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

Abstract

The invention discloses a regional carbon emission prediction method and device, equipment and a storage medium. The method comprises the following steps: acquiring time sequence carbon emission information and a county sequence characteristic matrix corresponding to each county in a target area; and constructing a county region adjacency matrix based on the geographical location relationship between the counties in the target region. And according to the county time sequence feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, determining neighborhood interference feature information of each county in the target region, the target interference feature determination model comprising a subspace feature determination model and a subtime feature determination model. According to the time sequence carbon emission information, the neighborhood interference feature information and the pre-trained target carbon emission prediction model, the region prediction carbon emission corresponding to the next moment of the target region is determined, the county region in the prediction region is used as a prediction object, and the accuracy of region carbon emission prediction is improved according to the carbon emission distribution and influence range.

Description

Technical Field

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

[0002] As an important national strategy, China's carbon reduction adopts a top-down approach, and regional differences will significantly affect the implementation of the carbon reduction strategy. The county is the most basic administrative unit in China and can better reflect regional heterogeneity than provinces and prefecture-level cities. Therefore, assessing county-level carbon dioxide emissions will help to more accurately portray China's carbon emission pattern and formulate carbon reduction policies according to local conditions.

[0003] Carbon emissions have the characteristics of cross-border transmission and boundary expansion. The carbon emission pattern is the result of the interaction between itself and the neighboring space. The scale is too large to analyze the impact of spatial expansion and spatial interaction. At present, the research on regional carbon emissions in China is mostly limited to the provincial and municipal scales. The existing methods are difficult to take into account the geographical location and time series factors between regions at the same time. Therefore, the prediction results of carbon emissions have large errors and cannot meet the actual use needs. Summary of the invention

[0004] The present invention provides a regional carbon emission prediction method, device, equipment and storage medium, which takes counties within the prediction area as prediction objects and improves the accuracy of regional carbon emission prediction based on carbon emission distribution and impact range.

[0005] According to one aspect of the present invention, a method for predicting regional carbon emissions is provided, the method comprising:

[0006] Obtaining time-series carbon emission information and a county time-series characteristic matrix corresponding to each county in the target area, wherein the county time-series characteristic matrix is ​​determined according to the county characteristics of the county;

[0007] Based on the geographical location relationship between the counties in the target area, a county adjacency matrix is ​​constructed; according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, the neighborhood interference feature information of each county in the target area is determined, wherein the target interference feature determination model includes a subspace feature determination model and a subtime feature determination model;

[0008] The regional predicted carbon emissions corresponding to the target area at the next moment are determined according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model.

[0009] According to another aspect of the present invention, a regional carbon emission prediction device is provided. The device comprises:

[0010] A target area information acquisition module, used to acquire the time series carbon emission information and the county time series characteristic matrix corresponding to each county in the target area, wherein the county time series characteristic matrix is ​​determined according to the county characteristics of the county;

[0011] A county adjacency matrix determination module, used to construct a county adjacency matrix based on the geographical location relationship between the counties in the target area;

[0012] An interference feature information determination module, used to determine the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, wherein the target interference feature determination model includes a subspace feature determination model and a subtime feature determination model;

[0013] The regional carbon emission prediction module is used to determine the regional predicted carbon emission corresponding to the target area at the next moment according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model.

[0014] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the regional carbon emission prediction method described in any embodiment of the present invention.

[0018] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the regional carbon emission prediction method described in any embodiment of the present invention when executed.

[0019] The technical solution of the embodiment of the present invention is to obtain the time-series carbon emission information and the county time-series feature matrix corresponding to each county in the target area. Based on the geographical location relationship between the counties in the target area, a county adjacency matrix is ​​constructed. According to the county time-series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, the neighborhood interference feature information of each county in the target area is determined. According to the time-series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model, the regional predicted carbon emissions corresponding to the target area at the next moment are determined, which solves the problem that the impact of spatial expansion and spatial interaction cannot be analyzed due to the large scale, and takes the counties in the prediction area as the prediction objects, thereby improving the accuracy of regional carbon emission prediction.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 is a flow chart of a method for predicting regional carbon emissions provided according to Embodiment 1 of the present invention;

[0023] Figure 2 is a county-adjacent schematic diagram of a target area provided according to Embodiment 1 of the present invention;

[0024] Figure 3 is a structural diagram of a regional carbon emission prediction device provided according to Embodiment 2 of the present invention;

[0025] Figure 4 It is a schematic diagram of the structure of an electronic device for implementing the regional carbon emission prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] Embodiment 1

[0029] Figure 1 A flowchart of a regional carbon emission prediction method is provided for the first embodiment of the present invention. This embodiment is applicable to the case of predicting carbon emissions in a target area. The method can be performed by a regional carbon emission prediction device. The regional carbon emission prediction device can be implemented in the form of hardware and / or software. The regional carbon emission prediction device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0030] S101. Obtain the time series carbon emission information and the county time series feature matrix corresponding to each county in the target area.

[0031] The target area may refer to the area where the carbon emissions are to be predicted. The target area may be a province, a city, or a combination of multiple provinces or cities. It should be noted that the target area includes at least two counties. The county time series feature matrix may refer to the matrix information established based on the feature information of the county over a period of time. The time series carbon emission information may refer to the continuous carbon emission information of the county over a period of time.

[0032] It should be noted that most of the factors affecting carbon emissions in the target area have time-series correlation, and the time-series effect will affect the evolution of carbon emissions in different regions. Therefore, taking into account the influence of spatial and temporal effects at the county scale is the key to scientifically analyzing the distribution and impact range of carbon emissions.

[0033] Specifically, the time series carbon emission information corresponding to each county is obtained through the database or county statistical information. For example, the time series carbon emission information can be obtained through the China Carbon Accounting Database website. On the other hand, the county characteristics of each county are obtained, and the county time series feature matrix is ​​constructed according to the county characteristics and time series information.

[0034] Exemplarily, the process of determining the county time series feature matrix includes: obtaining the county time series feature data corresponding to the county; processing the county time series feature data to determine the county features at each moment; for each moment, combining the county features at the moment into a moment feature vector; based on the time series, combining the feature vectors at each moment to obtain the county time series feature matrix.

[0035] Among them, the county time series feature data may refer to the data information of the county information after being arranged and combined according to the time series. Exemplarily, the county time series feature data at least includes night light data, GDP (Gross Domestic Product), NDVI (Normalized Difference Vegetation Index), precipitation data, population density, surface temperature, terrain complexity and elevation data and other information.

[0036] Specifically, it is necessary to collect all relevant county time series feature data and preprocess them, such as cleaning missing values, outlier detection, standardization or normalization, etc., to obtain county features at each moment. For the data at each moment, useful features are extracted. This can be achieved through statistical methods (such as mean, standard deviation), machine learning methods (such as PCA principal component analysis), deep learning methods (such as convolutional neural network CNN), etc. For example: night light data can be quantified by calculating the average brightness or total brightness. GDP can use growth rate or per capita GDP as a feature. NDVI can represent vegetation status by calculating the average value or rate of change. Precipitation can be statistically accumulated or frequency distributed. Population density can be directly used as a numerical value, or dynamic changes can be represented by growth rate. Surface temperature can use average temperature or temperature difference as a feature. Terrain complexity and elevation data can be represented by calculating the standard deviation or range of variation of altitude.

[0037] For each moment, all the features extracted above are combined into a feature vector, that is, the moment feature vector is obtained. Assuming k features are selected, the feature vector for each moment will be k-dimensional. For example, if night light intensity, GDP growth rate, NDVI average, precipitation, population density, surface temperature average, terrain complexity and average altitude are selected as features, then for each moment, an 8-dimensional moment feature vector will be obtained.

[0038] According to the given time series, all the moment feature vectors are combined into a matrix. If the time series has T moments, then the final county time series feature matrix will be of size T×k.

[0039] S102: construct a county adjacency matrix based on the geographical location relationship between the counties in the target area.

[0040] Specifically, collect geographic boundary information for each county in the target area, which can usually be obtained from public GIS data sources or found in government-published statistics. The data usually contains geographic coordinates (such as longitude and latitude) and / or polygon definitions (vertex coordinate lists). Calculate which counties are adjacent to each other, and then construct a county adjacency matrix based on the adjacent results.

[0041] Exemplarily, the step of constructing a county adjacency matrix based on the geographical location relationship between counties in the target area includes:

[0042] Based on the geographical location relationship between the counties in the target area, a regional topological map corresponding to the target area is constructed, wherein the regional topological map is used to represent the geographical adjacent relationship of the counties; based on the regional topological map, matrix element data in the adjacency matrix is ​​determined; according to the matrix element data, a county adjacency matrix is ​​constructed.

[0043] Figure 2 This is a schematic diagram of the counties adjacent to the target area provided by the present invention. Figure 2 The left picture is a county geographical map. Figure 2 The picture on the right is a county topology map. Figure 2 The geographical map in the target area can determine the geographical location relationship between the counties in the target area, and then construct a regional topological map corresponding to the target area according to the geographical location relationship between the counties. Figure 2 As shown in the topological diagram in , five counties are represented by A, B, C, D, and E, respectively. County A borders County B, County D, and County E. In the topological diagram, nodes A and B, D, and E are connected by edges, and the rest are analogous. The county adjacency matrix is ​​established according to the regional topological diagram. When two county nodes are connected, the corresponding element in the adjacency matrix is ​​1, otherwise it is 0. In the topological diagram, node A is connected to node B, node D, and node E. In the first row of the corresponding adjacency matrix, the elements at columns B, D, and E are 1; at the same time, the elements at rows B, D, and E in the first column are also 1, and the rest are analogous, so that the county adjacency matrix can be obtained.

[0044] S103, determining the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model.

[0045] Among them, the target interference feature determination model can be composed of a subspace feature determination model and a sub-temporal feature determination model. The subspace feature determination model can be a GCN (Graph Convolutional Network). The basic idea of ​​GCN is to apply convolution operations to the nodes in the graph, allowing each node to aggregate the information of its neighboring nodes to update its own feature representation. The sub-temporal feature determination model can be a TCN (Temporal Convolutional Network). TCN uses multiple convolutional layers to extract features in time series, and ensures the causality of the model through causal convolution, that is, the output will not be affected by future data. The county interference feature information can be the mutual interference information of carbon emissions in adjacent counties within the target area.

[0046] Specifically, for each county, the county time series feature matrix and the county adjacency matrix of the county are input into the target interference feature determination model, and the neighborhood interference feature information is obtained according to the output of the target interference feature determination model, and then the field interference feature information corresponding to each county in the target area can be obtained.

[0047] Exemplarily, determining the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model includes:

[0048] Inputting the county time series feature matrix and the county adjacency matrix into the subspace feature determination model, and determining the spatial interference feature information of each of the counties in the target area according to the output of the subspace feature determination model;

[0049] The spatial interference characteristic information is input into the sub-time characteristic determination model, and the neighborhood interference characteristic information of each of the counties in the target area is determined according to the output of the sub-time characteristic determination model.

[0050] The spatial interference characteristic information may refer to the spatial interference information between adjacent counties. It should be noted that the neighborhood interference characteristic information may include the spatial interference information and the temporal interference information between adjacent counties.

[0051] Specifically, the county time series feature matrix and the county adjacency matrix are input into the subspace feature determination model for spatial interference processing, so as to determine the spatial interference feature information between each county and the adjacent counties. The spatial interference feature information is input into the subtime feature determination model for temporal interference processing, and then the neighborhood interference feature information of each county in the target area can be determined.

[0052] S104: Determine the regional predicted carbon emissions corresponding to the target area at the next moment according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model.

[0053] Optionally, the target carbon emission prediction model can be obtained by training according to the training samples. The target carbon emission prediction model can be used to predict the carbon emission of the target area at the next moment.

[0054] Specifically, the time series carbon emission information and neighborhood interference feature information are input into the target carbon emission prediction model for carbon emission prediction. According to the output of the target emission prediction model, the regional predicted carbon emissions of the target area at the next moment can be obtained.

[0055] Exemplarily, the method of determining the regional predicted carbon emissions corresponding to the target area at the next moment based on the time-series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model includes: for each county, determining the carbon emissions at the current moment and the carbon emissions at the previous moment based on the time-series carbon emission information corresponding to the county; inputting the carbon emissions at the current moment, the carbon emissions at the previous moment and the neighborhood interference feature information into the target carbon emission prediction model for prediction processing; obtaining the county predicted carbon emissions corresponding to the next moment of the county based on the output of the target carbon emission prediction model; and determining the regional predicted carbon emissions corresponding to the target area at the next moment based on the predicted carbon emissions of each county.

[0056] That is to say, the time-series carbon emission information is parsed and processed, so that the carbon emissions of the county at the current moment and the carbon emissions at the previous moment can be obtained. The carbon emissions at the current moment and the carbon emissions at the previous moment, as well as the neighborhood interference feature information, are input into the target carbon emission prediction model, so that the county-level predicted carbon emissions corresponding to the next moment of the county can be obtained. Then, according to the county-level predicted carbon emissions of each county, the regional predicted carbon emissions corresponding to the next moment of the target area are determined. The technical solution of the embodiment of the present invention obtains the time-series carbon emission information and the county-level time-series feature matrix corresponding to each county in the target area. Based on the geographical location relationship between the counties in the target area, a county adjacency matrix is ​​constructed. According to the county-level time-series feature matrix, the county-level adjacency matrix and the pre-trained target interference feature determination model, the neighborhood interference feature information of each county in the target area is determined. According to the time series carbon emission information, the neighborhood interference feature information and the pre-trained target carbon emission prediction model, the regional predicted carbon emissions corresponding to the target area at the next moment are determined, which solves the problem that the impact of spatial expansion and spatial interaction cannot be analyzed due to the large scale. The counties within the prediction area are used as prediction objects, which improves the accuracy of regional carbon emission prediction.

[0057] Based on the above embodiments, the training process of the target carbon emission prediction model includes:

[0058] Acquire historical time series carbon emissions, and determine carbon emission training samples and carbon emission verification samples corresponding to the carbon emission training samples according to the historical time series carbon emissions.

[0059] Inputting the carbon emission training samples and the predetermined neighborhood interference feature samples into a preset carbon emission prediction model to predict carbon emissions;

[0060] Determining a training error of the preset carbon emission prediction model according to the carbon emission output information output by the preset carbon emission prediction model and the carbon emission verification sample;

[0061] Back-propagating the training error to the preset carbon emission prediction model to adjust the network parameters in the preset carbon emission prediction model;

[0062] When the preset convergence condition is met, it is determined that the training of the preset carbon emission prediction model is completed, and the target carbon emission prediction model is obtained.

[0063] It should be noted that the carbon emission training sample needs to include carbon emissions at least at two consecutive moments. When the carbon emission training sample cannot include carbon emissions at least at two consecutive moments, the preset initial value can be used as the carbon emissions at two consecutive moments.

[0064] Specifically, the historical time series carbon emissions are divided, and according to a preset ratio, a part of them is used as carbon emission training samples, and the other part is determined as carbon emission verification samples. According to the carbon emission training samples and the preset carbon emission prediction model, the training error can be determined based on the training function according to the carbon emission output information of the preset carbon emission prediction model and the carbon emission verification sample, and the training error is back-propagated to the preset carbon emission prediction model, and the network parameters in the preset carbon emission prediction model are adjusted until the preset convergence conditions are met, such as when the number of iterations reaches the preset number or the training error converges, it is determined that the training of the preset carbon emission prediction model is completed, and at this time, the preset carbon emission prediction model that has been trained can be used as the target carbon emission prediction model. By using carbon emission training samples and carbon emission verification samples for model training, the accuracy of the target carbon emission prediction model in predicting carbon emissions can be guaranteed, and then the accuracy of the carbon emissions in the target area can be predicted.

[0065] Embodiment 2

[0066] Figure 3 This is a schematic diagram of the structure of a regional carbon emission prediction device provided in Example 2 of the present invention. Figure 3 As shown, the device comprises:

[0067] The target area information acquisition module 301 is used to obtain the time series carbon emission information and the county time series characteristic matrix corresponding to each county in the target area, wherein the county time series characteristic matrix is ​​determined according to the county characteristics of the county;

[0068] A county adjacency matrix determination module 302 is used to construct a county adjacency matrix based on the geographical location relationship between the counties in the target area;

[0069] The interference feature information determination module 303 is used to determine the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, wherein the target interference feature determination model includes a subspace feature determination model and a subtime feature determination model;

[0070] The regional carbon emission prediction module 304 is used to determine the regional predicted carbon emission corresponding to the target area at the next moment according to the time series carbon emission information, the neighborhood interference feature information and the pre-trained target carbon emission prediction model.

[0071] The technical solution of the embodiment of the present invention is to obtain the time-series carbon emission information and the county time-series feature matrix corresponding to each county in the target area. Based on the geographical location relationship between the counties in the target area, a county adjacency matrix is ​​constructed. According to the county time-series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, the neighborhood interference feature information of each county in the target area is determined. According to the time-series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model, the regional predicted carbon emissions corresponding to the target area at the next moment are determined, which solves the problem that the impact of spatial expansion and spatial interaction cannot be analyzed due to the large scale, and takes the counties in the prediction area as the prediction objects, thereby improving the accuracy of regional carbon emission prediction.

[0072] Optionally, the target area information acquisition module 301 is specifically used to:

[0073] Obtaining county time series feature data corresponding to the county;

[0074] Processing the county time series characteristic data to determine the county characteristics at each moment;

[0075] For each moment, a moment feature vector is formed according to the county features of the moment;

[0076] Based on the time series, the feature vectors of each moment are combined and processed to obtain the county time series feature matrix.

[0077] Optionally, the county time series characteristic data includes at least night light data, GDP, NDVI, precipitation data, population density, surface temperature, terrain complexity and elevation data.

[0078] Optionally, the county adjacency matrix determination module 302 is specifically configured to:

[0079] Based on the geographical location relationship between the counties in the target area, construct a regional topological map corresponding to the target area, wherein the regional topological map is used to represent the geographical adjacent relationship of the counties; based on the regional topological map, determine the matrix element data in the adjacency matrix;

[0080] A county adjacency matrix is ​​constructed based on the matrix element data.

[0081] Optionally, the interference characteristic information determination module 303 is specifically configured to:

[0082] Inputting the county time series feature matrix and the county adjacency matrix into the subspace feature determination model, and determining the spatial interference feature information of each of the counties in the target area according to the output of the subspace feature determination model;

[0083] The spatial interference characteristic information is input into the sub-time characteristic determination model, and the neighborhood interference characteristic information of each of the counties in the target area is determined according to the output of the sub-time characteristic determination model.

[0084] Optionally, the regional carbon emission prediction module 304 is specifically used to:

[0085] For each county, determine the carbon emissions at the current moment and the carbon emissions at the previous moment according to the time series carbon emissions information corresponding to the county;

[0086] Inputting the carbon emissions at the current moment, the carbon emissions at the previous moment and the neighborhood interference feature information into the target carbon emissions prediction model for prediction processing;

[0087] According to the output of the target carbon emission prediction model, the predicted carbon emission of the county corresponding to the next moment is obtained;

[0088] According to the predicted carbon emissions of each of the counties, the predicted carbon emissions of the region corresponding to the target region at the next moment are determined.

[0089] Optionally, the device further comprises a prediction model training module.

[0090] Prediction model training module for:

[0091] Acquire historical time series carbon emissions, and determine a carbon emission training sample and a carbon emission verification sample corresponding to the carbon emission training sample according to the historical time series carbon emissions, wherein the carbon emission training sample includes carbon emissions at at least two consecutive moments;

[0092] Inputting the carbon emission training samples and the predetermined neighborhood interference feature samples into a preset carbon emission prediction model to predict carbon emissions;

[0093] Determining a training error of the preset carbon emission prediction model according to the carbon emission output information output by the preset carbon emission prediction model and the carbon emission verification sample;

[0094] Back-propagating the training error to the preset carbon emission prediction model to adjust the network parameters in the preset carbon emission prediction model;

[0095] When the preset convergence condition is met, it is determined that the training of the preset carbon emission prediction model is completed, and the target carbon emission prediction model is obtained.

[0096] The regional carbon emission prediction device provided in the embodiment of the present invention can execute the regional carbon emission prediction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0097] Embodiment 3

[0098] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0099] like Figure 4As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0100] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0101] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as the method for regional carbon emissions prediction.

[0102] In some embodiments, the method regional carbon emission prediction may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method regional carbon emission prediction described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the method regional carbon emission prediction in any other appropriate manner (e.g., by means of firmware).

[0103] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0104] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0105] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0106] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0107] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0108] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0109] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0110] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A regional carbon emission prediction method, characterized in that: include: Obtaining time-series carbon emission information and a county time-series characteristic matrix corresponding to each county in the target area, wherein the county time-series characteristic matrix is ​​determined according to the county characteristics of the county; Based on the geographical location relationship between the counties in the target area, construct a county adjacency matrix; Determine the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, wherein the target interference feature determination model includes a subspace feature determination model and a subtime feature determination model; The regional predicted carbon emissions corresponding to the target area at the next moment are determined according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model.

2. The method according to claim 1, characterized in that The process of determining the county time series feature matrix includes: Obtaining county time series feature data corresponding to the county; Processing the county time series characteristic data to determine the county characteristics at each moment; For each moment, a moment feature vector is formed according to the county features of the moment; Based on the time series, the feature vectors of each moment are combined and processed to obtain the county time series feature matrix.

3. The method according to claim 2, characterized in that The county time series characteristic data at least includes night light data, GDP, NDVI, precipitation data, population density, surface temperature, terrain complexity and elevation data.

4. The method according to claim 1, characterized in that The constructing of a county adjacency matrix based on the geographical location relationship between the counties in the target area includes: Based on the geographical location relationship between the counties in the target area, construct a regional topological map corresponding to the target area, wherein the regional topological map is used to represent the geographical adjacent relationship of the counties; based on the regional topological map, determine the matrix element data in the adjacency matrix; A county adjacency matrix is ​​constructed based on the matrix element data.

5. The method according to claim 1, characterized in that The determining of the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model includes: Inputting the county time series feature matrix and the county adjacency matrix into the subspace feature determination model, and determining the spatial interference feature information of each of the counties in the target area according to the output of the subspace feature determination model; The spatial interference characteristic information is input into the sub-time characteristic determination model, and the neighborhood interference characteristic information of each of the counties in the target area is determined according to the output of the sub-time characteristic determination model.

6. The method according to claim 1, characterized in that The determining, according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model, a regional predicted carbon emission amount corresponding to the target area at the next moment, includes: For each county, determine the carbon emissions at the current moment and the carbon emissions at the previous moment according to the time series carbon emissions information corresponding to the county; Inputting the carbon emissions at the current moment, the carbon emissions at the previous moment and the neighborhood interference feature information into the target carbon emissions prediction model for prediction processing; According to the output of the target carbon emission prediction model, the predicted carbon emission of the county corresponding to the next moment is obtained; According to the predicted carbon emissions of each of the counties, the predicted carbon emissions of the region corresponding to the target region at the next moment are determined.

7. The method according to claim 1, characterized in that The training process of the target carbon emission prediction model includes: Acquire historical time series carbon emissions, and determine a carbon emission training sample and a carbon emission verification sample corresponding to the carbon emission training sample according to the historical time series carbon emissions, wherein the carbon emission training sample includes carbon emissions at at least two consecutive moments; Inputting the carbon emission training samples and the predetermined neighborhood interference feature samples into a preset carbon emission prediction model to predict carbon emissions; Determining a training error of the preset carbon emission prediction model according to the carbon emission output information output by the preset carbon emission prediction model and the carbon emission verification sample; Back-propagating the training error to the preset carbon emission prediction model to adjust the network parameters in the preset carbon emission prediction model; When the preset convergence condition is met, it is determined that the training of the preset carbon emission prediction model is completed, and the target carbon emission prediction model is obtained.

8. A regional carbon emission prediction device, characterized in that: include: A target area information acquisition module, used to acquire the time series carbon emission information and the county time series characteristic matrix corresponding to each county in the target area, wherein the county time series characteristic matrix is ​​determined according to the county characteristics of the county; A county adjacency matrix determination module, used to construct a county adjacency matrix based on the geographical location relationship between the counties in the target area; An interference feature information determination module, used to determine the neighborhood interference feature information of each of the counties in the target area according to the county time series feature matrix, the county adjacency matrix and a pre-trained target interference feature determination model, wherein the target interference feature determination model includes a subspace feature determination model and a subtime feature determination model; The regional carbon emission prediction module is used to determine the regional predicted carbon emission corresponding to the target area at the next moment according to the time series carbon emission information, the neighborhood interference feature information and a pre-trained target carbon emission prediction model.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the regional carbon emission prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the regional carbon emission prediction method according to any one of claims 1 to 7 when executed.

Images