Carbon emission prediction method and device, electronic equipment and storage medium

Abstract

The invention provides a carbon emission prediction method and device, electronic equipment and a storage medium. Relates to the technical field of data processing. The method comprises the steps that at least one carbon emission influence factor of a to-be-predicted subject is acquired, and any carbon emission influence factor is timeliness data which corresponds to target time and affects the carbon emission of the to-be-predicted subject; and calling a pre-trained carbon emission prediction model, and processing the at least one carbon emission influence factor to predict the carbon emission of the to-be-predicted subject at the target time. According to the invention, the carbon emission can be predicted, so that the predicted subject can better formulate a production and operation strategy by using the predicted data.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a method, device, electronic device and storage medium for predicting carbon emissions. Background Art

[0002] Among environmental issues, carbon emissions are of paramount importance. In production and business activities, enterprises, groups or individuals will consume fuel to varying degrees, thus generating different amounts of carbon emissions.

[0003] In related technologies, the carbon emissions obtained are often result data, that is, only the carbon emissions that have been generated can be obtained. Therefore, how to predict carbon emissions so that the predicted entity can use the predicted data to better formulate production and operation strategies is a technical problem that needs to be solved in this field. Summary of the invention

[0004] In view of this, the purpose of the present disclosure is to propose a method for predicting carbon emissions, which can solve the existing problems in a targeted manner.

[0005] Based on the above-mentioned purposes, in the first aspect, the present disclosure proposes a method for predicting carbon emissions, comprising: obtaining at least one carbon emission influencing factor of the subject to be predicted, wherein the at least one carbon emission influencing factor includes carbon emission influencing factors of multiple dimensions, and any carbon emission influencing factor is time-sensitive data corresponding to a target time and having an impact on the carbon emissions of the subject to be predicted; calling a pre-trained carbon emission prediction model, processing at least one carbon emission influencing factor, to predict the carbon emissions of the subject to be predicted at the target time.

[0006] In the second aspect, a carbon emission prediction device is also provided, including: an acquisition unit, configured to acquire at least one carbon emission influencing factor of a subject to be predicted, wherein the at least one carbon emission influencing factor includes carbon emission influencing factors of multiple dimensions, and any carbon emission influencing factor is time-sensitive data corresponding to a target time and having an impact on the carbon emissions of the subject to be predicted; a prediction unit, configured to call a pre-trained carbon emission prediction model, process at least one carbon emission influencing factor, and predict the carbon emissions of the subject to be predicted at the target time.

[0007] According to a third aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method according to the first aspect.

[0008] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and the program is executed by a processor to implement any method of the first aspect.

[0009] Generally speaking, the present disclosure has at least the following beneficial effects: It is possible to predict carbon emissions, so that the predicted entity can use the predicted data to better formulate production and operation strategies. Moreover, the present disclosure can flexibly select the target time and the entity to be predicted for carbon emission prediction, thereby ensuring the flexibility of prediction. In addition, the present disclosure can also accurately predict the carbon emissions at the target time through timely carbon emission influencing factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In the drawings, unless otherwise specified, the same reference numerals throughout the several views denote the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in accordance with the present disclosure and should not be regarded as limiting the scope of the present disclosure.

[0011] Figure 1 Shows a flowchart of a method for predicting carbon emissions according to an embodiment of the present disclosure;

[0012] Figure 2 Shows another flowchart of a method for predicting carbon emissions according to an embodiment of the present disclosure;

[0013] Figure 3 Shows a schematic diagram of the prediction of carbon emissions according to an embodiment of the present disclosure;

[0014] Figure 4 Shows a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure;

[0015] Figure 5 Shows a schematic diagram of a storage medium provided by an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The present disclosure will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention and are not intended to limit the invention. Additionally, it should be noted that for the sake of description, only parts related to the relevant invention are shown in the drawings.

[0017] It should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other. The present disclosure will be described in detail below with reference to the drawings and embodiments.

[0018] Figure 1 Shows the method for predicting carbon emissions of the present disclosure. In an embodiment of the present disclosure, the method includes:

[0019] Step S101: Obtain at least one carbon emission influencing factor of the subject to be predicted, where any carbon emission influencing factor corresponds to the target time and is time-sensitive data that affects the carbon emissions of the subject to be predicted.

[0020] In this embodiment, the execution entity of the carbon emissions prediction method can obtain at least one carbon emission influencing factor of the subject to be predicted. A carbon emission influencing factor is a parameter that affects the carbon emissions of the carbon emission subject. For example, the carbon emission influencing factors can include at least one of the following: business volume, holidays, e-commerce festivals, etc.

[0021] Specifically, each carbon emission influencing factor has a corresponding additional condition, which is the time condition. That is, the carbon emission influencing factor does not exist alone and must correspond to a time. Therefore, the carbon emission influencing factor has timeliness. The time here can refer to the time when the carbon emissions are generated.

[0022] Step S102: Invoke the pre-trained carbon emissions prediction model to process at least one carbon emission influencing factor to predict the carbon emissions of the subject to be predicted at the target time.

[0023] In this embodiment, the above execution entity can invoke the pre-trained carbon emissions prediction model to process the above at least one carbon emission influencing factor to obtain the carbon emissions predicted by the carbon emissions prediction model. If the obtained carbon emissions are the carbon emissions of the subject to be predicted at the target time.

[0024] A carbon emissions prediction model is a model that can predict carbon emissions using carbon emission influencing factors. This model can be various deep neural networks.

[0025] After obtaining the predicted carbon emissions, the above execution entity or other electronic devices can compare the predicted carbon emissions with the planned carbon emissions. If the planned carbon emissions are smaller, the planned carbon emissions are used as the warning value. If the predicted carbon emissions are smaller, it is necessary to negotiate with the relevant business departments to determine the warning value.

[0026] This embodiment can predict carbon emissions, enabling the subject to be predicted to better formulate production and operation strategies using the predicted data. Moreover, this embodiment can flexibly select the target time and the subject to be predicted for carbon emissions prediction, thus ensuring the flexibility of the prediction. In addition, this embodiment can accurately predict the carbon emissions at the target time through time-sensitive carbon emission influencing factors.

[0027] Figure 2 Illustrates a training method for a carbon emissions prediction model according to an embodiment of the present disclosure. As Figure 2 shown, the training method includes:

[0028] Step S201: Among the historical carbon emission influencing factors, select the carbon emission influencing factors for training the carbon emission prediction model, and use the carbon emissions corresponding to the selected carbon emission influencing factors as training samples to obtain an initial sample set.

[0029] In this embodiment, the above-mentioned execution entity or other electronic devices can execute the training process of the carbon emission prediction model. Taking the above-mentioned execution entity as an example, the above-mentioned execution entity can select the carbon emission influencing factors for training the carbon emission prediction model from the historical data of the historical carbon emission influencing factors, that is, the carbon emission influencing factors. After that, the above-mentioned execution entity can use the carbon emission data corresponding to the selected carbon emission influencing factors as training samples. These training samples can form an initial sample set.

[0030] The above-mentioned execution entity can select the carbon emission influencing factors for training the carbon emission prediction model from the historical carbon emission influencing factors in various ways. For example, the above-mentioned execution entity can input the historical carbon emission influencing factors into a preset model and obtain the carbon emission influencing factors for training the carbon emission prediction model output from the model. Or, the above-mentioned execution entity can select according to a preset selection rule, such as selecting the carbon emission influencing factors within the latest preset time period.

[0031] Specifically, the above-mentioned execution entity can determine the historical carbon emissions corresponding to the selected carbon emission influencing factors from the historical data of carbon emissions.

[0032] The historical data of carbon emissions can be obtained through the following steps:

[0033] (1) Confirm the scenarios:

[0034] Confirm which scenarios the target enterprise, that is, the subject to be predicted, needs to include in carbon emissions and classify them. The carbon emission scenarios in the logistics industry can be divided into four categories: transportation, site, packaging, and employees. According to the specific situation of the enterprise, these four categories can be further subdivided. For example, transportation can be divided into: self-owned vehicles, outsourced vehicles, railways, airplanes, etc.; sites can be divided into: site electricity, site refrigerant, site heating, etc.

[0035] (2) Determine the analysis dimensions:

[0036] Determine the angles for analyzing carbon emissions according to the scenarios. By analyzing the business requirements, determine the following analysis angles, which can be divided into organizational, business, and carbon aspects. Organizationally, it includes branch codes, cities, provinces, etc.; business-wise, it includes license plate numbers, vehicle models, flight numbers, etc.; carbon-wise, it includes emission source ownership, emission source ownership, carbon emission sources, energy usage types, etc.

[0037] (3) Determine the data granularity:

[0038] Determine the data granularity according to the analysis dimension and determine the object for which carbon emissions are to be determined. Similar scenarios should be kept as consistent as possible. For example, the object corresponding to a vehicle is uniformly a combination of a network point code and a license plate number, and the object corresponding to a venue is the venue code, etc.

[0039] (4) Confirm relevant metrics

[0040] Since carbon emissions are a composite indicator data and cannot be directly obtained from the production activities or business processes, it is necessary to collect business data in the production activities or business processes, such as vehicle fuel consumption, venue electricity consumption, packaging material usage, etc. By multiplying by the corresponding carbon emission impact factors, carbon emissions can be obtained.

[0041] (5) Calculate and store

[0042] In the process of obtaining carbon emissions in step (4) above, it is necessary to use big data technology to perform cleaning processes on the data such as deduplication and elimination of invalid data, and also perform preset four arithmetic operations. Finally, store the operation results to form indicator data for display on the terminals of designated business personnel.

[0043] Step S202, call the sample generation model to process the training samples in the initial sample set to generate carbon emissions.

[0044] In this embodiment, the above execution subject can call the sample generation model to process the training samples in the initial sample set, so as to obtain the carbon emissions generated by the sample generation model. The sample generation model is a deep neural network that may generate new samples based on real training samples.

[0045] Step S203, add the generated carbon emissions to the initial sample set to obtain a training sample set.

[0046] In this embodiment, the above execution subject can add the generated carbon emissions to the initial sample set to obtain a training sample set. In this way, the training sample set can include the training samples in the initial sample set and also the generated carbon emissions.

[0047] Step S204, use the training sample set to train the carbon emission prediction model to be trained to obtain a carbon emission prediction model.

[0048] In this embodiment, the above execution subject can use the training sample set to train the carbon emission prediction model to be trained, and the result of the training is the above carbon emission prediction model.

[0049] In this embodiment, carbon emission influencing factors suitable for training can be selected from the carbon emission influencing factors. Moreover, the sample can be expanded through the sample generation model, avoiding the problem of poor training results caused by small data volume and large sample fluctuations.

[0050] Among these alternative implementation manners, in the historical carbon emission influencing factors in step S201, selecting the carbon emission influencing factors for training the carbon emission prediction model includes: determining the influence index of each historical carbon emission influencing factor on the carbon emission; according to the magnitude of the influence index, among each historical carbon emission influencing factor, selecting the carbon emission influencing factors for training the carbon emission prediction model, wherein the influence index of the selected carbon emission influencing factors is greater than the influence index of the unselected carbon emission influencing factors.

[0051] Among these alternative implementation manners, the above-mentioned execution subject can calculate the influence index of each historical carbon emission influencing factor on the carbon emission. The higher the influence index, the greater the influence of the historical carbon emission influencing factor on the carbon emission.

[0052] Specifically, the above-mentioned execution subject can use various methods to determine the influence index. For example, the above-mentioned execution subject can use the grey relational degree to represent the influence index. Using the grey relational degree can ensure accurate analysis results when the relationship between the carbon emission influencing factor and the carbon emission is non-linear. Or, the above-mentioned execution subject can use the Pearson correlation analysis to determine the influence index.

[0053] The above-mentioned execution subject can use various methods to select, according to the magnitude of the influence index, the carbon emission influencing factors for training the carbon emission prediction model from each historical carbon emission influencing factor. For example, the above-mentioned execution subject selects the historical carbon emission influencing factors whose influence index is greater than the preset threshold. For example, the carbon emission influencing factors with an influence index < 0.5 are excluded. Or, the above-mentioned execution subject can select a preset proportion of the historical carbon emission influencing factors in the order of the influence index from large to small.

[0054] These implementation manners can determine the historical carbon emission influencing factors with greater influence on the carbon emission, and use the training samples corresponding to these historical carbon emission influencing factors for model training, thereby improving the training efficiency and avoiding overfitting of the trained model.

[0055] In some alternative application scenarios of these alternative implementation manners, the above method of training the carbon emission prediction model to be trained by using the training sample set to obtain the carbon emission prediction model may include: optimizing the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters, where the training-related parameters include the structural parameters and learning parameters for training the carbon emission prediction model to be trained; using the training values of the training-related parameters to update the carbon emission prediction model to be trained to obtain the updated carbon emission prediction model to be trained; and using the training sample set to train the updated carbon emission prediction model to be trained to obtain the carbon emission prediction model.

[0056] In these application scenarios, the above execution subject may optimize the training-related parameters of the carbon emission prediction model, and the optimization result is the training value of the training-related parameters. The above execution subject may use various optimization models to perform the above optimization process. In some cases, the above execution subject may use the training sample set to perform the optimization.

[0057] The training-related parameters are key hyperparameters, which may specifically be structural parameters and learning parameters. Specifically, the structural parameters may include the number of hidden layer nodes. The learning parameters may include at least one of the learning rate and the batch size.

[0058] The above execution subject may use the values obtained by the optimization to update the initial carbon emission prediction model to obtain the carbon emission prediction model to be trained.

[0059] These application scenarios may first optimize the training-related parameters of the carbon emission prediction model, and then train the carbon emission prediction model updated by using the optimization result. Compared with directly training the entire model without optimization, it can accelerate the convergence of the model and improve the training efficiency of the model.

[0060] Optionally, the training structure of the carbon emission prediction model includes an optimization network and a prediction network; optimizing the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters includes: using the optimization network to optimize the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters; using the training values of the training-related parameters to update the carbon emission prediction model to be trained to obtain the updated carbon emission prediction model to be trained includes: using the training values of the training-related parameters to update the prediction network to obtain the updated prediction network; using the training sample set to train the updated carbon emission prediction model to be trained to obtain the carbon emission prediction model includes: using the training sample set to train the updated prediction network to obtain the carbon emission prediction model including the trained prediction network.

[0061] Specifically, the carbon emission prediction model may include an optimization network and a prediction network. During the training process, the optimization network can optimize the training-related parameters in the prediction network and obtain the optimization result. Subsequently, the above-mentioned execution entity can update the training-related parameters in the prediction network to the optimization result. Then, the above-mentioned execution entity can use the training sample set to train the updated prediction network to obtain the trained prediction network. After that, the above-mentioned execution entity can use the trained prediction network to predict the carbon emissions.

[0062] For example, the optimization network can be the whale optimization algorithm (WOA). Thus, the above-mentioned execution entity can perform optimization through the whale optimization algorithm to obtain the optimization result. The prediction network can be the Bidirectional Mass-Conserving Long Short-Term Memory (Bi-MC-LSTM). While this prediction network can correlate past and future data and mine temporal relationships, it uses the prior information of carbon emission influencing factors as a judgment index to further improve the accuracy of the model.

[0063] These specific application scenarios can optimize the parameters of the prediction network through the optimization network, effectively shortening the training process of the prediction network and improving the efficiency of obtaining the carbon emission prediction model.

[0064] Optionally, the sample generation model includes a generator, a discriminator, and a regressor; the training steps of the sample generation model include: performing a learning step: iterating the generator and the discriminator according to preset iteration parameters, processing the target training sample through the iterated generator to obtain a new carbon emission, where the target training sample is the carbon emission corresponding to the historical carbon emission influencing factors; inputting the new carbon emission into the regressor to obtain the regression value corresponding to the new carbon emission; training the regressor according to the regression value, and training the iterated discriminator according to the iterated generator and the trained regressor; repeating the learning step until the preset training termination condition is reached.

[0065] In these optional application scenarios, the sample generation model can include a generator, a discriminator, and a regressor. These three can be used to generate new training samples from the target training sample. The target training sample is the true carbon emission corresponding to the historical carbon emission influencing factors.

[0066] Specifically, the sample generation model can be various deep neural networks, such as Quantile Regression Conditional Generative Adversarial Nets (QRCGAN). QRCGAN is a virtual sample generation method that combines the advantages of the quantile regression network QRNN and the conditional generative adversarial network CGAN. By adding the conditional signal y, the unsupervised GAN model is extended to a supervised learning framework.

[0067] The sample generation model consists of three parts: a generator G, a discriminator D, and a regressor R.

[0068] (1) Set the initial states of the network parameters of the generator G, the discriminator D, and the regressor R. Define the dimension z_dim of the noise signal z, the batch size batch_size, the learning rate lr, and the ratio n_update of the update times of the generator and the discriminator and the maximum number of iterations max_it.

[0069] (2) In each round of iteration, randomly select a batch of actual data from the set of target training samples, and at the same time extract an equal amount of noise data according to its prior distribution.

[0070] (3) After fixing the generator and the discriminator, input the above actual data into the generator to obtain new training samples. The new training samples are sent to the regressor for calibration to obtain the corresponding output values. Then, use this batch of actual data and the generated simulated data, that is, the new training samples, to train the regressor.

[0071] (4) After fixing the generator and the regressor, use the same set of actual data and the generated simulated data to train the discriminator. After each training of the regressor, the discriminator is trained once.

[0072] (5) Repeat steps (2) to (4). After the discriminator is updated n_update times, while keeping the discriminator and the regressor unchanged, update the network parameters of the generator.

[0073] (6) Continuously repeat steps (2) to (5) until the network parameters stabilize or reach the preset maximum number of iterations max_it.

[0074] The models in these optional application scenarios include a generator, a discriminator, and a regressor. When the conditional signal is continuous numerical data, by adding a regressor, the difficulty of matching the new samples with the given conditional signal is reduced, and at the same time, the requirements for the generator and the discriminator are relaxed. Therefore, the sample generation model is suitable for the need to generate continuous numerical samples in this disclosure.

[0075] In some alternative implementation manners of this embodiment, the historical carbon emissions correspond to a single carbon emission scenario category, and the carbon emission prediction model is used to predict the carbon emissions corresponding to the carbon emission influencing factors in the single carbon emission scenario category.

[0076] In these alternative implementation manners, the historical carbon emissions are data in a single carbon emission scenario category. Each carbon emission scenario category may include multiple carbon emission scenarios. Each analysis perspective of carbon emissions may include multiple carbon emission scenarios, and each carbon emission scenario may have scenario information of different dimensions with an inclusion relationship. For example, the organization in the analysis perspective may include the city dimension and the province dimension.

[0077] Each carbon emission scenario category includes at least one specific carbon emission scenario under each analysis perspective. For example, each carbon emission scenario category may include the carbon emissions of the following items in all cities of Province A: all transportation outlets, diesel vehicles, and self-owned vehicles.

[0078] These implementation manners can model and predict carbon emissions for each carbon emission scenario category, thereby improving the pertinence of the prediction and contributing to improving the accuracy of carbon emission prediction.

[0079] In some alternative application scenarios of these alternative implementation manners, the above method may further include: obtaining a set of carbon emission data of the subject to be predicted, where the set of carbon emission data includes historical carbon emissions corresponding to multiple carbon emission influencing factors, and the set of carbon emission data corresponds to multiple carbon emission scenarios; through the carbon emission scenarios, merging the historical carbon emissions of multiple carbon emission scenarios in the set of carbon emission data into at least two carbon emission scenario categories, where each carbon emission scenario category corresponds to at least one carbon emission scenario, and the single carbon emission scenario category is any one of the at least two carbon emission scenario categories.

[0080] In these alternative application scenarios, the above execution subject may merge the carbon emission scenarios to obtain carbon emission scenario categories. For example, the two carbon emission scenarios of diesel vehicles and self-owned vehicles may be merged into the same carbon emission scenario category.

[0081] These application scenarios can obtain carbon emission scenarios through the merging of scenarios, thereby realizing the modeling and prediction of carbon emissions for multiple scenarios in a large category.

[0082] The embodiments of the present disclosure provide a carbon emission prediction device, which is used to execute the carbon emission prediction method of the above embodiments, such as Figure 3As shown, the device includes: an acquisition unit 301 configured to acquire at least one carbon emission influencing factor of the subject to be predicted, where any carbon emission influencing factor is timeliness data corresponding to a target time and having an impact on the carbon emissions of the subject to be predicted; and a prediction unit 302 configured to call a pre-trained carbon emission prediction model to process at least one carbon emission influencing factor to predict the carbon emissions of the subject to be predicted at the target time.

[0083] Optionally, the training steps of the carbon emission prediction model include: selecting, from historical carbon emission influencing factors, the carbon emission influencing factors for training the carbon emission prediction model, and using the historical carbon emissions corresponding to the selected carbon emission influencing factors as training samples to obtain an initial sample set; calling a sample generation model to process the training samples in the initial sample set to generate carbon emissions; adding the generated carbon emissions to the initial sample set to obtain a training sample set; and using the training sample set to train the carbon emission prediction model to be trained to obtain the carbon emission prediction model.

[0084] Optionally, selecting, from historical carbon emission influencing factors, the carbon emission influencing factors for training the carbon emission prediction model includes: determining the influence index of each historical carbon emission influencing factor on carbon emissions; and selecting, from each historical carbon emission influencing factor according to the magnitude of the influence index, the carbon emission influencing factors for training the carbon emission prediction model, where the influence index of the selected carbon emission influencing factors is greater than that of the unselected carbon emission influencing factors.

[0085] Optionally, using the training sample set to train the carbon emission prediction model to be trained to obtain the carbon emission prediction model includes: optimizing the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters, where the training-related parameters include the structural parameters and learning parameters for training the carbon emission prediction model to be trained; updating the carbon emission prediction model to be trained with the training values of the training-related parameters to obtain an updated carbon emission prediction model to be trained; and using the training sample set to train the updated carbon emission prediction model to be trained to obtain the carbon emission prediction model.

[0086] Optionally, the training structure of the carbon emission prediction model includes an optimization network and a prediction network; optimizing the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters, including: using the optimization network to optimize the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters; using the training values of the training-related parameters to update the carbon emission prediction model to be trained to obtain the updated carbon emission prediction model to be trained, including: using the training values of the training-related parameters to update the prediction network to obtain the updated prediction network; using the training sample set to train the updated carbon emission prediction model to be trained to obtain the carbon emission prediction model, including: using the training sample set to train the updated prediction network to obtain the carbon emission prediction model including the trained prediction network.

[0087] Optionally, the sample generation model includes a generator, a discriminator, and a regressor; the training steps of the sample generation model include: performing a learning step: iterating the generator and the discriminator according to preset iteration parameters, processing the target training samples through the iterated generator to obtain new carbon emissions, where the target training samples are the carbon emissions corresponding to the historical carbon emission influencing factors; inputting the new carbon emissions into the regressor to obtain the regression value corresponding to the new carbon emissions; training the regressor according to the regression value, and training the iterated discriminator according to the iterated generator and the trained regressor; repeating the learning step until the preset training termination condition is reached.

[0088] Optionally, the historical carbon emissions correspond to a single carbon emission scenario category, and the carbon emission prediction model is used to predict the carbon emissions corresponding to the carbon emission influencing factors in the single carbon emission scenario category.

[0089] Optionally, the method further includes: obtaining a carbon emission data set of the subject to be predicted, where the carbon emission data set includes historical carbon emissions corresponding to multiple carbon emission influencing factors, and the carbon emission data set corresponds to multiple carbon emission scenarios; through the carbon emission scenarios, merging the historical carbon emissions of multiple carbon emission scenarios in the carbon emission data set into at least two carbon emission scenario categories, where each carbon emission scenario category corresponds to at least one carbon emission scenario, and the single carbon emission scenario category is any one of the at least two carbon emission scenario categories.

[0090] The carbon emission prediction device provided by the above embodiments of the present disclosure and the carbon emission prediction method provided by the embodiments of the present disclosure are based on the same inventive concept and have the same beneficial effects as the methods adopted, run, or implemented by the application programs stored therein.

[0091] The embodiments of the present disclosure also provide an electronic device corresponding to the carbon emission prediction method provided by the foregoing embodiments to execute the carbon emission prediction method. The embodiments of the present disclosure are not limited thereto.

[0092] Please refer to Figure 4 , which shows a schematic diagram of an electronic device provided by some embodiments of the present disclosure. As Figure 4 shown, the electronic device 40 includes: a processor 400, a memory 401, a bus 402, and a communication interface 403. The processor 400, the communication interface 403, and the memory 401 are connected through the bus 402; a computer program that can run on the processor 400 is stored in the memory 401, and when the processor 400 runs the computer program, it executes the method provided by any one of the foregoing embodiments of the present disclosure.

[0093] Among them, the memory 401 may include a high-speed random access memory (RAM: Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 403 (which can be wired or wireless), a communication connection is established between the system network element and at least one other network element, and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0094] The bus 402 may be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. Among them, the memory 401 is used to store a program. After receiving an execution instruction, the processor 400 executes the program. The carbon emission prediction method disclosed in any one of the foregoing embodiments of the present disclosure can be applied to the processor 400 or implemented by the processor 400.

[0095] The processor 400 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method may be completed by the integrated logic circuit of the hardware in the processor 400 or the instructions in the form of software. The above-mentioned processor 400 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present disclosure. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present disclosure may be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 401, and the processor 400 reads the information in the memory 401 and combines its hardware to complete the steps of the above method.

[0096] The electronic device provided by the embodiments of the present disclosure and the carbon emission prediction method provided by the embodiments of the present disclosure are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by them.

[0097] The embodiments of the present disclosure also provide a computer-readable storage medium corresponding to the carbon emission prediction method provided by the foregoing embodiments. Please refer to Figure 5 , which shows that the computer-readable storage medium is an optical disc 50, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it will execute the carbon emission prediction method provided by any of the foregoing embodiments.

[0098] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here one by one.

[0099] The computer-readable storage medium provided by the above embodiments of the present disclosure and the method for predicting carbon emissions provided by the embodiments of the present disclosure are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0100] It should be noted that:

[0101] In the above text, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "including a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising such element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present disclosure is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, the features described with reference to certain examples may be combined in other examples.

[0102] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present disclosure, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc) and includes several instructions for causing a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of the various embodiments of the present disclosure.

[0103] The embodiments of the present disclosure have been described above in conjunction with the accompanying drawings, which are only specific embodiments of the present disclosure. However, the present disclosure is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present disclosure, those of ordinary skill in the art can also make many forms without departing from the purpose of the present disclosure and the scope protected by the claims, and all of them fall within the protection scope of the present disclosure.

Claims

1. A method for predicting carbon emissions, characterized in that, Including: Obtain at least one carbon emission influencing factor of the subject to be predicted, where any carbon emission influencing factor is timeliness data corresponding to the target time and having an impact on the carbon emissions of the subject to be predicted; Call a pre-trained carbon emission prediction model to process the at least one carbon emission influencing factor to predict the carbon emissions of the subject to be predicted at the target time.

2. The method according to claim 1, wherein The training steps of the carbon emission prediction model include: Among the historical carbon emission influencing factors, select the carbon emission influencing factors for training the carbon emission prediction model, and use the historical carbon emissions corresponding to the selected carbon emission influencing factors as training samples to obtain an initial sample set; Call a sample generation model to process the training samples in the initial sample set to generate carbon emissions; Add the generated carbon emissions to the initial sample set to obtain a training sample set; Use the training sample set to train the carbon emission prediction model to be trained to obtain the carbon emission prediction model.

3. The method according to claim 2, wherein Among the historical carbon emission influencing factors, selecting the carbon emission influencing factors for training the carbon emission prediction model includes: Determine the influence index of each historical carbon emission influencing factor on carbon emissions; According to the magnitude of the influence index, select the carbon emission influencing factors for training the carbon emission prediction model among each historical carbon emission influencing factor, where the influence index of the selected carbon emission influencing factor is greater than that of the unselected carbon emission influencing factor.

4. The method according to claim 3, characterized in that, Using the training sample set to train the carbon emission prediction model to be trained to obtain the carbon emission prediction model includes: Optimize the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters, where the training-related parameters include the structural parameters and learning parameters for training the carbon emission prediction model to be trained; Use the training values of the training-related parameters to update the carbon emission prediction model to be trained to obtain an updated carbon emission prediction model to be trained; Use the training sample set to train the updated carbon emission prediction model to be trained to obtain the carbon emission prediction model.

5. The method according to claim 4, wherein The training structure of the carbon emission prediction model includes an optimization network and a prediction network; Optimizing the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters includes: using the optimization network to optimize the training-related parameters of the carbon emission prediction model to obtain the training values of the training-related parameters; Using the training values of the training-related parameters to update the carbon emission prediction model to be trained to obtain an updated carbon emission prediction model to be trained includes: using the training values of the training-related parameters to update the prediction network to obtain an updated prediction network; Training the updated carbon emission prediction model to be trained using the training sample set to obtain the carbon emission prediction model includes: training the updated prediction network using the training sample set to obtain the carbon emission prediction model including the trained prediction network.

6. The method according to claim 2, characterized in that, The sample generation model includes a generator, a discriminator, and a regressor; the training steps of the sample generation model include: Performing a learning step: iterating the generator and the discriminator according to preset iteration parameters, processing the target training sample through the iterated generator to obtain a new carbon emission, where the target training sample is the carbon emission corresponding to the historical carbon emission influencing factors; inputting the new carbon emission into the regressor to obtain a regression value corresponding to the new carbon emission; training the regressor according to the regression value, and training the iterated discriminator according to the iterated generator and the trained regressor; Repeatedly performing the learning step until a preset training termination condition is reached.

7. The method according to claim 2, wherein The historical carbon emissions correspond to a single carbon emission scenario category, and the carbon emission prediction model is used to predict the carbon emissions corresponding to the carbon emission influencing factors in the single carbon emission scenario category.

8. The method according to claim 7, wherein The method further includes: Obtaining a carbon emission data set of the subject to be predicted, where the carbon emission data set includes historical carbon emissions corresponding to multiple carbon emission influencing factors, and the carbon emission data set corresponds to multiple carbon emission scenarios; Merging the historical carbon emissions of the multiple carbon emission scenarios in the carbon emission data set into at least two carbon emission scenario categories through the carbon emission scenarios, where each carbon emission scenario category corresponds to at least one carbon emission scenario, and the single carbon emission scenario category is any one of the at least two carbon emission scenario categories.

9. A prediction device for carbon emissions, characterized in that, Including: An acquisition unit configured to acquire at least one carbon emission influencing factor of the subject to be predicted, where any carbon emission influencing factor is timeliness data corresponding to a target time and having an impact on the carbon emissions of the subject to be predicted; A prediction unit configured to call a pre-trained carbon emission prediction model to process the at least one carbon emission influencing factor to predict the carbon emissions of the subject to be predicted at the target time.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor runs the computer program to implement the method according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the method according to any one of claims 1-8.

Images