period, obtaining a first impact factor corresponding to the first activity level fluctuation event information before the first missing sub-time period; wherein the first impact factor represents the degree of influence of the first activity level fluctuation event information on the historical data of time-series activity levels corresponding to the first missing sub-time period; Obtain the time series activity level historical data corresponding to the preceding missing sub-time period before the first activity level fluctuation event information, and take the product of the activity level historical data at each time point of the time series activity level historical data corresponding to the preceding missing sub-time period and the first influencing factor as the time series activity level historical data of the first missing sub-time period under similar production conditions.

3. The method according to claim 2, characterized in that Obtain historical data on time-series activity levels under similar production conditions for each missing sub-period, including: When the first impact factor changes with time, obtaining a time-series first impact factor corresponding to the first activity level fluctuation event information before the first missing sub-time period; The product of the historical data of the time series activity level and the first influencing factor corresponding to the previous missing sub-time period is calculated according to the time correspondence, and is used as the historical data of the time series activity level under similar production conditions for the first missing sub-time period.

4. The method according to claim 3, characterized in that The activity level fluctuation event information also includes technology improvement event information. When the activity level fluctuation event information is technology improvement event information, the method further includes: Obtaining first emission factor data corresponding to the first activity level fluctuation event information before the first missing sub-time period; wherein the first emission factor data is the emission factor data after the first activity level fluctuation event information; Based on the time series activity level forecast data in each missing sub-time period and the emission factor data corresponding to the activity level forecast data at different time points, the carbon emission forecast value in each missing sub-time period is determined, including: The predicted carbon emission value for the first missing sub-time period is determined based on the time-series activity level prediction data for the first missing sub-time period and the first emission factor data corresponding to the activity level prediction data at different time points.

5. A device for predicting missing data in carbon emission inventory based on neural network, characterized in that: The method comprises means for performing the method according to any one of claims 1 to 4.

6. A device for predicting missing data in a carbon emission inventory based on a neural network, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images