Flight carbon emission prediction method and device, equipment and storage medium

Abstract

The embodiment of the invention provides a flight carbon emission prediction method and device, equipment and a storage medium, and relates to the technical field of carbon emission prediction. The method comprises the following steps: acquiring flight basic information of a target area in a preset historical time period, and calculating corresponding carbon emission historical data; determining a carbon emission sequence signal corresponding to the carbon emission historical data, and decomposing the carbon emission sequence signal into carbon emission sub-sequence signals with different frequencies; performing prediction based on the plurality of carbon emission sub-sequence signals by using the trained carbon emission prediction model to obtain a plurality of corresponding carbon emission decomposition prediction results; and combining the plurality of carbon emission decomposition prediction results to obtain a carbon emission prediction result of the target area in a preset target time period. According to the embodiment of the invention, the carbon emission sequence signal is decomposed, the decomposed signal is predicted in combination with the machine learning model, and the decomposition prediction results are combined to obtain the final prediction result, so that the accuracy and flexibility of the carbon emission prediction result are effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of carbon emission prediction, and in particular to a method, device, equipment and storage medium for predicting carbon emissions of flights. Background Art

[0002] As global warming and climate change become more and more serious, the study of carbon emissions has gradually become the focus of people's attention. Especially with the development of the aviation industry, it is gradually becoming an important source of carbon dioxide emissions.

[0003] At present, the scenario analysis method is usually used to predict carbon emissions. This method mainly predicts the possible situations or consequences of the prediction object by assuming that a certain phenomenon or trend will continue into the future. Although the scenario analysis method can analyze the emission trends under different scenarios based on the driving factors of aviation carbon emissions to predict aviation carbon emissions, the scenario assumptions of this method are highly subjective and require high accuracy, logic and causal relationship of scenario data. Therefore, this method has the problem of insufficient prediction accuracy and flexibility in actual application. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a flight carbon emission prediction method, device, equipment and storage medium to improve the accuracy and flexibility of carbon emission prediction.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting carbon emissions of a flight, comprising: Obtaining basic flight information of the target area within a preset historical period, and calculating corresponding carbon emission historical data based on the basic flight information; Determine a carbon emission sequence signal corresponding to the carbon emission historical data, and decompose the carbon emission sequence signal into a plurality of carbon emission subsequence signals; Using the trained carbon emission prediction model to perform prediction based on the multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results; The multiple carbon emission decomposition prediction results are combined to obtain the carbon emission prediction result of the target area within a preset target time period.

[0006] In an embodiment of the present application, aviation carbon emission data is converted into a sequence signal, and the decomposed sequence signal is input into a prediction model for prediction, and then these decomposed prediction results are merged to obtain a final prediction result, thereby effectively improving the accuracy and flexibility of the carbon emission prediction results.

[0007] In some possible embodiments, the method for obtaining the carbon emission prediction model includes: Based on the preset real carbon emission data, determine the corresponding real carbon emission sequence signal; Based on the carbon emission real sequence signal, several preset initial prediction models are trained respectively to obtain several trained candidate prediction models; Using each candidate prediction model to predict the carbon emission real sequence signal respectively, to obtain candidate prediction results corresponding to each candidate prediction model; The prediction errors between each candidate prediction result and the real sequence signal of the carbon emission are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

[0008] In an embodiment of the present application, multiple candidate models are trained, and based on the error between the predicted results of each candidate model and the actual results, the model with the smallest error is selected as the final prediction model to further improve the accuracy of the prediction results.

[0009] In some possible embodiments, the training of several preset initial prediction models based on the real sequence signal of carbon emissions to obtain several trained candidate prediction models includes: Decomposing the carbon emission real sequence signal into a plurality of carbon emission real subsequence signals; The multiple prediction modules of each initial prediction model are trained based on the multiple carbon emission real subsequence signals to obtain a plurality of trained candidate prediction models.

[0010] In an embodiment of the present application, multiple prediction modules of the initial model are trained after decomposing the sequence signal, thereby further improving the accuracy of the prediction model output.

[0011] In some possible embodiments, the method of using each candidate prediction model to predict the carbon emission real sequence signal to obtain a candidate prediction result corresponding to each candidate prediction model includes: Using each candidate prediction model to predict the multiple carbon emission real subsequence signals respectively, to obtain multiple decomposition prediction results corresponding to each candidate prediction model; The multiple decomposed prediction results of each candidate prediction model are merged respectively to obtain the candidate prediction results corresponding to each candidate prediction model.

[0012] In the embodiment of the present application, decomposition prediction results are obtained by predicting the decomposed multiple subsequence signals respectively, and then error calculation is performed based on the prediction results after combining these decomposition prediction results, thereby improving the accuracy of the selected prediction model.

[0013] In some possible embodiments, decomposing the carbon emission sequence signal into a plurality of carbon emission sub-sequence signals includes: The carbon emission sequence signal is decomposed into a plurality of carbon emission subsequence signals using a variational mode decomposition method.

[0014] In the embodiment of the present application, the variational mode decomposition method is specifically used to decompose the carbon emission sequence signal, thereby further improving the accuracy of the prediction result.

[0015] In some possible embodiments, respectively determining the prediction errors between each candidate prediction result and the real sequence signal of carbon emissions, and determining the corresponding candidate prediction model with the smallest prediction error as the carbon emission prediction model, includes: Based on at least one of the root mean square error calculation method and the mean absolute percentage error calculation method, the prediction errors between each candidate prediction result and the carbon emission real sequence signal are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

[0016] In an embodiment of the present application, the error between each candidate prediction result and the actual data is determined based on different error calculation indicators, thereby further improving the accuracy of the selected prediction model.

[0017] In a second aspect, an embodiment of the present application provides a flight carbon emission prediction device, comprising: A data acquisition module is used to acquire basic flight information of a target area within a preset historical period, and calculate corresponding carbon emission historical data based on the basic flight information; A sequence decomposition module, used to determine a carbon emission sequence signal corresponding to the carbon emission historical data, and decompose the carbon emission sequence signal into a plurality of carbon emission subsequence signals; A data prediction module, used to use a trained carbon emission prediction model to make predictions based on the multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results; The result merging module is used to merge the multiple carbon emission decomposition prediction results to obtain the carbon emission prediction result of the target area within a preset target time period.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the method described in any embodiment of the first aspect when executing the program.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any embodiment of the first aspect can be implemented.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program product, wherein the computer program product includes a computer program, wherein when the computer program is executed by a processor, the method described in any embodiment of the first aspect can be implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A schematic diagram of a method for predicting carbon emissions of a flight provided in an embodiment of the present application; Figure 2 A schematic diagram of a carbon emission prediction result comparison process provided in an embodiment of the present application; Figure 3 A schematic diagram of variational modal decomposition of a carbon emission signal provided in an embodiment of the present application; Figure 4 A schematic diagram showing the comparison between carbon emission prediction results of different models and original signals provided in the embodiments of the present application; Figure 5 A schematic diagram of the structure of a flight carbon emission prediction device provided in an embodiment of the present application; Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0024] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0025] It should be noted that carbon emissions have caused global warming and climate change, becoming one of the key factors exacerbating environmental problems in the 21st century, and have attracted widespread attention from the scientific and political communities. In this context, the development of practical methods to predict carbon dioxide emissions is crucial for policymakers to understand carbon emissions and propose precise emission reduction strategies, especially for the aviation industry, which is gradually becoming an important source of carbon dioxide emissions. According to 2020 data, the aviation industry is expected to continue to grow at a rate of 4.3%, exceeding other transportation sectors. In 2022, global aviation industry carbon dioxide emissions accounted for 2% of total energy-related carbon dioxide emissions. If global commercial aviation is regarded as a country, its emissions will be enough to rank sixth in the global carbon dioxide emission inventory, which shows the seriousness of the aviation industry's carbon emission problem. Faced with a series of emission reduction targets and practical explorations, there is an increasing need for practical carbon emission prediction schemes to assist in dynamic policy adjustment.

[0026] At present, the scenario analysis method is mainly used to solve the carbon emission prediction task. This method is to analyze a certain problem by constructing a corresponding scenario. It is an analysis method that conceives various possible future plans by making detailed and rigorous reasoning and description of the future on the basis of various key assumptions about the major evolution of the economy, industry or technology. However, in the process of using this method to predict carbon emissions, due to the strong subjectivity of scenario assumptions and the high requirements for the accuracy, logic and causal relationship of scenario data, this method has the problem of low prediction accuracy and flexibility in actual application.

[0027] In response to the problems existing in the above-mentioned prior art, an embodiment of the present application provides a flight carbon emission prediction method, which decomposes the sequence signal and then combines the prediction results based on the decomposed signal into the final carbon emission prediction result, thereby effectively improving the accuracy and flexibility of carbon emission prediction.

[0028] like Figure 1 As shown, the embodiment of the present application provides a method for predicting carbon emissions of flights, which may include the steps of: S1. Obtain basic flight information of a target area within a preset historical period, and calculate corresponding carbon emission historical data based on the basic flight information.

[0029] Specifically, first obtain the basic flight information of the target area within the preset historical period, where the target area can be one or more cities, provinces, countries and other regional ranges; the preset historical period can be any historical period in the past week, month or year; the basic flight information may include departure place, destination, aircraft type, great circle distance, airline name, etc., and its data source can be the OAG (Official Airlines Guide) database, which is a resource that provides global capacity and traffic flow data, which contains data on nearly 1,000 global airlines and nearly 4,500 airports.

[0030] Based on the basic flight information obtained, the corresponding carbon emissions can be calculated according to the ICAO carbon emissions calculator method. The formula is as follows: Carbon dioxide emissions per seat = 3.16 × (total fuel × seat freight factor) / (number of seats × seat passenger capacity factor); From this, the formula for the total CO2 emissions of all passengers can be deduced as: Total CO2 emissions = 3.16 × (total fuel × seat freight factor).

[0031] Among them, you can refer to Appendices B and C of the International Civil Aviation Organization (ICAO) carbon emission calculator document to determine the total fuel consumption of each flight based on the aircraft model and the adjusted great circle distance; you can refer to Appendix A of the ICAO carbon emission calculator document to determine the "passenger factor" of each route based on the origin and destination of each flight, including the seat freight coefficient, number of seats and seat passenger capacity coefficient.

[0032] Finally, by summarizing the unit flight carbon emissions of all flights in the destination area, we can obtain the historical carbon emissions data representing the target area in a preset historical period. For example, if each day is used as the unit time of the historical period, each data point represents the daily flight carbon emissions.

[0033] S2. Determine a carbon emission sequence signal corresponding to the carbon emission historical data, and decompose the carbon emission sequence signal into a plurality of carbon emission subsequence signals. Further, the carbon emission sequence signal may be decomposed into a plurality of carbon emission subsequence signals of different frequencies using a variational mode decomposition method.

[0034] It should be noted that by using the variational mode decomposition method to decompose the carbon emission sequence signal, multiple intrinsic mode functions (IMFs) and a residual component can be obtained. Figure 3As shown in the figure, the signals in the figure represent the changing trend of carbon emissions. The top signal is the original signal, that is, the carbon emission sequence signal corresponding to the historical data of carbon emissions; the following five signals are 4 IMF signals and one residual signal obtained after decomposition.

[0035] Specifically, variational mode decomposition (VMD) is a widely used decomposition method that decomposes the input signal into a discrete number of intrinsic mode functions (IMFs), also known as intrinsic mode functions and intrinsic mode components. For each mode, the VMD process has the following steps: first, construct an analytical signal through the Hilbert transform to obtain a one-sided spectrum; second, use the displacement characteristics of the Fourier transform to shift the modal spectrum to the baseband; third, estimate the bandwidth of the signal through the H1 Gaussian smoothness. Then minimize the sum of the spectral widths of all IMFs by solving:

[0036]

[0037] in, For different modes, t represents the time variable, is the frequency center, f is the original signal, K is the mode number, j is a unit imaginary number, is the Dirac increment; st The abbreviation of subject to indicates the constraint condition. This minimization problem is solved by Lagrange multipliers in a series of iterative sub-optimizations, known as the alternative direction method of multipliers.

[0038] It should be noted that compared with another empirical mode decomposition (EMD) method, VMD is more robust to sampling and noise.

[0039] S3. Use the trained carbon emission prediction model to make predictions based on multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results.

[0040] Specifically, the decomposed multiple carbon emission subsequence signals can be respectively input into the trained carbon emission prediction model to obtain the carbon emission decomposition prediction result corresponding to each carbon emission subsequence signal.

[0041] It can be understood that the carbon emission sequence signal and the multiple carbon emission sub-sequence signals after decomposition are both time series signals, which refers to a sequence composed of the observation values ​​of one or a group of statistical indicators x(t) at a series of moments t1, t2,…, tn arranged in chronological order. It is also called dynamic data, which has important applications in nature, economy, society and other fields.

[0042] In time series analysis, commonly used models include seasonal decomposition, exponential smoothing methods, and autoregressive moving average models. Among them: seasonal decomposition is a method to separate seasonal and trend components in time series. By observing and identifying seasonal patterns in time series, we can understand the inherent structure of data and predict future trends; exponential smoothing is a non-parametric prediction method that predicts future trends by giving a larger weight to recent observations and a smaller weight to more distant observations. This method can effectively handle noise and outliers in time series and give smooth prediction results; the autoregressive integrated moving average model (ARIMA) is an improvement based on the autoregressive integrated moving average model. It predicts future trends by identifying and modeling the autoregressive and moving average components in time series. The ARIMA model can handle non-stationary time series data and give accurate prediction results.

[0043] S4. Merge multiple carbon emission decomposition prediction results to obtain a carbon emission prediction result for the target area within a preset target period.

[0044] Finally, by merging the carbon emission decomposition prediction results corresponding to each carbon emission subsequence signal, the carbon emission prediction result of the target area within the preset target period can be obtained. Among them, the preset target period and the preset historical period are corresponding periods, which respectively represent the time scales of input data and output data for the prediction model, that is, the prediction result of the next period is obtained based on the signal prediction of the past period. It can be understood that the prediction result and the input data are of the same data type, that is, the sequence signal of carbon emissions with respect to time. It should be noted that the time scale of input and output can be set according to specific needs.

[0045] Based on this, aviation carbon emission data is converted into sequence signals, and the decomposed sequence signals are input into the prediction model for prediction. Then these decomposed prediction results are merged to obtain the final prediction results, thereby effectively improving the accuracy and flexibility of carbon emission prediction results.

[0046] In some possible embodiments, the method for obtaining the carbon emission prediction model includes: Based on the preset real carbon emission data, determine the corresponding real carbon emission sequence signal; Based on the real sequence signal of carbon emissions, several preset initial prediction models are trained respectively to obtain several trained candidate prediction models; Using each candidate prediction model to predict the real sequence signal of carbon emissions, respectively, to obtain candidate prediction results corresponding to each candidate prediction model; The prediction errors between each candidate prediction result and the real sequence signal of carbon emissions are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

[0047] It should be noted that, in this embodiment, a machine learning (ML) method can be used to obtain a corresponding carbon emission prediction model.

[0048] Specifically, firstly, the real data of carbon emissions in a preset historical period (such as the past year) is obtained, and the corresponding real sequence signal of carbon emissions is determined. The data acquisition method and the signal conversion method can be the same as the method in step S1 above, which will not be repeated here.

[0049] It should be noted that the candidate prediction models can be selected from but are not limited to the traditional machine learning method Random Forest Model (RF), the neural network method Multilayer Perceptron Network Model (MLP) and the Long Short-Term Memory Neural Network Model (LSTM).

[0050] It can be understood that during the training process, the above-mentioned real sequence signal can be divided into a training set and a test set in a certain ratio (such as 8:2 or 9:1, etc.) to train and test the model.

[0051] After obtaining the trained candidate prediction models, the real sequence signals of carbon emissions in the test set are used to input each candidate prediction model to obtain the candidate prediction results output by each candidate prediction model. By comparing these candidate prediction results with the real sequence signals of carbon emissions, the prediction errors corresponding to each candidate prediction model can be calculated, and the candidate prediction model with the smallest error is selected as the final carbon emission prediction model.

[0052] It should be noted that random forest is an ensemble algorithm that uses a collection of multiple decision classification trees to randomly select features or average the results before voting. As a traditional machine learning method, random forest is a collection of tree predictors or weak learners: ,in Represents the observed inputs, either point values ​​or vectors (IMFs). The forecast is the unweighted average of the set:

[0053] in, J Represents the number of decision trees.

[0054] The Multilayer Perceptron (MLP) is an artificial neural network model for time series forecasting that is inspired by biological neural networks in the brain and simplified to mimic them. We can describe the output and input lag The relationship between p is the lag order, and its expression is as follows:

[0055] here, It can be a point value, that is , or vectors (IMFs); are the initial coefficients to be estimated, It is the first i The coefficients of the hidden layers; is a linear function that depends on the model structure or hidden layers, q is the number of nodes in the last hidden layer; this embodiment mainly uses the feedforward neural network (feedforwardnet) in MATLAB software for training, and adopts the conjugate gradient error minimization method to train the MLP network.

[0056] Long short-term memory (LSTM) is a special recursive neural network that can learn long-term dependencies and handle the gradient vanishing problem in traditional RNNs. As a widely used prediction neural network, the input to LSTM in this embodiment is historical data, i.e., point values ​​or sub-signals (IMFs), and the output data is the prediction result.

[0057] In addition, the candidate prediction models involved in the comparison may also include the traditional time series model Autoregressive Integrated Moving Average (ARIMA) model and Vector Autoregression (VAR) model.

[0058] It should be noted that the ARIMA model is mainly composed of three parts, namely, the autoregressive model (AR), the integrated process (I) and the moving average model (MA). It can be understood that in the autoregressive integrated moving average model (ARIMA), it is assumed that the future value of the time series is a linear function of the past observations and random errors, and its expression is:

[0059] in, is the time series data we are considering, d represents the order of the difference; arrive are the parameters of the AR model, which are used to describe the current value and the past p The relationship between the values ​​at each time point; arrive are the parameters of the MA model, which are used to describe the current value and the past q The relationship between the errors at different time points; is t The error term at the time point; c is a constant term.

[0060] It should be noted that the form of the VAR model is that the value of a time series variable is a linear function of the past value of the variable itself and the past values ​​of other variables, and its expression is:

[0061] in, represents a time series column vector; represents the parameter matrix; represents a column vector of constant terms.

[0062] In some possible embodiments, several preset initial prediction models are trained based on the real sequence signal of carbon emissions to obtain several trained candidate prediction models, including: Decomposing the carbon emission real sequence signal into multiple carbon emission real subsequence signals; The multiple prediction modules of each initial prediction model are trained based on multiple carbon emission real subsequence signals to obtain several trained candidate prediction models.

[0063] It should be noted that for the training of the candidate prediction model, the carbon emission real sequence signal can be decomposed, and then the multiple prediction modules of each initial prediction model can be trained based on the decomposed multiple carbon emission real subsequence signals. Specifically, for each initial prediction model, it can be regarded as a collection of multiple prediction modules (sub-models), and the number of prediction modules corresponds to the number of carbon emission real subsequence signals after the decomposition of the carbon emission real sequence signal, that is, each carbon emission real subsequence signal corresponds to one prediction module for training.

[0064] In some possible embodiments, each candidate prediction model is used to predict the real sequence signal of carbon emissions respectively, and the candidate prediction results corresponding to each candidate prediction model are obtained, including: Using each candidate prediction model to predict multiple carbon emission real subsequence signals respectively, and obtaining multiple decomposition prediction results corresponding to each candidate prediction model; The multiple decomposed prediction results of each candidate prediction model are merged respectively to obtain the candidate prediction results corresponding to each candidate prediction model.

[0065] Specifically, since each candidate prediction model can be regarded as a collection of multiple prediction modules, when predicting the real sequence signal of carbon emissions, the real sequence signal of carbon emissions can be decomposed into multiple corresponding real sub-sequence signals of carbon emissions, and then these real sub-sequence signals of carbon emissions can be predicted respectively using different prediction modules of the candidate prediction model to obtain multiple corresponding decomposition prediction results. Finally, these multiple decomposition prediction results are merged to obtain the candidate prediction result corresponding to the candidate prediction model.

[0066] In some possible embodiments, respectively determining the prediction errors between each candidate prediction result and the real sequence signal of carbon emissions, and determining the corresponding candidate prediction model with the smallest prediction error as the carbon emission prediction model, includes: Based on at least one of the root mean square error calculation method and the mean absolute percentage error calculation method, the prediction errors between each candidate prediction result and the real sequence signal of carbon emissions are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

[0067] It should be noted that in order to evaluate the performance of each prediction model, the root mean square error (RMSE) and mean absolute percentage error (MAPE) can be used to verify the accuracy of each prediction model. The calculation expression of RMSE is: ; The calculation expression of MAPE is: .in, represents the true value, represents the predicted value, n is the number of prediction samples, that is, the number of observations.

[0068] Specifically, for each candidate prediction result output by the candidate prediction model, the root mean square error and / or mean absolute percentage error between the candidate prediction result and the real sequence signal of carbon emissions can be calculated respectively, and the prediction error corresponding to the candidate prediction model can be determined based on the root mean square error and / or the mean absolute percentage error. Exemplarily, the weighted average of the root mean square error and the mean absolute percentage error can be calculated according to preset weights as the prediction error corresponding to the candidate prediction model.

[0069] See also Figure 2For example, this embodiment uses five different types of models, including traditional time series models autoregressive integrated moving average (ARIMA) and vector autoregression (VAR), traditional machine learning method random forest (RF), neural network method multilayer perceptron network (MLP) and long short-term memory neural network (LSTM), to predict the carbon emissions of flights in 2019, and then calculates the errors between them and the real data to compare the prediction effects of various models.

[0070] As shown in Table 1 below, it is a comparison of the RMSE and MAPE of the prediction results of various models with the actual results. It can be seen that the RMSE values ​​corresponding to the carbon emissions predicted by each model are between 0.06765 and 0.22254, and another evaluation index MAPE is between 0.010639 and 0.036326. It should be noted that VMD RF, VMD MLP, and VMD LSTM are models that decompose the input data into sub-signals for prediction, while RF, MLP, and LSTM are models that predict directly based on the original input data.

[0071] Table 1: Comparison of model RMSE and MAPE results

[0072] It is not difficult to see from Table 1 that according to the RMSE and MAPE results, the prediction performance of each model is ranked from high to low as follows: VMDLSTM, VAR, VMD MLP, VMD RF, RF, LSTM, ARIMA, MLP. It can be seen that the prediction error of the VMD LSTM model is much different from that of other models. Its RMSE result is 0.06765 and its MAPE result is 0.010639. It can be seen that the VMD LSTM model has the best prediction performance. In addition, compared with only using machine learning models for raw point value prediction, using VMD for data decomposition before prediction usually has higher accuracy. This proves the importance of using the VMD method for data decomposition, which can effectively improve prediction accuracy. The comparison between the prediction results output by each model and the original real data is shown in the figure. Figure 4 As shown by Figure 4 It can be seen that the prediction results corresponding to the VMD LSTM model have the highest overlap with the original real data, indicating that the VMD LSTM model has the best prediction performance.

[0073] Please refer to Figure 5 , Figure 5 The following is a block diagram showing the composition of the flight carbon emission prediction device provided by some embodiments of the present application. It should be understood that the flight carbon emission prediction device is similar to the above-mentioned Figure 1Corresponding to the method embodiment, it is able to execute each step involved in the above method embodiment. The specific functions of the flight carbon emission prediction device can be found in the description above. To avoid repetition, the detailed description is appropriately omitted here.

[0074] Figure 5 The flight carbon emission prediction device includes at least one software function module that can be stored in a memory in the form of software or firmware or fixed in the flight carbon emission prediction device, and the flight carbon emission prediction device includes: The data acquisition module 510 is used to acquire basic flight information of the target area within a preset historical period, and calculate corresponding carbon emission historical data based on the basic flight information; A sequence decomposition module 520, for determining a carbon emission sequence signal corresponding to the carbon emission historical data, and decomposing the carbon emission sequence signal into a plurality of carbon emission subsequence signals; A data prediction module 530 is used to use a trained carbon emission prediction model to make predictions based on multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results; The result merging module 540 is used to merge multiple carbon emission decomposition prediction results to obtain the carbon emission prediction result of the target area within a preset target time period.

[0075] It can be understood that the above-mentioned device embodiment corresponds to the method embodiment of the present invention. The flight carbon emission prediction device provided by the embodiment of the present invention can implement the flight carbon emission prediction method provided by any method embodiment of the present invention.

[0076] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method, and will not be described in detail here.

[0077] like Figure 6 As shown, some embodiments of the present application provide an electronic device 600, which includes: a memory 610, a processor 620, and a computer program stored in the memory 610 and executable on the processor 620, wherein the processor 620 reads the program from the memory 610 via a bus 630 and executes the program to implement a method such as any embodiment of the flight carbon emission prediction method described above.

[0078] Processor 620 can process digital signals and can include various computing structures, such as complex instruction set computer structure, reduced instruction set computer structure, or a structure that implements a combination of multiple instruction sets. In some examples, processor 620 can be a microprocessor.

[0079] The memory 610 may be used to store instructions executed by the processor 620 or data related to the execution of instructions. These instructions and / or data may include codes for implementing some or all functions of one or more modules described in the embodiments of the present application. The processor 620 of the disclosed embodiment may be used to execute instructions in the memory 610 to implement the method shown above. The memory 610 includes a dynamic random access memory, a static random access memory, a flash memory, an optical memory, or other memory known to those skilled in the art.

[0080] Some embodiments of the present application further provide a computer-readable storage medium having a computer program stored thereon. The computer program is executed by a processor to execute the method described in the method embodiment.

[0081] Some embodiments of the present application further provide a computer program product, which, when executed on a computer, enables the computer to execute the method described in the method embodiment.

[0082] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0083] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0084] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0085] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0086] The above description is only an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0087] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0088] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

Claims

1. A method for predicting flight carbon emissions, characterized in that: include: Obtaining basic flight information of the target area within a preset historical period, and calculating corresponding carbon emission historical data based on the basic flight information; Determine a carbon emission sequence signal corresponding to the carbon emission historical data, and decompose the carbon emission sequence signal into a plurality of carbon emission subsequence signals; Using the trained carbon emission prediction model to perform prediction based on the multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results; The multiple carbon emission decomposition prediction results are combined to obtain the carbon emission prediction result of the target area within a preset target time period.

2. The flight carbon emission prediction method according to claim 1, characterized in that: The method for obtaining the carbon emission prediction model includes: Based on the preset real carbon emission data, determine the corresponding real carbon emission sequence signal; Based on the carbon emission real sequence signal, several preset initial prediction models are trained respectively to obtain several trained candidate prediction models; Using each candidate prediction model to predict the carbon emission real sequence signal respectively, to obtain candidate prediction results corresponding to each candidate prediction model; The prediction errors between each candidate prediction result and the real sequence signal of the carbon emission are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

3. The flight carbon emission prediction method according to claim 2, characterized in that: The method of training a plurality of preset initial prediction models based on the carbon emission real sequence signal to obtain a plurality of trained candidate prediction models includes: Decomposing the carbon emission real sequence signal into a plurality of carbon emission real subsequence signals; The multiple prediction modules of each initial prediction model are trained based on the multiple carbon emission real subsequence signals to obtain a plurality of trained candidate prediction models.

4. The flight carbon emission prediction method according to claim 3, characterized in that: The method of using each candidate prediction model to predict the carbon emission real sequence signal respectively to obtain candidate prediction results corresponding to each candidate prediction model includes: Using each candidate prediction model to predict the multiple carbon emission real subsequence signals respectively, to obtain multiple decomposition prediction results corresponding to each candidate prediction model; The multiple decomposed prediction results of each candidate prediction model are merged respectively to obtain the candidate prediction results corresponding to each candidate prediction model.

5. The flight carbon emission prediction method according to claim 2, characterized in that: The step of decomposing the carbon emission sequence signal into a plurality of carbon emission subsequence signals comprises: The carbon emission sequence signal is decomposed into a plurality of carbon emission subsequence signals using a variational mode decomposition method.

6. The flight carbon emission prediction method according to claim 2, characterized in that: The step of respectively determining the prediction errors between each candidate prediction result and the real sequence signal of carbon emissions, and determining the corresponding candidate prediction model with the smallest prediction error as the carbon emissions prediction model, comprises: Based on at least one of the root mean square error calculation method and the mean absolute percentage error calculation method, the prediction errors between each candidate prediction result and the carbon emission real sequence signal are determined respectively, and the corresponding candidate prediction model with the smallest prediction error is determined as the carbon emission prediction model.

7. A flight carbon emission prediction device, characterized in that: include: A data acquisition module is used to acquire basic flight information of a target area within a preset historical period, and calculate corresponding carbon emission historical data based on the basic flight information; A sequence decomposition module, used to determine a carbon emission sequence signal corresponding to the carbon emission historical data, and decompose the carbon emission sequence signal into a plurality of carbon emission subsequence signals; A data prediction module, used to use a trained carbon emission prediction model to make predictions based on the multiple carbon emission subsequence signals to obtain corresponding multiple carbon emission decomposition prediction results; The result merging module is used to merge the multiple carbon emission decomposition prediction results to obtain the carbon emission prediction result of the target area within a preset target time period.

8. An electronic device, characterized in that: It comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the flight carbon emission prediction method according to any one of claims 1 to 6 can be implemented when the processor executes the program.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the flight carbon emission prediction method according to any one of claims 1 to 6 is executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the flight carbon emission prediction method according to any one of claims 1 to 6 is implemented.

Images