CN113657661A - Enterprise carbon emission prediction method and device, computer equipment and storage medium - Google Patents

Enterprise carbon emission prediction method and device, computer equipment and storage medium Download PDF

Info

Publication number
CN113657661A
CN113657661A CN202110923844.3A CN202110923844A CN113657661A CN 113657661 A CN113657661 A CN 113657661A CN 202110923844 A CN202110923844 A CN 202110923844A CN 113657661 A CN113657661 A CN 113657661A
Authority
CN
China
Prior art keywords
data
carbon emission
enterprise
analyzed
historical
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110923844.3A
Other languages
Chinese (zh)
Inventor
周珑
浦迪
曾梦妤
奚建飞
殷梓恒
张茹
穆文杰
黎灿兵
陈浩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southern Power Grid Digital Grid Research Institute Co Ltd
Original Assignee
Southern Power Grid Digital Grid Research Institute Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southern Power Grid Digital Grid Research Institute Co Ltd filed Critical Southern Power Grid Digital Grid Research Institute Co Ltd
Priority to CN202110923844.3A priority Critical patent/CN113657661A/en
Publication of CN113657661A publication Critical patent/CN113657661A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/80—Management or planning
    • Y02P90/84—Greenhouse gas [GHG] management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Tourism & Hospitality (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Public Health (AREA)
  • Evolutionary Biology (AREA)
  • Primary Health Care (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Water Supply & Treatment (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application relates to an enterprise carbon emission prediction method, an enterprise carbon emission prediction device, computer equipment and a storage medium. The method comprises the following steps: acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and power consumption of the target enterprise; performing empirical mode decomposition on data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence; respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences; and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result. By adopting the method, the carbon emission prediction precision can be improved.

Description

Enterprise carbon emission prediction method and device, computer equipment and storage medium
Technical Field
The present application relates to the field of data processing technologies, and in particular, to a method and an apparatus for predicting carbon emissions of an enterprise, a computer device, and a storage medium.
Background
Reducing carbon emission is the core important meaning of 'carbon neutralization' and the key direction of force for realizing the target, and realizing accurate measurement and comprehensive monitoring of carbon emission is the primary task of reducing carbon emission. Although the energy supply side is an important area for carbon emission generation, the production thereof serves for consumption, and the energy consumption side directly or indirectly obtains economic and environmental benefits from high energy consumption and high emission of the energy supply side. Based on the method, in order to balance the supply side responsibility and the consumption side responsibility of the carbon emission and avoid the unfair problem caused by the transfer of the carbon emission responsibility, the accurate measurement and the comprehensive monitoring of the carbon emission at the energy consumption side are enhanced, and a credible data support is provided for a 'double-carbon' target.
Currently, the carbon emission estimation method includes a model estimation method, which means that the carbon emission is predicted by a constructed analysis model. Although the model estimation method can effectively analyze and predict the carbon emission, the method needs factors such as comprehensive policy and economy, is more suitable for carbon emission prediction at a macroscopic level, and has larger measurement and calculation errors at the industrial and enterprise levels.
Disclosure of Invention
In view of the above, it is necessary to provide a method, an apparatus, a computer device and a storage medium for predicting carbon emissions of an enterprise, which can improve the accuracy of carbon emissions prediction.
A method of enterprise carbon emission prediction, the method comprising:
acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and a power consumption of the target enterprise;
performing empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
An enterprise carbon emissions prediction apparatus, the apparatus comprising an acquisition module, a processing module, a prediction module, and an output module, wherein:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring data to be analyzed, and the data to be analyzed comprises at least one of a total domestic production value of everyone, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and power consumption of the target enterprise;
the processing module is used for carrying out empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
the prediction module is used for respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and the output module is used for carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and a power consumption of the target enterprise;
performing empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and a power consumption of the target enterprise;
performing empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
The enterprise carbon emission prediction method, the enterprise carbon emission prediction device, the computer equipment and the storage medium predict the enterprise carbon emission at the energy consumption side based on the empirical mode decomposition and the carbon emission prediction model, and effectively reduce errors caused by a single model. On one hand, the data to be analyzed is decomposed based on empirical mode decomposition, and the stationary signals and the non-stationary signals of the data to be analyzed are separated under the condition that manual setting and intervention are not needed, so that the data have higher analysis value. On the other hand, a power consumption-carbon emission association mechanism is established by analyzing historical carbon emission sources, compositions and characteristics of various industries, and a power consumption-based carbon emission total amount measurement model is established based on the association mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, carbon emission data generated by primary energy consumption are fused, accurate measuring and calculating of the carbon emission total amount of each enterprise are achieved, and carbon emission prediction accuracy is improved.
Drawings
FIG. 1 is a diagram of an environment in which a method for predicting carbon emissions from an enterprise may be implemented, according to one embodiment;
FIG. 2 is a schematic flow chart diagram of a method for enterprise carbon emissions prediction in one embodiment;
FIG. 3 is a block diagram of an overall computing framework for an enterprise carbon emissions prediction method in one embodiment;
FIG. 4 is a flowchart illustrating the empirical mode decomposition step performed on data to be analyzed according to one embodiment;
FIG. 5 is a block diagram of a carbon emissions prediction model in one embodiment;
FIG. 6 is a block diagram of an enterprise carbon emissions prediction unit in accordance with one embodiment;
FIG. 7 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The enterprise carbon emission prediction method provided by the application can be applied to the application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The server 104 acquires data to be analyzed transmitted via the terminal 102, wherein the data to be analyzed includes at least one of a total per capita domestic production value, energy intensity, an industry added value to which the target enterprise belongs, an industry energy consumption total amount to which the target enterprise belongs, a business income of the target enterprise, a profit of the target enterprise business, and a power consumption of the target enterprise. Server 104 performs empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences of different frequencies and a residual component sequence. The server 104 inputs a plurality of sequences formed by the subsequences and the residual component sequence into the carbon emission prediction model, and obtains sequence prediction results corresponding to the sequences. And the server 104 performs sequence reconstruction on the obtained sequence prediction results to obtain a reconstruction result, and determines an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices, and the server 104 may be implemented by an independent server or a server cluster formed by a plurality of servers.
In one embodiment, as shown in fig. 2, an enterprise carbon emission prediction method is provided, which is illustrated by applying the method to the server in fig. 1, and includes the following steps:
step S202, obtaining data to be analyzed, wherein the data to be analyzed comprises at least one of a total domestic production value of everyone, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and a power consumption of the target enterprise.
Specifically, acquiring the data to be analyzed includes: acquiring input data, and screening abnormal data from the input data based on abnormal data screening rules according to values correspondingly acquired by the input data at each moment; determining a first numerical value corresponding to the abnormal data at the last moment and a second numerical value corresponding to the abnormal data at the next moment; when the first numerical value and the second numerical value both meet the normal data screening rule, determining a substitute numerical value based on the first numerical value and the second numerical value, and substituting the abnormal data by the substitute numerical value; and determining the data to be analyzed according to the input data after the substitution processing.
In one embodiment, the abnormal data filtering rule may further be understood that when the value obtained at a certain time is obviously not matched with the values obtained at other times, the value obtained at the certain time is considered to be an abnormal value, for example, 2005-2019, the power consumption of the enterprise is basically 150 ten thousand watts, whereas in 2018, the power consumption of the enterprise is only 5 ten thousand watts, and at this time, the power consumption of the enterprise in 2018 may be considered to be abnormal data, and the abnormal data needs to be deleted from the input data to avoid affecting subsequent carbon emission prediction. It should be noted that the normal data screening rule can be understood according to the above-described embodiment, and this is not limited in the examples of the present application.
In one embodiment, on one hand, when the server performs screening of the abnormal data based on the input data, the server may extract the numerical values corresponding to the corresponding time from the input data in order according to a time sequence, and each time the numerical value is obtained, determine whether the numerical value is the abnormal data based on the abnormal data screening rule, and of course, in this embodiment, the server may also perform screening of the abnormal data based on a plurality of numerical values obtained many times, which is not considered in the present embodiment. On the other hand, the server determines a substitute value based on the average value of the first value and the second value, and substitutes the abnormal data by the substitute value, wherein when the substitute value is determined to be substituted successfully, the corresponding abnormal data can be considered to be successfully converted into normal data, and at the moment, the conversion of the normal data is completed. Of course, the server may also determine the alternative value based on a weighted average of the first value and the second value, which is not limited in this embodiment.
Step S204, empirical mode decomposition is carried out on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence.
The empirical mode decomposition is a novel self-adaptive signal time-frequency processing method for signal decomposition according to the time scale characteristics of data, and is suitable for analysis and processing of nonlinear non-stationary signals. The essence of the method is to decompose a non-stationary signal into a finite number of eigenmode functions, and this decomposition is based on the local characteristics of the time scale of the original signal, so that the individual frequency components contain the local characteristic signals of different time scales of the original signal.
Specifically, when the server performs empirical mode decomposition on data to be analyzed, the server needs to consider the following two aspects: 1. the number of zero crossing points and the number of extreme points differ by at most one; 2. the average of the upper envelope formed by connecting the local maximum points and the lower envelope formed by connecting the local minimum points is zero, that is, both are symmetrical with respect to the time axis. I.e. in case both of the above-mentioned conditions are fulfilled, the server may resolve the corresponding sub-sequence and residual component sequence from the data to be analyzed.
Therefore, based on the empirical mode decomposition method, the stationary signal and the non-stationary signal of the original data are separated, so that the data has higher analysis value, and the prediction accuracy can be effectively improved.
Step S206, a plurality of sequences formed by the subsequences and the residual component sequence are respectively input into the carbon emission prediction model, and sequence prediction results corresponding to the sequences are obtained.
Specifically, the carbon emission of the enterprise refers to carbon emission generated in the process of consuming energy by the enterprise, and comprises carbon dioxide, methane, nitrous oxide, hydrofluorocarbon, perfluorocarbon, sulfur hexafluoride and the like. These types of energy consumption include primary energy consumption (e.g., raw coal, oil, natural gas) and secondary energy consumption (e.g., heat, electricity). The method for calculating the carbon emission of the enterprise adopts an emission factor method, namely the product of the energy consumption activity level and the emission factor.
In one embodiment, in consideration of the fact that fuel combustion, production processes and the like are different in production processes of various enterprises, not only can comprehensive data not be acquired, but also accurate emission factor values cannot be obtained for various fuels mixed with combustion and complex production processes. Therefore, in the current embodiment, in order to accurately measure and calculate the carbon emission of the enterprise, the server establishes a correlation mechanism of power consumption-carbon emission by analyzing historical carbon emission sources, compositions and characteristics of various industries on the theoretical basis that the energy utilization structures of the enterprises in the same industry have similarity, and constructs a carbon emission total measurement model based on the power consumption based on the correlation mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, and carbon emission data generated by primary energy consumption is fused to realize measuring and calculating of the carbon emission total amount of each enterprise.
And S208, carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
Specifically, the prediction results of each sequence of the server are superposed to obtain a corresponding reconstruction result. Wherein, referring to fig. 3, if the number of subsequences is n, each subsequence can be understood as the IMF illustrated in fig. 31～IMFnThe residual component sequence can be understood as R illustrated in fig. 3nWherein, IMF1～IMFnAnd RnInputting the sub-sequences into a DBN (Deep Belief Network) Network respectively, and obtaining corresponding sub-sequence prediction values through the processing of the DBN Network to obtain each sequence prediction result; finally, the server carries out superposition reconstruction on each obtained subsequence prediction value to obtain final predictionThe result is the reconstructed result. In one embodiment, the server will determine a business carbon emissions prediction corresponding to the target business based on the reformulation. It should be noted that the DBN network is a deep model in which feature learning and a classifier are combined in a training process, features are automatically extracted from sample data through model training, and conventional and complex signal processing is not required, so that interference and influence of human factors on feature extraction are avoided, and complexity of feature selection and uncertainty of a diagnosis result are solved.
In the enterprise carbon emission prediction method, the enterprise-level carbon emission of the energy consumption side is predicted based on the empirical mode decomposition and the carbon emission prediction model, so that errors caused by a single model are effectively reduced. On one hand, the data to be analyzed is decomposed based on empirical mode decomposition, and the stationary signals and the non-stationary signals of the data to be analyzed are separated under the condition that manual setting and intervention are not needed, so that the data have higher analysis value. On the other hand, a power consumption-carbon emission association mechanism is established by analyzing historical carbon emission sources, compositions and characteristics of various industries, and a power consumption-based carbon emission total amount measurement model is established based on the association mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, carbon emission data generated by primary energy consumption are fused, accurate measuring and calculating of the carbon emission total amount of each enterprise are achieved, and carbon emission prediction accuracy is improved.
In one embodiment, determining data to be analyzed from the input data after the substitution process includes: normalizing the input data after the substitution processing to obtain corresponding normalized data; and taking the corresponding normalized data as the data to be analyzed.
Specifically, the server normalizes the input data after the substitution processing according to the following formula so as to constrain the value range of the input data after the substitution processing to be 0-1:
Figure BDA0003208450820000071
wherein, X' is normalized data obtained after processing, X is input data after substitution processing, mu is a preset mean value, and sigma is a preset variance. It should be noted that the normalization process is a way to simplify the calculation, i.e. to convert the original series of data into a certain range, or a certain form.
In the embodiment, the input data after the substitution processing is subjected to the normalization processing, so that the data are mapped to the range of 0-1 for processing, the data processing is facilitated, and the data processing speed is increased.
In one embodiment, referring to fig. 4, performing an empirical mode decomposition on data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence, includes:
step S402, taking the data to be analyzed as an original input signal, and acquiring a local maximum point and a local minimum point corresponding to the original input signal.
Specifically, the server takes the data to be analyzed as an original input signal, and obtains a local maximum point and a local minimum point corresponding to the original input signal. The calculation of the local maximum value point and the local minimum value point may refer to the following modes: first, for sequence x1,x2,x3,...,xi-1,xi,xi+1,...,xn(i 2, 3.., n-1), t will be calculatedi-1＝xi-xi-1,ti＝xi+1-xi. Thereafter, when t is determinedi-1*ti<＝0,ti-1>When 0, then x is considerediFor maximum point, when t is determinedi-1*ti<＝0,ti-1<When 0, then x is considerediIs a minimum point. Of course, in the current embodiment, the local maximum point and the local minimum point may also be calculated in other manners, which is not limited in the embodiment of the present application.
Step S404, determining a corresponding initial frequency component based on the local maximum point, the local minimum point and the original input signal.
Specifically, determining a corresponding initial frequency component based on the local maximum point, the local minimum point and the original input signal includes: determining a corresponding upper envelope line based on the local maximum value point; determining a corresponding lower envelope line based on the local minimum value point; an average value between the upper and lower envelope is calculated and the corresponding initial frequency component is determined based on the difference between the original input signal and the average value.
In one embodiment, the server will determine the corresponding initial frequency component based on:
(1) after the server determines the corresponding local maximum point and local minimum point, all the local maximum points are connected to form the corresponding upper envelope E1(t) and connecting all local minima points to form a corresponding lower envelope E2(t) of (d). In one embodiment, the server calculates the average m between the upper and lower envelopes based on the following formula1(t)：
m1(t)＝[E2(t)+E1(t)]/2。(2)
(2) The server is based on combining the original input signal x (t) and the mean value m1The difference h between (t)1(t), the initial frequency component is calculated, and the specific calculation formula can refer to the following formula (3), it should be noted that, in the embodiment of the present application, x (t) and m1The calculation formula of the difference between (t) is not limited, and in different application scenarios, the adaptive adjustment may be performed based on the following formula (3):
h1(t)＝x(t)-m1(t)。(3)
in step S406, when it is determined that the initial frequency component satisfies the preset signal decomposition condition, the initial frequency component is taken as a decomposed subsequence.
Wherein, the preset signal decomposition condition refers to: 1. the number of zero crossing points and the number of extreme points differ by at most one; 2. the average of the upper envelope formed by connecting the local maximum points and the lower envelope formed by connecting the local minimum points is zero, that is, both are symmetrical with respect to the time axis. Initial frequency component obtained based on equation (3)h1(t) if the 2-item signal decomposition condition is satisfied, the initial frequency component h is divided into two1(t) as a first frequency component c1(t), which represents the highest frequency component in the original input signal x (t).
Specifically, the method further comprises: in each iteration cycle process, when the initial frequency component is determined not to meet the preset signal decomposition condition, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal to continue executing until the corresponding subsequence is decomposed, and then calculating the updating signal.
In one embodiment, the initial frequency component h is determined1(t) if the signal decomposition condition is not satisfied, the initial frequency component h is divided into1(t) repeating the steps of calculating the upper and lower envelope means and the difference between the original input signal and the upper and lower envelope means as the input signal, wherein the initial frequency component h is obtained if it is in the corresponding repeated calculating step1(t) has a mean value of the upper and lower envelope lines of m11(t), then the corresponding difference h needs to be further judged11(t)＝h1(t)-m11(t) whether a signal decomposition condition can be satisfied; if h11(t) if the signal decomposition condition is not satisfied, repeating the above step k times until h is obtained1k(t)＝h1(k-1)(t)-m1k(t) (k is 1,2 …, N) satisfies the signal decomposition condition, the corresponding subsequence is decomposed.
Step S408, separating the decomposed subsequence from the original input signal to obtain a corresponding update signal.
Specifically, the server may separate the decomposed subsequence from the original input signal based on a difference between the original input signal and the decomposed subsequence, and obtain an updated signal from which the high frequency component is removed.
And step S410, when entering the next iteration cycle, taking the updating signal as the original input signal in the next iteration cycle, and returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal to continue the execution until the decomposition of all the subsequences and the residual component sequence is completed, and ending the iteration cycle.
Specifically, the server sends an update signal r1(t) as the original input signal, and returning to the step of obtaining the local maximum point and the local minimum point corresponding to the original input signal to continue the execution, at this time, the second frequency component c will be obtained2(t) of (d). After repeating the above steps n times, n frequency components are obtained correspondingly, and the calculation formula of each component may specifically refer to the following formula (4):
Figure BDA0003208450820000101
in one embodiment, when the server determines the nth update signal rnAnd (t), when the function is a monotone function, ending the loop, and outputting each subsequence obtained by decomposition and the residual component sequence. Wherein, the following formula (4) can be obtained:
Figure BDA0003208450820000102
in the formula (5), the residual function rn(t) represents the average trend of the signal. And each frequency component c1(t),c2(t)....cnAnd (t) respectively containing components of different time characteristic scales of the signal, wherein the time scales of the components are sequentially from small to large. Therefore, each component correspondingly contains components of different frequency bands from high to low, and the frequency component contained in each frequency band is different and can change along with the change of the signal.
In the above embodiment, for an unknown signal, decomposition can be directly started based on the empirical mode decomposition method without performing preliminary analysis and study. In the process of being applicable to the empirical mode decomposition mode, the method can automatically well classify the stable signals and the non-stable signals of the original data according to certain fixed modes without manual setting and intervention, so that the data has higher analysis value, and the prediction precision is improved.
In one embodiment, the method further comprises: acquiring historical data to be analyzed of a sample, and determining reference carbon emission data of the sample which is adaptive to the historical data to be analyzed of the sample; inputting the historical data to be analyzed of the sample into a carbon emission prediction model to be trained, processing the historical data to be analyzed of the sample through the carbon emission prediction model to be trained, and outputting a corresponding prediction result; and adjusting model parameters of the carbon emission prediction model to be trained according to the difference between the prediction result and the sample reference carbon emission data, and stopping training until the training stopping condition is reached to obtain the trained carbon emission prediction model.
Specifically, referring to fig. 5, the DBN network is a deep learning model formed by stacking a plurality of RBM networks (Restricted Boltzmann machines), and as shown in fig. 4, the DBN network adopted in the embodiment of the present application is a neural network model formed by stacking three RBM networks (i.e., RBM1, RBM2, and RBM 3). Wherein, the neural network model comprises a visual layer v and three hidden layers (h)1、h2、h3) And a label layer. The learning process is mainly divided into two stages: firstly, carrying out unsupervised layer-by-layer pre-training by adopting a limited Boltzmann machine, effectively excavating the characteristics implied by data, and when needing to be explained, adopting a greedy layer-by-layer training algorithm at the stage, which can also be called pre-training, wherein only one RBM network is trained each time, the output of the trained RBM network is used as the input of a higher RBM network, and then the next RBM network is trained until the training of all RBM networks is completed; thus, by means of the RBM layer-by-layer training method, the implicit characteristics of the original data can be extracted, and the deep expression of the characteristics is obtained, so that more abstract, more characterization capability and more sensitive characteristics to the sample label are formed. And then, a back propagation algorithm is adopted, label data of the sample is combined to be supervised and optimized, and a complex nonlinear relation between the characteristics and the state label is established, so that the performance of the deep belief network is optimized, and the classification function of the deep belief network is realized.
Specifically, the step of determining sample reference carbon emission data corresponding to the sample historical data to be analyzed includes: calculating an industry historical carbon emission proportion through the following formula (the industry historical carbon emission proportion refers to the ratio of the carbon emission generated by the industry historical electricity to the industry historical total carbon emission):
Figure BDA0003208450820000111
wherein E isA,total＝EA,electricity+EA,other，EA,electricity＝CA,electricity*αA,powergrid；RA,electricityCarbon proportion for industry history of industry A, EA,totalFor historical total carbon emissions of industry A, EA,electricityHistorical carbon emissions for industry-wide electricity usage in industry A, EA,otherHistorical carbon emissions for other energy consumptions in industry A, CA,electricityFor the industry-wide historical power consumption, alpha, of the A industryA,powergridThe average carbon dioxide emission factor of the power grid in the historical region of the area where the industry A is located.
In one embodiment, the server calculates the historical carbon emission generated by the power consumption of the enterprise (the historical carbon emission generated by the power consumption of the enterprise refers to the historical net purchased electricity consumption of the enterprise, and the generated carbon dioxide emission) by the following formula:
Ex,electricity＝Cx,electricity*αx,powergrid；
wherein E isx,electricityHistorical carbon emissions, C, for x enterprise electricity usagex,electricityNet purchase of electricity usage, alpha, for x enterprisesx,powergridAverage carbon dioxide emission factors of power grids in historical regions of x enterprises;
in one embodiment, the server will base the calculated RA,electricityAnd Ex,electricityCalculating the historical total carbon emission of the enterprise by the following formula, and taking the calculated historical total carbon emission of the enterprise as sample reference carbon emission data:
Figure BDA0003208450820000121
in the embodiment, a power consumption-carbon emission association mechanism is established, a carbon emission total amount measurement model based on power consumption is established, and carbon emission data generated by primary energy consumption is fused on the basis of industrial-level and enterprise-level power big data, so that the measurement and calculation of the carbon emission total amount of each enterprise are realized, and the prediction precision is improved.
It should be understood that although the steps in the flowcharts of fig. 2 and 4 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2 and 4 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps.
In one embodiment, as shown in fig. 6, an enterprise carbon emissions prediction apparatus 600 is provided, comprising: an obtaining module 601, a processing module 602, a predicting module 603 and an outputting module 604, wherein:
the obtaining module 601 is configured to obtain data to be analyzed, where the data to be analyzed includes at least one of a total domestic production value, an energy intensity, an industry added value to which a target enterprise belongs, an industry energy consumption total amount to which the target enterprise belongs, a business income of the target enterprise, a profit of the target enterprise, and a power consumption of the target enterprise.
The processing module 602 is configured to perform empirical mode decomposition on data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence.
And the predicting module 603 is configured to input a plurality of sequences formed by each subsequence and the residual component sequence into the carbon emission prediction model, respectively, to obtain sequence prediction results corresponding to each sequence, respectively.
And the output module 604 is configured to perform sequence reconstruction on the obtained sequence prediction results to obtain a reconstruction result, and determine an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
In one embodiment, the obtaining module 601 is further configured to obtain input data, and screen abnormal data from the input data based on an abnormal data screening rule according to a value that is obtained by the input data at each time; determining a first numerical value corresponding to the abnormal data at the last moment and a second numerical value corresponding to the abnormal data at the next moment; when the first numerical value and the second numerical value both meet the normal data screening rule, determining a substitute numerical value based on the first numerical value and the second numerical value, and substituting the abnormal data by the substitute numerical value; and determining the data to be analyzed according to the input data after the substitution processing.
In one embodiment, the obtaining module 601 is further configured to perform normalization processing on the input data after the substitution processing to obtain corresponding normalized data; and taking the corresponding normalized data as the data to be analyzed.
In one embodiment, the processing module 602 is further configured to use data to be analyzed as an original input signal, and obtain a local maximum point and a local minimum point corresponding to the original input signal; determining a corresponding initial frequency component based on the local maximum value point, the local minimum value point and the original input signal; when the initial frequency component is determined to meet the preset signal decomposition condition, taking the initial frequency component as a decomposed subsequence; separating the decomposed subsequence from the original input signal to obtain a corresponding updating signal; and when entering the next iteration cycle, taking the updating signal as the original input signal in the next iteration cycle, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal and continuing to execute until the decomposition of all the subsequences and the residual component sequence is completed, and finishing the iteration cycle.
In one embodiment, the processing module 602 is further configured to determine a corresponding upper envelope line based on the local maximum value point; determining a corresponding lower envelope line based on the local minimum value point; an average value between the upper and lower envelope is calculated and the corresponding initial frequency component is determined based on the difference between the original input signal and the average value.
In one embodiment, the processing module 602 is further configured to, in each iteration cycle, when it is determined that the initial frequency component does not satisfy the preset signal decomposition condition, return to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal and continue to be executed until the corresponding subsequence is decomposed, and then perform calculation on the update signal.
In one embodiment, the apparatus further comprises a model training module, wherein:
the model training module is used for acquiring the historical data to be analyzed of the sample and determining the reference carbon emission data of the sample which is adaptive to the historical data to be analyzed of the sample; inputting the historical data to be analyzed of the sample into a carbon emission prediction model to be trained, processing the historical data to be analyzed of the sample through the carbon emission prediction model to be trained, and outputting a corresponding prediction result; and adjusting model parameters of the carbon emission prediction model to be trained according to the difference between the prediction result and the sample reference carbon emission data, and stopping training until the training stopping condition is reached to obtain the trained carbon emission prediction model.
In one embodiment, the model training module is further configured to calculate the carbon emission proportion for industry history according to the following formula:
Figure BDA0003208450820000141
wherein E isA,total＝EA,electricity+EA,other，EA,electricity＝CA,electricity*αA,powergrid；RA,electricityCarbon proportion for industry history of industry A, EA,totalAs a history of industry ACarbon emission, EA,electricityHistorical carbon emissions for industry-wide electricity usage in industry A, EA,otherHistorical carbon emissions for other energy consumptions in industry A, CA,electricityFor the industry-wide historical power consumption, alpha, of the A industryA,powergridAverage carbon dioxide emission factors of a power grid in a historical region of an area where the industry A is located;
calculating the historical carbon emission generated by the power utilization of the enterprise according to the following formula:
Ex,electricity＝Cx,electricity*αx,powergrid；
wherein E isx,electricityHistorical carbon emissions, C, for x enterprise electricity usagex,electricityNet purchase of electricity usage, alpha, for x enterprisesx,powergridAverage carbon dioxide emission factors of power grids in historical regions of x enterprises;
based on calculated RA,electricityAnd Ex,electricityCalculating the historical total carbon emission of the enterprise by the following formula, and taking the calculated historical total carbon emission of the enterprise as sample reference carbon emission data:
Figure BDA0003208450820000142
the enterprise carbon emission prediction device predicts the enterprise carbon emission on the energy consumption side based on the empirical mode decomposition and the carbon emission prediction model, and effectively reduces errors caused by a single model. On one hand, the data to be analyzed is decomposed based on empirical mode decomposition, and the stationary signals and the non-stationary signals of the data to be analyzed are separated under the condition that manual setting and intervention are not needed, so that the data have higher analysis value. On the other hand, a power consumption-carbon emission association mechanism is established by analyzing historical carbon emission sources, compositions and characteristics of various industries, and a power consumption-based carbon emission total amount measurement model is established based on the association mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, carbon emission data generated by primary energy consumption are fused, accurate measuring and calculating of the carbon emission total amount of each enterprise are achieved, and carbon emission prediction accuracy is improved.
For specific limitations of the enterprise carbon emission prediction device, reference may be made to the above limitations of the enterprise carbon emission prediction method, which are not described herein again. The various modules of the enterprise carbon emissions prediction unit described above may be implemented in whole or in part by software, hardware, and combinations thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing data to be analyzed and enterprise carbon emission prediction results. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of enterprise carbon emissions prediction.
Those skilled in the art will appreciate that the architecture shown in fig. 7 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program: acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and power consumption of the target enterprise; performing empirical mode decomposition on data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence; respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences; and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
In one embodiment, the processor, when executing the computer program, further performs the steps of: acquiring input data, and screening abnormal data from the input data based on abnormal data screening rules according to values correspondingly acquired by the input data at each moment; determining a first numerical value corresponding to the abnormal data at the last moment and a second numerical value corresponding to the abnormal data at the next moment; when the first numerical value and the second numerical value both meet the normal data screening rule, determining a substitute numerical value based on the first numerical value and the second numerical value, and substituting the abnormal data by the substitute numerical value; and determining the data to be analyzed according to the input data after the substitution processing.
In one embodiment, the processor, when executing the computer program, further performs the steps of: normalizing the input data after the substitution processing to obtain corresponding normalized data; and taking the corresponding normalized data as the data to be analyzed.
In one embodiment, the processor, when executing the computer program, further performs the steps of: taking data to be analyzed as an original input signal, and acquiring a local maximum value point and a local minimum value point corresponding to the original input signal; determining a corresponding initial frequency component based on the local maximum value point, the local minimum value point and the original input signal; when the initial frequency component is determined to meet the preset signal decomposition condition, taking the initial frequency component as a decomposed subsequence; separating the decomposed subsequence from the original input signal to obtain a corresponding updating signal; and when entering the next iteration cycle, taking the updating signal as the original input signal in the next iteration cycle, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal and continuing to execute until the decomposition of all the subsequences and the residual component sequence is completed, and finishing the iteration cycle.
In one embodiment, the processor, when executing the computer program, further performs the steps of: determining a corresponding upper envelope line based on the local maximum value point; determining a corresponding lower envelope line based on the local minimum value point; an average value between the upper and lower envelope is calculated and the corresponding initial frequency component is determined based on the difference between the original input signal and the average value.
In one embodiment, the processor, when executing the computer program, further performs the steps of: in each iteration cycle process, when the initial frequency component is determined not to meet the preset signal decomposition condition, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal to continue executing until the corresponding subsequence is decomposed, and then calculating the updating signal.
In one embodiment, the processor, when executing the computer program, further performs the steps of: acquiring historical data to be analyzed of a sample, and determining reference carbon emission data of the sample which is adaptive to the historical data to be analyzed of the sample; inputting the historical data to be analyzed of the sample into a carbon emission prediction model to be trained, processing the historical data to be analyzed of the sample through the carbon emission prediction model to be trained, and outputting a corresponding prediction result; and adjusting model parameters of the carbon emission prediction model to be trained according to the difference between the prediction result and the sample reference carbon emission data, and stopping training until the training stopping condition is reached to obtain the trained carbon emission prediction model.
In one embodiment, the processor, when executing the computer program, further performs the steps of: calculating the carbon emission proportion for industry history according to the following formula:
Figure BDA0003208450820000171
wherein E isA,total＝EA,electricity+EA,other，EA,electricity＝CA,electricity*αA,powergrid；RA,electricityCarbon proportion for industry history of industry A, EA,totalFor historical total carbon emissions of industry A, EA,electricityHistorical carbon emissions for industry-wide electricity usage in industry A, EA,otherHistorical carbon emissions for other energy consumptions in industry A, CA,electricityFor the industry-wide historical power consumption, alpha, of the A industryA,powergridAverage carbon dioxide emission factors of a power grid in a historical region of an area where the industry A is located;
calculating the historical carbon emission generated by the power utilization of the enterprise according to the following formula:
Ex,electricity＝Cx,electricity*αx,powergrid；
wherein E isx,electricityHistorical carbon emissions, C, for x enterprise electricity usagex,electricityNet purchase of electricity usage, alpha, for x enterprisesx,powergridAverage carbon dioxide emission factors of power grids in historical regions of x enterprises;
based on calculated RA,electricityAnd Ex,electricityCalculating the historical total carbon emission of the enterprise by the following formula, and taking the calculated historical total carbon emission of the enterprise as sample reference carbon emission data:
Figure BDA0003208450820000172
the computer equipment predicts the enterprise-level carbon emission of the energy consumption side based on the empirical mode decomposition and the carbon emission prediction model, and effectively reduces errors caused by a single model. On one hand, the data to be analyzed is decomposed based on empirical mode decomposition, and the stationary signals and the non-stationary signals of the data to be analyzed are separated under the condition that manual setting and intervention are not needed, so that the data have higher analysis value. On the other hand, a power consumption-carbon emission association mechanism is established by analyzing historical carbon emission sources, compositions and characteristics of various industries, and a power consumption-based carbon emission total amount measurement model is established based on the association mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, carbon emission data generated by primary energy consumption are fused, accurate measuring and calculating of the carbon emission total amount of each enterprise are achieved, and carbon emission prediction accuracy is improved.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of: acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and power consumption of the target enterprise; performing empirical mode decomposition on data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence; respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences; and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring input data, and screening abnormal data from the input data based on abnormal data screening rules according to values correspondingly acquired by the input data at each moment; determining a first numerical value corresponding to the abnormal data at the last moment and a second numerical value corresponding to the abnormal data at the next moment; when the first numerical value and the second numerical value both meet the normal data screening rule, determining a substitute numerical value based on the first numerical value and the second numerical value, and substituting the abnormal data by the substitute numerical value; and determining the data to be analyzed according to the input data after the substitution processing.
In one embodiment, the computer program when executed by the processor further performs the steps of: normalizing the input data after the substitution processing to obtain corresponding normalized data; and taking the corresponding normalized data as the data to be analyzed.
In one embodiment, the computer program when executed by the processor further performs the steps of: taking data to be analyzed as an original input signal, and acquiring a local maximum value point and a local minimum value point corresponding to the original input signal; determining a corresponding initial frequency component based on the local maximum value point, the local minimum value point and the original input signal; when the initial frequency component is determined to meet the preset signal decomposition condition, taking the initial frequency component as a decomposed subsequence; separating the decomposed subsequence from the original input signal to obtain a corresponding updating signal; and when entering the next iteration cycle, taking the updating signal as the original input signal in the next iteration cycle, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal and continuing to execute until the decomposition of all the subsequences and the residual component sequence is completed, and finishing the iteration cycle.
In one embodiment, the computer program when executed by the processor further performs the steps of: determining a corresponding upper envelope line based on the local maximum value point; determining a corresponding lower envelope line based on the local minimum value point; an average value between the upper and lower envelope is calculated and the corresponding initial frequency component is determined based on the difference between the original input signal and the average value.
In one embodiment, the computer program when executed by the processor further performs the steps of: in each iteration cycle process, when the initial frequency component is determined not to meet the preset signal decomposition condition, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal to continue executing until the corresponding subsequence is decomposed, and then calculating the updating signal.
In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring historical data to be analyzed of a sample, and determining reference carbon emission data of the sample which is adaptive to the historical data to be analyzed of the sample; inputting the historical data to be analyzed of the sample into a carbon emission prediction model to be trained, processing the historical data to be analyzed of the sample through the carbon emission prediction model to be trained, and outputting a corresponding prediction result; and adjusting model parameters of the carbon emission prediction model to be trained according to the difference between the prediction result and the sample reference carbon emission data, and stopping training until the training stopping condition is reached to obtain the trained carbon emission prediction model.
In one embodiment, the computer program when executed by the processor further performs the steps of: calculating the carbon emission proportion for industry history according to the following formula:
Figure BDA0003208450820000191
wherein E isA,total＝EA,electricity+EA,other，EA,electricity＝CA,electricity*αA,powergrid；RA,electricityCarbon proportion for industry history of industry A, EA,totalFor historical total carbon emissions of industry A, EA,electricityHistorical carbon emissions for industry-wide electricity usage in industry A, EA,otherHistorical carbon emissions for other energy consumptions in industry A, CA,electricityFor the industry-wide historical power consumption, alpha, of the A industryA,powergridAverage carbon dioxide emission factors of a power grid in a historical region of an area where the industry A is located;
calculating the historical carbon emission generated by the power utilization of the enterprise according to the following formula:
Ex,electricity＝Cx,electricity*αx,powergrid；
wherein E isx,electricityHistorical carbon emissions, C, for x enterprise electricity usagex,electricityNet purchase of electricity usage, alpha, for x enterprisesx,powergridIs a company place of xAverage carbon dioxide emission factor of the power grid in historical regions of the region;
based on calculated RA,electricityAnd Ex,electricityCalculating the historical total carbon emission of the enterprise by the following formula, and taking the calculated historical total carbon emission of the enterprise as sample reference carbon emission data:
Figure BDA0003208450820000201
the storage medium predicts the enterprise-level carbon emission of the energy consumption side based on the empirical mode decomposition and the carbon emission prediction model, and effectively reduces errors caused by a single model. On one hand, the data to be analyzed is decomposed based on empirical mode decomposition, and the stationary signals and the non-stationary signals of the data to be analyzed are separated under the condition that manual setting and intervention are not needed, so that the data have higher analysis value. On the other hand, a power consumption-carbon emission association mechanism is established by analyzing historical carbon emission sources, compositions and characteristics of various industries, and a power consumption-based carbon emission total amount measurement model is established based on the association mechanism. The carbon emission total amount measuring and calculating model is based on industry-level and enterprise-level electric power big data, carbon emission data generated by primary energy consumption are fused, accurate measuring and calculating of the carbon emission total amount of each enterprise are achieved, and carbon emission prediction accuracy is improved.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (10)

1. A method for predicting carbon emissions of an enterprise, the method comprising:
acquiring data to be analyzed, wherein the data to be analyzed comprises at least one of a total production value, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and a power consumption of the target enterprise;
performing empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
2. The method of claim 1, wherein the obtaining data to be analyzed comprises:
acquiring input data, and screening abnormal data from the input data based on abnormal data screening rules according to values correspondingly acquired by the input data at various moments;
determining a first numerical value corresponding to the abnormal data at the last moment and a second numerical value corresponding to the abnormal data at the next moment;
when the first numerical value and the second numerical value both meet a normal data screening rule, determining a substitute numerical value based on the first numerical value and the second numerical value, and substituting the abnormal data by the substitute numerical value;
and determining the data to be analyzed according to the input data after the substitution processing.
3. The method of claim 1, wherein determining data to be analyzed from the input data after the substitution process comprises:
normalizing the input data after the substitution processing to obtain corresponding normalized data;
and taking the corresponding normalized data as data to be analyzed.
4. The method of claim 1, wherein said performing an empirical mode decomposition of said data to be analyzed to obtain a plurality of subsequences of different frequencies and a sequence of residual components comprises:
taking the data to be analyzed as an original input signal, and acquiring a local maximum value point and a local minimum value point corresponding to the original input signal;
determining a corresponding initial frequency component based on the local maximum point, the local minimum point and the original input signal;
when the initial frequency component is determined to meet a preset signal decomposition condition, taking the initial frequency component as a decomposed subsequence;
separating the decomposed subsequence from the original input signal to obtain a corresponding update signal;
and when entering the next iteration cycle, taking the updating signal as an original input signal in the next iteration cycle, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal and continuously executing the steps until the decomposition of all the subsequences and the residual component sequence is completed, and finishing the iteration cycle.
5. The method of claim 4, wherein determining the corresponding initial frequency component based on the local maxima points, local minima points, and the original input signal comprises:
determining a corresponding upper envelope line based on the local maximum value point;
determining a corresponding lower envelope line based on the local minimum value point;
calculating an average value between the upper envelope and the lower envelope, and determining a corresponding initial frequency component based on a difference between the original input signal and the average value;
the method further comprises the following steps:
in each iteration cycle, when the initial frequency component is determined not to meet the preset signal decomposition condition, returning to the step of obtaining the local maximum value point and the local minimum value point corresponding to the original input signal to continue executing until the corresponding subsequence is decomposed, and then calculating the updating signal.
6. The method according to any one of claims 1-5, further comprising:
acquiring sample historical data to be analyzed, and determining sample reference carbon emission data adaptive to the sample historical data to be analyzed;
inputting the historical data to be analyzed of the sample into a carbon emission prediction model to be trained, processing the historical data to be analyzed of the sample through the carbon emission prediction model to be trained, and outputting a corresponding prediction result;
and adjusting model parameters of the carbon emission prediction model to be trained according to the difference between the prediction result and the sample reference carbon emission data until the training stopping condition is reached, and stopping training to obtain the trained carbon emission prediction model.
7. The method of claim 6, wherein determining sample reference carbon emission data that is appropriate for the sample historical data to be analyzed comprises:
calculating the carbon emission proportion for industry history according to the following formula:
Figure FDA0003208450810000031
wherein E isA,total＝EA,electricity+EA,other，EA,electricity＝CA,electricity*αA,powergrid；RA,electricityCarbon proportion for industry history of industry A, EA,totalFor historical total carbon emissions of industry A, EA,electricityHistorical carbon emissions for industry-wide electricity usage in industry A, EA,otherHistorical carbon emissions for other energy consumptions in industry A, CA,electricityFor the industry-wide historical power consumption, alpha, of the A industryA,powergridAverage carbon dioxide emission factors of a power grid in a historical region of an area where the industry A is located;
calculating the historical carbon emission generated by the power utilization of the enterprise according to the following formula:
Ex,electricity＝Cx,electricity*αx,powergrid；
wherein E isx,electricityHistorical carbon emissions, C, for x enterprise electricity usagex,electricityFor the historical net purchase and use of power of x enterprisesAmount, αx,powergridAverage carbon dioxide emission factors of power grids in historical regions of x enterprises;
based on calculated RA,electricityAnd Ex,electricityCalculating the historical total carbon emission of the enterprise by the following formula, and taking the calculated historical total carbon emission of the enterprise as sample reference carbon emission data:
Figure FDA0003208450810000032
8. an enterprise carbon emission prediction apparatus, comprising an acquisition module, a processing module, a prediction module, and an output module, wherein:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring data to be analyzed, and the data to be analyzed comprises at least one of a total domestic production value of everyone, energy intensity, an industry added value of a target enterprise, an industry energy consumption total amount of the target enterprise, a main business income of the target enterprise, a profit of the target enterprise and power consumption of the target enterprise;
the processing module is used for carrying out empirical mode decomposition on the data to be analyzed to obtain a plurality of subsequences with different frequencies and a residual component sequence;
the prediction module is used for respectively inputting a plurality of sequences formed by the subsequences and the residual component sequence into a carbon emission prediction model to obtain sequence prediction results respectively corresponding to the sequences;
and the output module is used for carrying out sequence reconstruction on each obtained sequence prediction result to obtain a reconstruction result, and determining an enterprise carbon emission prediction result corresponding to the target enterprise based on the reconstruction result.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
CN202110923844.3A 2021-08-12 2021-08-12 Enterprise carbon emission prediction method and device, computer equipment and storage medium Pending CN113657661A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110923844.3A CN113657661A (en) 2021-08-12 2021-08-12 Enterprise carbon emission prediction method and device, computer equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110923844.3A CN113657661A (en) 2021-08-12 2021-08-12 Enterprise carbon emission prediction method and device, computer equipment and storage medium

Publications (1)

Publication Number Publication Date
CN113657661A true CN113657661A (en) 2021-11-16

Family

ID=78479519

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110923844.3A Pending CN113657661A (en) 2021-08-12 2021-08-12 Enterprise carbon emission prediction method and device, computer equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113657661A (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114548481A (en) * 2021-12-26 2022-05-27 特斯联科技集团有限公司 Power equipment carbon neutralization processing apparatus based on reinforcement learning
CN115062872A (en) * 2022-08-11 2022-09-16 国网(宁波)综合能源服务有限公司 Industry energy consumption prediction method and prediction system based on electric power big data
CN115511332A (en) * 2022-09-30 2022-12-23 南方电网能源发展研究院有限责任公司 Carbon emission determination method, carbon emission determination device, computer equipment and storage medium
CN115545450A (en) * 2022-09-27 2022-12-30 广东师大维智信息科技有限公司 Carbon emission collaborative prediction method based on digital twinning
CN116579546A (en) * 2023-04-15 2023-08-11 浙江容大电力工程有限公司 Intelligent electricity consumption data analysis and management system based on park
CN117291628A (en) * 2023-09-21 2023-12-26 南通大学 Comprehensive energy system carbon emission amount calculation method, device and storage medium
CN117494063A (en) * 2023-09-08 2024-02-02 国网江苏省电力有限公司仪征市供电分公司 Novel method, system, terminal and medium for monitoring enterprise carbon emission under power system
CN117764214A (en) * 2023-10-23 2024-03-26 国网甘肃省电力公司天水供电公司 Multi-dimensional target dynamic early warning method and system for energy storage power station considering double-carbon targets

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106600027A (en) * 2016-10-31 2017-04-26 上海市政工程设计研究总院（集团）有限公司 Urban traffic carbon emission measurement and calculation system, and measurement and calculation method

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106600027A (en) * 2016-10-31 2017-04-26 上海市政工程设计研究总院（集团）有限公司 Urban traffic carbon emission measurement and calculation system, and measurement and calculation method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
孙佶: "企业节能和温室气体减排项目决策方法研究", 中国优秀硕士学位论文全文数据库工程科技Ⅰ辑(月刊), no. 2014, 15 February 2014 (2014-02-15), pages 027 - 201 *
张国兴;张振华;刘鹏;刘明星;: "我国碳排放增长率的运行机理及预测", 中国管理科学, no. 12 *
张国兴等: "我国碳排放增长率的运行机理及预测", 中国管理科学, vol. 23, no. 12, 31 December 2015 (2015-12-31), pages 86 - 93 *
田中华;杨泽亮;蔡睿贤;: "电力行业对地区节能和碳排放强度下降目标贡献分析", 中国电力, vol. 48, no. 3, 5 March 2015 (2015-03-05), pages 150 - 155 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114548481A (en) * 2021-12-26 2022-05-27 特斯联科技集团有限公司 Power equipment carbon neutralization processing apparatus based on reinforcement learning
CN114548481B (en) * 2021-12-26 2022-08-23 特斯联科技集团有限公司 Power equipment carbon neutralization processing apparatus based on reinforcement learning
CN115062872A (en) * 2022-08-11 2022-09-16 国网(宁波)综合能源服务有限公司 Industry energy consumption prediction method and prediction system based on electric power big data
CN115062872B (en) * 2022-08-11 2022-11-08 国网(宁波)综合能源服务有限公司 Industry energy consumption prediction method and prediction system based on electric power big data
CN115545450B (en) * 2022-09-27 2023-06-06 广东师大维智信息科技有限公司 Carbon emission collaborative prediction method based on digital twin
CN115545450A (en) * 2022-09-27 2022-12-30 广东师大维智信息科技有限公司 Carbon emission collaborative prediction method based on digital twinning
CN115511332A (en) * 2022-09-30 2022-12-23 南方电网能源发展研究院有限责任公司 Carbon emission determination method, carbon emission determination device, computer equipment and storage medium
CN116579546A (en) * 2023-04-15 2023-08-11 浙江容大电力工程有限公司 Intelligent electricity consumption data analysis and management system based on park
CN116579546B (en) * 2023-04-15 2023-12-12 浙江容大电力工程有限公司 Intelligent electricity consumption data analysis and management system based on park
CN117494063A (en) * 2023-09-08 2024-02-02 国网江苏省电力有限公司仪征市供电分公司 Novel method, system, terminal and medium for monitoring enterprise carbon emission under power system
CN117494063B (en) * 2023-09-08 2024-06-07 国网江苏省电力有限公司仪征市供电分公司 Novel enterprise carbon emission monitoring method under power system
CN117291628A (en) * 2023-09-21 2023-12-26 南通大学 Comprehensive energy system carbon emission amount calculation method, device and storage medium
CN117764214A (en) * 2023-10-23 2024-03-26 国网甘肃省电力公司天水供电公司 Multi-dimensional target dynamic early warning method and system for energy storage power station considering double-carbon targets

Similar Documents

Publication Publication Date Title
CN113657661A (en) Enterprise carbon emission prediction method and device, computer equipment and storage medium
CN111860982A (en) Wind power plant short-term wind power prediction method based on VMD-FCM-GRU
US20160239592A1 (en) Data-driven battery aging model using statistical analysis and artificial intelligence
CN113009077B (en) Gas detection method, gas detection device, electronic equipment and storage medium
CN110826803A (en) Electricity price prediction method and device for electric power spot market
CN117389824A (en) Cloud server load prediction method based on signal decomposition and mixing model
WO2024109760A1 (en) Battery life prediction method, prediction system and computer device
Xie et al. Residual life prediction of lithium-ion batteries based on data preprocessing and a priori knowledge-assisted CNN-LSTM
CN111461445A (en) Short-term wind speed prediction method and device, computer equipment and storage medium
CN115238855A (en) Completion method of time sequence knowledge graph based on graph neural network and related equipment
Tessoni et al. Advanced statistical and machine learning methods for multi-step multivariate time series forecasting in predictive maintenance
CN117078480A (en) Carbon emission monitoring method, device, equipment, storage medium and computer product
CN116757062A (en) Power load frequency characteristic analysis method, device, equipment and storage medium
CN116822374A (en) Wind speed multi-step prediction method and device, computer equipment and storage medium
Liu et al. A novel dual-attention optimization model for points classification of power quality disturbances
CN111697560A (en) Method and system for predicting load of power system based on LSTM
CN114781532A (en) Evaluation method and device of machine learning model, computer equipment and medium
Javier et al. Causal network inference in a dam system and its implications on feature selection for machine learning forecasting
Sun et al. Short-term power load prediction based on VMD-SG-LSTM
CN114118570A (en) Service data prediction method and device, electronic equipment and storage medium
CN117561502A (en) Method and device for determining failure reason
Chauvet et al. AI approaches for electricity price forecasting in stable/unstable markets: EU improvement project
CN112183848A (en) Power load probability prediction method based on DWT-SVQR integration
Zhao et al. An interpretable ultra-short-term wind power prediction model based on the feature matrix reconstruction through regression trees
Liu Research on Crude Oil Price Forecasting Based on MLR-SVR Decomposition Integrated Model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination