A carbon emission dynamic prediction method and system based on multi-energy load data

Abstract

A carbon emission dynamic prediction method and system based on multi-energy load data, the method comprising: collecting and preprocessing the historical data of carbon emission, energy consumption and multi-energy load consumption of the park, calculating the joint correlation coefficient and screening the energy type; calculating the initial carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix; constructing a carbon emission coupling objective function, solving and updating the corresponding conversion matrix by alternately fixing the carbon emission-energy and energy-multi-energy load conversion matrix; establishing a carbon emission coupling model, calculating the prediction carbon emission error, constructing a prediction error correction objective function, and correcting the conversion matrix; predicting the multi-energy load consumption in the M+1 month and correcting the error, combining the corrected conversion matrix in S4, and predicting the carbon emission in the M+1 month. The present application can dynamically adjust the conversion relationship between carbon emission and multi-energy load, significantly improve the accuracy of carbon emission prediction and the response ability to actual working condition changes.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of carbon emissions, and relates to a carbon emission dynamic prediction method and system based on multi-energy load data. BACKGROUND

[0002] Industrial parks, science and technology parks and comprehensive economic development zones and other park energy-intensive enterprises are concentrated, the energy structure is complex and the total carbon emission is large, which is the key area of carbon emission management and optimization. At present, some parks introduce carbon emission prediction methods for carbon emission monitoring, but most carbon emission prediction methods still stay in the historical statistics or static evaluation stage, and it is difficult to cope with the changes in carbon emission trends caused by the adjustment of industrial structure and the dynamic changes of energy use in the park. Therefore, how to accurately predict the carbon emission based on the relevant data of the park is of great significance for realizing the fine management of carbon emission in the park and formulating differentiated emission reduction paths.

[0003] At present, in the aspect of carbon emission accounting, the patent with publication number CN120069335A realizes intelligent analysis of carbon emission data based on checking, abnormal detection, prediction modeling and knowledge base question answering; the patent with publication number CN120069487A provides a smart park power carbon emission prediction method based on artificial intelligence, which calculates the purchased power carbon emission, direct power carbon emission and total carbon emission according to the power carbon emission accounting standard by acquiring the data of purchased power activity and direct power consumption in the smart park; and the Chinese patent CN118071164A discloses a carbon emission reduction potential estimation method based on unit key indicator carbon emission change prediction, which expands the STIRPAT model by using data such as park greening area and non-fossil energy proportion, and constructs a mathematical model combined with the ridge regression algorithm, and then uses BP neural network to predict the target year data, which can estimate the carbon emission reduction potential of each unit; in addition, the patent with publication number CN119005405A provides a dynamic carbon emission factor prediction method based on data driving, which uses the labels of training samples generated by the improved dynamic carbon emission factor calculation method, adopts historical data, simulation power flow and Fea2Vec-BiLSTM model, and can accurately estimate the carbon emission reduction potential and optimize the carbon emission reduction strategy of the power grid.

[0004] In summary, the existing carbon emission accounting and prediction methods mostly rely on life cycle assessment, and it is difficult to dynamically adjust in combination with real-time load changes, and the prediction accuracy of future carbon emission trends is low. At the same time, the conversion coefficient in the existing method is fixed, and it lacks the response ability to actual working condition changes. The existing carbon emission prediction methods mostly focus on the calculation and prediction of power carbon emission, and lack dynamic consideration of multi-energy system conversion. SUMMARY

[0005] In order to solve the problems of difficulty in combining real-time load change dynamic prediction, lack of dynamic correction of multi-energy system conversion and the like in the prior art, the present application provides a carbon emission dynamic prediction method and system based on multi-energy load data, which comprises the following steps: collecting and preprocessing carbon emission, energy consumption and multi-energy load consumption historical data of a park, calculating joint correlation coefficients and screening energy types; calculating initial carbon emission-energy conversion matrix and energy-multi-load conversion matrix; constructing a carbon emission coupling objective function, solving and updating the corresponding conversion matrix by alternately fixing the carbon emission-energy and energy-multi-energy load conversion matrix; establishing a carbon emission coupling model, calculating the prediction carbon emission error, constructing a prediction error correction objective function and correcting the conversion matrix; predicting the multi-energy load consumption in the M+1 month and correcting the error, and predicting the carbon emission in the M+1 month in combination with the corrected conversion matrix of S4. The present application can dynamically adjust the conversion relationship between carbon emission and multi-energy load, and significantly improve the accuracy of carbon emission prediction and the response ability to actual working condition changes.

[0006] The present application adopts the following technical solutions.

[0007] The present application provides a carbon emission dynamic prediction method based on multi-energy load data, which comprises the following steps:

[0008] S1, collecting and preprocessing carbon emission, energy consumption and multi-energy load consumption historical data of a park, calculating joint correlation coefficients between the preprocessed multi-energy load consumption and energy consumption, and screening energy types;

[0009] S2, based on the screened energy types, the joint correlation coefficients between the corresponding energy types and different types of load consumption, calculating initial carbon emission-energy conversion matrix and energy-multi-load conversion matrix;

[0010] S3, based on the initial conversion matrix, taking minimizing the average error of carbon emission in the previous M-1 months as the target, constructing a carbon emission coupling objective function; solving and updating the conversion matrix by alternately fixing different conversion matrices;

[0011] S4, establishing a carbon emission coupling model, predicting the carbon emission in the M month based on the conversion matrix obtained in S3; calculating the prediction carbon emission error, constructing a prediction error correction objective function and correcting the conversion matrix;

[0012] S5, obtaining historical temperature data and multi-energy load consumption historical data, predicting the multi-energy load consumption in the M+1 month and correcting the error, and predicting the carbon emission in the M+1 month in combination with the corrected conversion matrix.

[0013] Preferably, in S1, the Pearson correlation coefficient and Spearman correlation coefficient between different types of load consumption and different types of energy consumption are calculated, and the joint correlation coefficient is obtained by weighted fusion.

[0014] When the joint correlation coefficient falls within a predetermined interval, the corresponding energy type is the energy type that is strongly correlated with the consumption of multi-energy loads.

[0015] Preferably, in S2, an initial carbon emission-energy conversion matrix is ​​calculated based on the carbon emission factor and the correlation coefficients between different energy types and carbon emissions;

[0016] Based on the joint correlation coefficients of selected energy types and the energy efficiency conversion coefficients of different energy types, the initial energy-multi-energy load conversion matrix of the park is calculated.

[0017] Preferably, the specific calculation formula for the initial energy-multi-energy load conversion matrix is ​​as follows:

[0018] ;

[0019] In the formula, Represents the energy-multi-energy load conversion matrix, where l To select the number of energy categories that are strongly correlated with multi-energy loads. K Number of load types; For the energy-multi-energy load conversion matrix, the first j Line 1 k The values ​​in the column; For the first Energy efficiency conversion coefficient of energy types; For different energy types; For the first Type of energy and the first k Joint correlation coefficient of load consumption.

[0020] Preferably, the specific formula for the carbon emission coupling objective function in S3 is as follows:

[0021] ;

[0022] In the formula, This represents the park's carbon emissions in month t. This is the basic energy matrix, where n is the number of energy types; and These are the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix, respectively. Indicates the first t Monthly multi-energy load consumption matrix; This represents the L2 norm function.

[0023] Preferably, in S3, the initial carbon emission-energy conversion matrix is fixed, the energy-multiple energy load conversion matrix is solved, the initial energy-multiple energy load conversion matrix is fixed, the carbon emission-energy conversion matrix is solved, the carbon emission-energy conversion matrix and the energy-multiple energy load conversion matrix are updated, the corresponding carbon emission coupling objective function value is calculated, and the updating is stopped until the objective function value is less than a predetermined error threshold, and the final carbon emission-energy conversion matrix and the energy-multiple energy load conversion matrix are output.

[0024] Preferably, in S4, a carbon emission coupling model is established based on the carbon emission-energy conversion relationship and the energy-multiple energy load conversion relationship; and the carbon emission amount of the Mth month is predicted through the carbon emission coupling model based on the final carbon emission-energy conversion matrix and the energy-multiple energy load conversion matrix of S3.

[0025] Preferably, in S4, a carbon emission coupling model is established based on the carbon emission-energy conversion relationship and the energy-multiple energy load conversion relationship; and the carbon emission amount of the Mth month is predicted through the carbon emission coupling model based on the final carbon emission-energy conversion matrix and the energy-multiple energy load conversion matrix of S3.

[0026] Preferably, historical temperature data are collected and preprocessed, the preprocessed historical temperature data and the multiple energy load consumption historical data are subjected to multivariate phase space reconstruction to obtain a high-dimensional feature matrix; and the multiple energy load consumption matrix of the M+1th month is predicted based on the high-dimensional feature matrix.

[0027] The difference between the multiple energy load consumption historical data of the first M months and the predicted multiple energy load consumption corresponding to the time is calculated to obtain the prediction error of the first M months, and the prediction error of the M+1th month is predicted through the prediction model to correct the predicted multiple energy load consumption matrix of the M+1th month.

[0028] Another aspect of the present application provides a carbon emission dynamic prediction system based on multiple energy load data, comprising:

[0029] A data acquisition and processing module collects and preprocesses the carbon emission amount, energy consumption amount and multiple energy load consumption historical data of a park, calculates the joint correlation coefficient between the preprocessed multiple energy load consumption and energy consumption, and screens energy types.

[0030] An initial conversion matrix calculation module calculates the initial carbon emission-energy conversion matrix and the energy-multiple load conversion matrix based on the screened energy types and the joint correlation coefficient between the corresponding energy types and different types of load consumption.

[0031] The conversion matrix updating module, based on the initial conversion matrix, constructs a carbon emission coupling target function with the objective of minimizing the average error of carbon emissions in the previous M-1 months; and solves and updates the conversion matrix by alternately fixing different conversion matrices;

[0032] The conversion matrix correction module establishes a carbon emission coupling model, predicts the carbon emissions in the Mth month based on the conversion matrix obtained by S3, calculates the prediction error of carbon emissions, constructs a prediction error correction target function, and corrects the conversion matrix;

[0033] The carbon emission prediction module obtains historical temperature data and historical multi-energy load consumption data, predicts the multi-energy load consumption in the M+1th month and corrects the error, and predicts the carbon emissions in the M+1th month in combination with the corrected conversion matrix.

[0034] Compared with the prior art, the beneficial effects of the present application at least include:

[0035] 1、The present application introduces a joint correlation coefficient analysis method to realize the quantification and screening of the relationship between multi-energy load and energy consumption in historical data, ensuring that the data dimension of the input model is reasonable and strongly related, thereby improving the effectiveness and computational efficiency of the overall prediction.

[0036] 2、The present application constructs a coupling conversion matrix between carbon emissions-energy and energy-load, and dynamically updates the matrix during model training, so that the model can continuously optimize the carbon emission mapping relationship according to historical and real-time data. Compared with the traditional fixed conversion coefficient carbon emission prediction method, the present application can more accurately reflect the changes of actual working conditions, effectively improving the adaptability of carbon emission prediction under multi-energy load.

[0037] 3、The present application updates the carbon emission-energy conversion matrix and the energy-load conversion matrix in each cycle (such as month) through an alternating optimization and error correction mechanism, and introduces historical prediction error as a feedback item for adaptive correction, effectively reducing the model prediction error. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a flowchart of a carbon emission dynamic prediction method based on multi-energy load data provided by an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions will be described clearly and completely below in conjunction with the accompanying drawings of the embodiments of the present application. The embodiments described in the present application are only a part of the embodiments of the present application, but not all the embodiments. Based on the spirit of the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0040] Embodiment 1

[0041] The present embodiment provides a carbon emission dynamic prediction method based on multi-energy load data, as shown in the following specific steps: Figure 1

[0042] S1, collect the carbon emission, energy consumption and multi-energy load consumption historical data of the park and preprocess, calculate the joint correlation coefficient between the preprocessed multi-energy load consumption and different types of energy consumption, and screen the energy types; the specific process is as follows:

[0043] The Pearson correlation coefficient and the Spearman correlation coefficient between different types of load consumption and different types of energy consumption are calculated, and the joint correlation coefficient is obtained by weighted fusion; the specific calculation formula is as follows:

[0044] ;

[0045] In the formula, is the joint correlation coefficient; is the Pearson correlation coefficient; is the Spearman correlation coefficient;

[0046] When , it indicates that the corresponding energy type and load consumption are strongly correlated, and the energy consumption of the energy type strongly correlated with the multi-energy load consumption is screened out.

[0047] S2, based on the screened energy type, the joint correlation coefficient between the corresponding energy type and different types of load consumption, the initial carbon emission-energy conversion matrix and the energy-multi-load conversion matrix are calculated.

[0048] Since the product of chemical reaction is proportional to energy absorption or release, there is a significant linear relationship between industrial carbon emission and energy consumption, so in the present application, based on the carbon emission factor, the correlation coefficient between different energy types and carbon emission, the initial carbon emission-energy conversion matrix is calculated; the specific calculation formula is as follows:

[0049] ;

[0050] In the formula, is the carbon emission-energy conversion matrix, wherein n ​The number of energy types distinguishes between different types of fossil fuels and electricity; For the first Carbon emission factors from fossil fuels; For different fossil fuels; For the first Energy consumption of fossil fuels; For the first The average lower heating value of fossil fuels; For the first Carbon content per unit calorific value of a fossil fuel; For the first The carbon oxidation rate of fossil fuels; 3.67 is the gasification coefficient of CO2; For the first The correlation coefficient between fossil fuels and carbon emissions can be calculated based on industry standards; Carbon emissions from fossil fuels; Carbon emission factor for electricity; Carbon emissions from purchasing electricity from outside the industrial park; This refers to the amount of electricity consumed through external purchases. The correlation coefficient between electricity and carbon emissions can be calculated according to industry standards.

[0051] Based on the joint correlation coefficients of selected energy types and the energy efficiency conversion coefficients of different energy types, the initial energy-multi-energy load conversion matrix of the park is calculated; the specific calculation formula is as follows:

[0052] ;

[0053] In the formula, Represents the energy-multi-energy load conversion matrix, where l To select the number of energy categories that are strongly correlated with multi-energy loads. K As for the number of load types, this embodiment only considers heating, cooling, electrical, and load loads, i.e., K=4; For the energy-multi-energy load conversion matrix, the first j Line number k The values ​​in the column; For the first Energy efficiency conversion coefficient of energy types; For different energy types; For the first Type of energy and the first k The joint correlation coefficient of load consumption.

[0054] S3. Based on the initial transformation matrix, construct a carbon emission coupled objective function with the goal of minimizing the average error of carbon emission prediction for the first M-1 months; solve and update the corresponding transformation matrix by alternately fixing the carbon emission-energy and energy-multi-energy load transformation matrices.

[0055] Based on the carbon emissions and multi-energy load consumption in the first M-1 months, and the initial carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix, a carbon emission coupled objective function is constructed; the specific formula is as follows:

[0056] ;

[0057] In the formula, This represents the carbon emissions of the park in month t. The basic energy matrix represents fixed energy consumption that is independent of real-time load, such as the basic energy consumption of equipment standby and the fixed energy consumption of park infrastructure. and These are the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix, respectively. Indicates the first t Monthly multi-energy load consumption matrix; Represents a L2 norm function;

[0058] By fixing the initial carbon emission-energy conversion matrix, the energy-multi-energy load conversion matrix is ​​solved, and the carbon emission-energy conversion matrix is ​​solved. The carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix are updated, and the corresponding carbon emission coupling objective function value is recalculated. The updated carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix are solved again. The updating process continues until the objective function value is less than a predetermined error threshold, at which point the update stops, and the final carbon emission-energy conversion matrix is ​​output. Energy-Multi-Energy Load Conversion Matrix That is, the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix obtained through the first M-1 months of training.

[0059] S4. Establish a carbon emission coupling model. Based on the transformation matrix obtained in S3, predict the carbon emission amount in month M. Calculate the predicted carbon emission error. Combine the multi-energy load consumption matrix in month M to construct the prediction error correction objective function and correct the carbon emission-energy transformation matrix and the energy-multi-energy load transformation matrix.

[0060] Furthermore, S4 includes:

[0061] Based on the carbon emission-energy conversion relationship and the energy-multi-energy load conversion relationship, a carbon emission coupling model is established. Based on the final carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix of S3, the carbon emission amount for month M is predicted through the carbon emission coupling model. The specific formula is as follows:

[0062] ;

[0063] In the formula, the predicted carbon emission of the Mth month; the carbon emission-energy conversion matrix obtained by training the previous M-1 months; the energy-multi-energy load conversion matrix obtained by training the previous M-1 months; the multi-energy load consumption matrix of the Mth month in the historical data;

[0064] Based on the predicted carbon emission of the Mth month and the actual carbon emission of the Mth month, a predicted carbon emission error is calculated; according to the predicted carbon emission error, the energy-multi-energy load conversion matrix, the multi-energy load consumption matrix of the Mth month and the basic energy consumption matrix, a predicted error correction target function is constructed, and a carbon emission-energy conversion correction matrix is solved; the specific formula is as follows:

[0065] ;

[0066] In the formula, the carbon emission-energy conversion correction matrix of the predicted error correction; the predicted carbon emission error; F represents the F norm;

[0067] Based on the carbon emission-energy conversion correction matrix, the carbon emission-energy conversion matrix is updated as the carbon emission-energy conversion matrix for carbon emission prediction of the next month, and the specific formula is as follows:

[0068] ;

[0069] In the formula, the carbon emission-energy conversion matrix after the predicted error correction; according to the predicted carbon emission error, the updated carbon emission-energy conversion matrix, the multi-energy load consumption matrix of the Mth month and the basic energy consumption matrix, an energy-multi-energy load conversion correction matrix is solved through the predicted error correction target function, and the energy-multi-energy load conversion matrix is updated; the specific formula is as follows:

[0070] ;

[0071] ;

[0072] In the formula, the energy-multi-energy load conversion matrix after the predicted error correction;

[0073] The corrected carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix are used as the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix for predicting the park carbon emission of the M+1th month.

[0074] S5, acquire historical temperature data and multi-energy load consumption history data, predict the multi-energy load consumption of the M+1 month and correct the error, combine the corrected conversion matrix of S4, and predict the carbon emission of the M+1 month through the carbon emission coupling model.

[0075] Collect historical temperature data and preprocess, input the preprocessed historical temperature data and multi-energy load consumption history data into the multivariate phase space reconstruction to obtain a high-dimensional feature matrix; input the high-dimensional feature matrix into the gated recurrent unit network to predict the multi-energy load consumption of the M+1 month; in this embodiment, the daily average temperature of each month is taken as the temperature data of the corresponding month;

[0076] Calculate the difference between the multi-energy load consumption history data of the previous M months and the predicted multi-energy load consumption at the corresponding time to obtain the predicted load error of the previous M months, and input it into the LSTM network to predict the predicted load error of the M+1 month; based on the predicted load error of the M+1 month and the multi-energy load consumption of the M+1 month, update the predicted multi-energy load consumption matrix of the M+1 month;

[0077] Based on the multi-energy load consumption matrix, the updated carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix of S4, and the basic energy matrix, the carbon emission of the M+1 month is predicted through the carbon emission coupling model.

[0078] Embodiment 2

[0079] The embodiment provides a carbon emission dynamic prediction system based on multi-energy load data, which comprises:

[0080] The data acquisition and processing module acquires the carbon emission, energy consumption and multi-energy load consumption history data of the park and preprocesses them, calculates the joint correlation coefficient between the preprocessed multi-energy load consumption and energy consumption, and screens the energy types;

[0081] The initial conversion matrix calculation module calculates the initial carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix based on the screened energy types, the joint correlation coefficient between the corresponding energy types and different types of load consumption;

[0082] The conversion matrix updating module, based on the initial conversion matrix, takes minimizing the average error of the previous M-1 month carbon emission prediction as the target, constructs a carbon emission coupling objective function; by alternately fixing the carbon emission-energy and energy-multi-energy load conversion matrix, the corresponding conversion matrix is solved and updated;

[0083] The conversion matrix correction module establishes a carbon emission coupling model, predicts the carbon emission in the Mth month based on the conversion matrix obtained by S3, calculates the prediction carbon emission error, combines the multi-energy load consumption matrix in the Mth month, constructs a prediction error correction objective function, and corrects the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix;

[0084] The carbon emission prediction module obtains historical temperature data and multi-energy load consumption historical data, predicts the multi-energy load consumption in the M+1th month and corrects the error, combines the conversion matrix corrected by S4, and predicts the carbon emission in the M+1th month through the carbon emission coupling model.

[0085] The present disclosure can be a system, a method, and / or a computer program product. The computer program product can include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present disclosure.

[0086] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium can be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or punched-tape, a holographic storage medium, or any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0087] Computer readable program instructions described herein can be downloaded to respective computing / processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and / or a wireless network. The network can comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing / processing device.

[0088] Computer readable program instructions for carrying out operations of the present disclosure can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages such as the "C" programming language or similar programming languages. The computer readable program instructions can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate array (FPGA), or programmable logic array (PLA) can execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

[0089] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting the present application. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modification or equivalent replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A method for dynamic prediction of carbon emissions based on multi-energy load data, characterized in that, The method comprises the following steps: S1, collecting and preprocessing the historical data of carbon emissions, energy consumption and multi-energy load consumption of the park, calculating the joint correlation coefficient between the preprocessed multi-energy load consumption and energy consumption, and screening the energy types; the specific process is as follows: Pearson correlation coefficient and Spearman correlation coefficient between different types of load consumption and different types of energy consumption are calculated, and a joint correlation coefficient is obtained by weighted fusion; when the joint correlation coefficient belongs to a predetermined interval, the corresponding energy type is an energy type that is strongly correlated with the multi-energy load consumption; S2, based on the screened energy type, the joint correlation coefficient of the corresponding energy type and different types of load consumption, the initial carbon emission-energy conversion matrix and the energy-multi-load conversion matrix are calculated; wherein the specific calculation formula of the initial energy-multi-energy load conversion matrix is: ； In the formula, represents the energy-multiple energy load conversion matrix, wherein l is the number of energy categories screened out and strongly correlated with the multiple energy load, K is the number of load types; is the value of the energy-multiple energy load conversion matrix in the j row and the k column; is the energy efficiency conversion coefficient of the energy category; is the different energy type; is the joint correlation coefficient of the energy category and the k load consumption. S3, based on the initial conversion matrix, the carbon emission coupling objective function is constructed by minimizing the average error of carbon emissions in the previous M-1 months; by alternately fixing different conversion matrices, the conversion matrix is solved and updated; wherein the specific formula of the carbon emission coupling objective function is: ； wherein, represents the park carbon emission in the tth month; is the base energy matrix, n is the number of energy types; and are the carbon emission-energy conversion matrix and the energy-multiple energy load conversion matrix, respectively; represents the multiple energy load consumption matrix in the tth month; t represents the multiple energy load consumption matrix in the tth month; represents the two-norm function; S4, a carbon emission coupling model is established, and based on the conversion matrix obtained in S3, the carbon emissions in the Mth month are predicted; the prediction error of carbon emissions is calculated, a prediction error correction objective function is constructed, and the conversion matrix is corrected; S5, the historical temperature data and multi-energy load consumption historical data are obtained, the multi-energy load consumption in the M+1th month is predicted and the error is corrected, and the carbon emissions in the M+1th month are predicted combined with the corrected conversion matrix.

2. The carbon emission dynamic prediction method based on multi-energy load data according to claim 1, wherein: In S2, based on the carbon emission factor and the correlation coefficient of different energy types and carbon emissions, the initial carbon emission-energy conversion matrix is calculated; Based on the joint correlation coefficient of the screened energy type and the energy efficiency conversion coefficient of different types of energy, the initial energy-multi-energy load conversion matrix of the park is calculated.

3. The carbon emission dynamic prediction method based on multi-energy load data according to claim 1, wherein: In S3, by fixing the initial carbon emission-energy conversion matrix, the energy-multi-energy load conversion matrix is solved, and the initial energy-multi-energy load conversion matrix is fixed to solve the carbon emission-energy conversion matrix; the carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix are updated, the corresponding carbon emission coupling objective function value is calculated, and the updating is stopped until the objective function value is less than a predetermined error threshold, and the final carbon emission-energy conversion matrix and energy-multi-energy load conversion matrix are output.

4. The carbon emission dynamic prediction method based on multi-energy load data according to claim 1, wherein: In S4, based on the carbon emission-energy conversion relationship and the energy-multi-energy load conversion relationship, a carbon emission coupling model is established; based on the final carbon emission-energy conversion matrix and the energy-multi-energy load conversion matrix in S3, the carbon emissions in the Mth month are predicted through the carbon emission coupling model.

5. The carbon emission dynamic prediction method based on multi-energy load data according to claim 1 or 4, wherein: In S4, a predicted carbon emission error is calculated based on the predicted Mth month carbon emission and the actual Mth month carbon emission; a predicted error correction objective function is constructed according to the predicted carbon emission error, the energy-multiple energy load conversion matrix, the Mth month multiple energy load consumption matrix and the basic energy consumption matrix; a carbon emission-energy conversion correction matrix is solved by fixing the energy-multiple energy load conversion matrix, and the carbon emission-energy conversion matrix is updated; the energy-multiple energy load conversion correction matrix is solved by fixing the updated carbon emission-energy conversion matrix, and the energy-multiple energy load conversion matrix is updated.

6. The carbon emission dynamic prediction method based on multiple energy load data according to claim 1, characterized in that: Collect historical temperature data and pre-process, reconstruct the pre-processed historical temperature data and multiple energy load consumption historical data through multivariate phase space, and obtain a high-dimensional feature matrix; Based on the high-dimensional feature matrix, the M+1 month multiple energy load consumption matrix is predicted; The difference between the M-1 month multiple energy load consumption historical data and the corresponding time predicted multiple energy load consumption is calculated to obtain the M-1 month predicted load error, and the M+1 month predicted error is predicted through the prediction model to correct the predicted M+1 month multiple energy load consumption matrix.

7. A dynamic carbon emission prediction system based on multi-energy load data, performing the method according to any one of claims 1 to 6, characterized in that, It includes: A data acquisition and processing module acquires and pre-processes the carbon emission, energy consumption and multiple energy load consumption historical data of the park, calculates the joint correlation coefficient between the pre-processed multiple energy load consumption and energy consumption, and selects the energy type; An initial conversion matrix calculation module calculates the initial carbon emission-energy conversion matrix and energy-multiple load conversion matrix based on the selected energy type, the joint correlation coefficient between the corresponding energy type and different types of load consumption; A conversion matrix updating module constructs a carbon emission coupling objective function based on the initial conversion matrix to minimize the average error of the M-1 month carbon emission; the conversion matrix is solved and updated by alternately fixing different conversion matrices; A conversion matrix correction module establishes a carbon emission coupling model, predicts the Mth month carbon emission based on the conversion matrix obtained in S3, calculates the predicted carbon emission error, constructs a predicted error correction objective function, and corrects the conversion matrix; A carbon emission prediction module acquires historical temperature data and multiple energy load consumption historical data, predicts the M+1 month multiple energy load consumption and corrects the error, and predicts the M+1 month carbon emission combined with the corrected conversion matrix.

Images