Coal mine carbon emission prediction method and system based on digital twinning and LSTM

Abstract

The invention discloses a coal mine carbon emission prediction method and system based on digital twinning and LSTM, and the method comprises the steps: building a carbon emission source information coding standard based on a business demand and a digital twinning model of a mining area in combination with a production link and an auxiliary production link of the mining area; performing carbon emission edge calculation on the carbon emission energy consumption data of the carbon emission source by using a coal mine carbon emission accounting method to obtain carbon emission data; establishing a corresponding mapping relation between the coal mine digital twinborn model and a database of carbon edge computing equipment by utilizing a carbon emission source information coding standard; according to the mapping relation, mapping the carbon emission data of each link to a coal mine digital twinborn model one by one; and establishing a carbon emission prediction model by using an LSTM algorithm, inputting historical actual carbon emission data into the model, and predicting carbon emission data of the current day. According to the method, the current carbon emission situation of each link of coal mine production and auxiliary production can be known, the carbon emission trend of each emission point is mastered in real time, and the emission reduction strength of enterprise energy-saving measures is quantified.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission prediction in various production and auxiliary production links of coal mines, and specifically to a coal mine carbon emission prediction method and system based on digital twin and LSTM. Background Art

[0002] High-energy-consuming production industries have accelerated the process of global warming. The rapid development of emerging technologies such as "big data, cloud computing, the Internet of Things, and mobile Internet" has promoted the deep integration of digital technology and energy technology, improved the digitalization and intelligence level in the fields of energy and carbon emissions, and changed the management mode in the fields of energy and carbon emissions. However, in the process of promoting digital transformation in the energy field, some difficulties have also been encountered, mainly manifested in the insufficient sharing of energy data and carbon emission data, and the lack of advanced digital technology monitoring means to grasp the energy consumption and carbon emissions of users in real time.

[0003] As a high-energy-consuming and high-carbon-emitting enterprise, coal mines have overall carbon emissions that are too high to be predicted and controlled. This is because the carbon emission data in various production and auxiliary production links of coal mines is difficult to monitor, and carbon emission data transparency has not yet been achieved, making it impossible to control the carbon emission data of coal mining enterprises in real time.

[0004] Based on the survey of the actual needs of coal mines, the following pain points in carbon emission monitoring in the production chain of the coal mining industry are summarized: (1) The carbon measurement system is not comprehensive enough, making it difficult to grasp the detailed carbon emissions in the coal production process; (2) There is a large space for energy conservation and emission reduction in the coal production process. Currently, there is a lack of carbon emission-related management systems to monitor and predict carbon emissions in real time, making it difficult to quantify emission reduction efforts, which is not conducive to the sustainable development of the coal mining industry; (3) Traditional carbon emission monitoring data has poor credibility and timeliness, data correlation is not high enough, and there is a lack of a sharing and access mechanism. Summary of the invention

[0005] The purpose of the present invention is to provide a coal mine carbon emission prediction method and system based on digital twins and LSTM, use carbon edge metering equipment to perform edge calculation, storage and classification of carbon emissions in various links of coal mines, use digital twin technology to establish a production process model, realize real-time monitoring of carbon data in each link, and combine the LSTM algorithm to predict the carbon emission data of each link, so as to help enterprises grasp the current status of carbon emissions, grasp the carbon emission situation and trend of each emission point in real time, quantify the emission reduction efforts of the enterprise's energy-saving measures, and ultimately achieve total carbon emissions control.

[0006] The present invention adopts the following technical solution: a coal mine carbon emission prediction method based on digital twin and LSTM, comprising the following steps:

[0007] S1. Based on the business needs and digital twin model of the mining area, the carbon emission source information coding standard is established in combination with the production links and auxiliary production links of the mining area.

[0008] S2. Use the coal mine carbon emission accounting method to perform carbon emission edge calculation on the carbon emission energy consumption data of the carbon emission source to obtain carbon emission data.

[0009] S3. Using the carbon emission source information coding standard in step S1, establish a mapping relationship between the coal mine digital twin model and the database of the carbon edge computing device.

[0010] S4. According to the mapping relationship obtained in step S3, the carbon emission data of each link is mapped one by one to the digital twin model of the coal mine.

[0011] S5. Use the LSTM algorithm to establish a carbon emission prediction model, input the historical actual carbon emission data into the model, and predict the carbon emission data for the day.

[0012] Furthermore, in step S1, establishing the carbon emission source information coding standard includes the following contents:

[0013] The production links of the mining area include geological support system, tunneling system, coal mining system, main transportation system, auxiliary transportation system, ventilation and compressed air system, water supply and drainage system and washing system; the auxiliary production links of the mining area include electricity consumption link and fuel energy consumption link.

[0014] Establish the linkage between carbon emission sources in the production links and auxiliary production links of the mining area, and obtain the carbon emission source information coding standard.

[0015] Furthermore, in step S2, the carbon emission data obtained includes the following:

[0016] A carbon edge computing device is set up at the energy consumption metering instrument end on the carbon emission source side of the mining area, and a carbon emission factor database and a coal mine carbon emission accounting method are established based on the carbon emission factor method, and the method is built into the carbon edge computing device; energy consumption data from the energy consumption metering instrument is collected in real time and stored in the database of the carbon edge computing device, and the coal mine carbon emission accounting method is used to perform edge computing on carbon emissions to obtain carbon emission data and store them in the database of the carbon edge computing device.

[0017] Furthermore, in step S3, establishing a mapping relationship includes the following contents:

[0018] Digital twin technology is used to establish a digital twin model of a coal mine, which includes a data acquisition system, a feedback system, an information synchronization system and a simulation system. The data acquisition system is used to keep the collected energy consumption data synchronized with the model in real time. The information synchronization system is used to maintain state synchronization between the twin system and the physical system. The feedback system promptly feeds back carbon emission data to management personnel.

[0019] Establish j groups of mapping relationships between each coded information and the database, where j represents the energy side attribute in the information coding. The mapping relationship is expressed as BPVj={BNi, BVi, PNi, PVi}, BNi represents the physical construction entity of the i-th device side attribute in the j-th group of mapping relationships, BVi represents the virtual model constructed by the i-th device side attribute in the j-th group of mapping relationships, PNi represents the coding information of the i-th device side attribute in the j-th group of mapping relationships, and PVi represents the carbon emission value corresponding to the i-th device side attribute code in the j-th group of mapping relationships.

[0020] Furthermore, in step S5, the predicted carbon emission data for the day includes the following contents:

[0021] S501. The carbon emission prediction model includes an input gate, an output gate and a forget gate.

[0022] Among them, the forget gate is used to determine the information to be deleted from the cell state. The specific expression is:

[0023] n t =σ(W n ·[h t-1 ,x t ]+b n );

[0024] The input gate is used to determine the new information retained in the cell state, and the specific expression is:

[0025] l t =σ(W1·[h t-1 ,x t ]+b1);

[0026] r t =tanh(W r ·[h t-1 ,x t ]+b r );

[0027] c t =n t *c t-1 +l t *r t ;

[0028] The output gate is used to determine the information in the cell state as output. The specific expression is:

[0029] m t =σ(W m ·[h t-1 ,x t ]+b m );

[0030] h t =m t *tanh(c t );

[0031] Among them, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, and x t represents the sample input at time t, W n , W1, W r , W m Both represent weight matrices, b n , b1, b r 、b m Both represent parameter matrices, n t represents the output of the forget gate, l t represents the input of the input gate, m t represents the output of the output gate, r t Indicates the information saved at the current moment, h t 、h t-1 Respectively represent the sample output at time t and time t-1, c t 、c t-1 They represent the sample states at time t and time t-1 respectively.

[0032] S502. The parameters of the carbon emission prediction model include hidden layer learning rate, time step, batch processing, number of hidden layers and number of training times, and the value range of the parameters is set.

[0033] S503. Extract historical carbon emission data from the database of the carbon edge computing device as a data set, and divide it into a training set and a test set; use the training set to train the carbon emission prediction model, and calculate the mean square error, root mean square error and mean absolute error. When the errors reach the set threshold, stop the training and obtain a trained carbon emission prediction model.

[0034] The test set is input into the trained carbon emission prediction model, and the mean square error, root mean square error and mean absolute error are calculated. When the errors reach the set threshold, it indicates that the model can be used to predict carbon emission data and obtain the final carbon emission prediction model. Otherwise, the data set is replaced and re-trained and tested.

[0035] S504, normalize the data set using min-max standardization, the specific formula is:

[0036]

[0037] Among them, x represents the original value of the data; x represents the normalized value; T min and T max Respectively represent the maximum and minimum values ​​of the data in the sample.

[0038] S505: Input the processed data set into the final carbon emission prediction model to obtain the predicted carbon emission data for the day.

[0039] Furthermore, the present invention also proposes a coal mine carbon emission prediction system based on digital twins and LSTM, including:

[0040] Energy consumption management module is used to measure, collect, transmit, count and analyze energy consumption data.

[0041] The carbon emission management module includes a carbon report generation module and a data visualization module. The carbon report generation module is used to obtain carbon emission data and produce carbon reports for coal mining enterprises; the data visualization module is used to use the coal mine digital twin model and combine the production links and auxiliary production links of the mining area to perform real-time visualization of carbon emission data.

[0042] The carbon emission prediction module includes a carbon emission prediction module and an over-standard warning module. The carbon emission prediction module is used to establish a carbon emission prediction model using the LSTM algorithm and perform real-time prediction of carbon emission data; the over-standard warning module is used to issue an alarm when the carbon emission prediction value exceeds the threshold.

[0043] The carbon analysis module is used to use big data analysis technology to conduct data analysis and mining on each link in the coal production chain. It combines the real-time mapping data of the digital twin to establish different carbon emission analysis models for different scenarios and explore the space for energy conservation and emission reduction in each link.

[0044] Furthermore, the present invention also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the coal mine carbon emission prediction method based on digital twins and LSTM when executing the computer program.

[0045] Furthermore, the present invention also proposes a computer-readable storage medium, which stores a computer program. When the computer program is run by a processor, it executes the coal mine carbon emission prediction method based on digital twins and LSTM.

[0046] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0047] (1) The present invention uses edge computing technology to embed the carbon emission accounting method into the carbon edge metering equipment, and performs edge computing, storage and classification on the carbon emissions of each link in the coal mine on the edge side, thereby realizing edge computing of carbon emission data, which can not only reduce the computing and storage pressure of the carbon chain management system, but also improve data security.

[0048] (2) The present invention uses digital twin technology to establish production process models for each link of coal production and auxiliary production, realizes real-time monitoring of carbon data in each link, ensures the effective real-time nature of carbon emission data, and realizes the visual display of carbon emission data.

[0049] (3) The present invention constructs a carbon emission prediction model based on the LSTM algorithm to predict the carbon emission data of each link, help to understand the current status of carbon emissions in each link of coal mine production and auxiliary production, grasp the carbon emission trend of each emission point in real time, quantify the emission reduction efforts of the enterprise's energy-saving measures, and ultimately achieve total carbon emission control. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is an overall implementation flow chart of the present invention.

[0051] Figure 2 It is a coal mine carbon emission source link relationship diagram of the present invention.

[0052] Figure 3 It is a flow chart of carbon emission edge computing of the present invention.

[0053] Figure 4 This is a structural diagram of the LSTM carbon emission prediction model of the present invention.

[0054] Figure 5 It is a structural diagram of the coal mine carbon emission prediction system of the present invention. DETAILED DESCRIPTION

[0055] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0056] To achieve the above objectives, the present invention proposes a coal mine carbon emission prediction method based on digital twin and LSTM, such as Figure 1 As shown, the specific steps are as follows:

[0057] S1. Based on the business application needs of the mining area throughout its life cycle and the application characteristics of the digital twin model, the carbon emission source information coding standard is established in combination with the production links and auxiliary production links of the mining area to meet the needs of carbon emission calculation; the specific contents are:

[0058] The main energy consumption in the mining area is electricity consumption generated by production equipment, and the rest includes coal, oil, gas and other energy consumption. Each energy consumption link corresponds to each carbon emission link. Establish the link relationship between carbon emission sources in each production and auxiliary production link of the coal mine, and obtain the carbon emission source information coding standard, such as Figure 2 shown.

[0059] The production links of the mining area are the various production subsystems involved in the coal production links, including geological support system, tunneling system, coal mining system, main transportation system, auxiliary transportation system, ventilation and compressed air system, water supply and drainage system and washing system.

[0060] Among them, the geological support system includes drilling equipment, geophysical equipment and remote sensing equipment, the tunneling system includes tunneling machines, anchor bolters, crushers and shuttle cars, the coal mining system includes coal mining machines, scraper conveyors, hydraulic supports, transfer machines and crushers, the main transportation system includes belt conveyors, belt conveyors and elevators, the auxiliary transportation system includes monorail cranes and trackless rubber-tyred vehicles, the ventilation and compressed air system includes main fans, local fans and air compressors, the water supply and drainage system includes pump rooms, control cabinets and transformers, and the washing system includes magnetic separators, concentrators, crushers, vibrating screens, filter presses and blowers.

[0061] The auxiliary production links in the mining area mainly involve the energy consumption generated by the living and office areas such as the ground office, living, and canteen of the coal mine, including electricity consumption and fuel consumption. The electricity consumption link includes the lighting system, elevator system, HVAC system, electric hot water system, etc., and the fuel consumption link includes coal-fired boiler room, gas boiler, and fuel for mining area vehicles.

[0062] The carbon emission sources of equipment in different energy consumption chains are encoded, and the attribute information of different energy sides is represented by numbers. As shown in Table 1, the attribute information of some carbon emission sources in coal mines is shown.

[0063] Table 1 Attribute information of some carbon emission sources in coal mines

[0064]

[0065]

[0066] S2. Using the carbon emission accounting method of coal mining enterprises, carbon emission energy consumption data of carbon emission sources are calculated at the edge of the carbon emission to obtain carbon emission data, which can realize functions such as energy consumption data collection, establishment of carbon emission accounting methods, and database establishment. Figure 3 As shown; the specific content is:

[0067] A carbon edge computing device is set up at the energy consumption metering instrument end on the carbon emission source side of the mining area, and a carbon emission factor database and a carbon emission accounting method for coal mining enterprises are established based on the carbon emission factor method in the "Guidelines for Greenhouse Gas Emission Accounting Methods and Reporting for Mining Enterprises (Trial)", and this method is built into the carbon edge computing device; energy consumption data from energy consumption metering instruments are collected in real time through the coal mine industrial ring network and stored in the database of the carbon edge computing device, and the carbon emission accounting method of coal mining enterprises is used to perform edge computing on carbon emissions, and the carbon emission data is obtained and stored in the database of the carbon edge computing device, thereby realizing real-time measurement and storage of carbon emission data.

[0068] S3. Using the carbon emission source information coding standard in step S1, a mapping relationship between the coal mine digital twin model and the database of the carbon edge computing device is established to achieve a one-to-one correspondence between the carbon emission source equipment and the digital twin model; the specific contents are:

[0069] Digital twin technology is used to establish a digital twin model of a coal mine, which includes a data acquisition system, a feedback system, an information synchronization system and a simulation system. The data acquisition system is used to keep the collected energy consumption data synchronized with the model in real time. The information synchronization system is used to keep the state synchronization between the twin system and the physical system. The feedback system promptly feeds back carbon emission data to management personnel, who then take corresponding energy-saving measures.

[0070] Establish j groups of mapping relationships between each coded information and the database, where j represents the energy side attributes in the information coding, including electricity, natural gas, oil, etc., and is composed of numbers. The mapping relationship is expressed as BPVj={BNi, BVi, PNi, PVi}, i represents the device side attributes in the information coding, and is composed of letters. BNi represents the physical construction entity of the i-th device side attribute in the j-th group of mapping relationships, BVi represents the virtual model constructed by the i-th device side attribute in the j-th group of mapping relationships, PNi represents the coding information of the i-th device side attribute in the j-th group of mapping relationships, and PVi represents the carbon emission value corresponding to the i-th device side attribute code in the j-th group of mapping relationships.

[0071] S4. According to the mapping relationship obtained in step S3, the carbon emission data of each link are mapped one by one to the digital twin model of the coal mine, so as to realize the visualization, dynamic monitoring and digital management of carbon emission data, and more intuitively display the carbon emission situation in each production link of the coal mining enterprise.

[0072] S5. Use the LSTM algorithm to establish a carbon emission prediction model, input the actual carbon emission data of the previous month into the model, and predict the carbon emission data of the day; the specific content is:

[0073] S501, such as Figure 4 As shown, the carbon emission prediction model includes an input gate, an output gate and a forget gate.

[0074] Among them, the forget gate is used to determine the information to be deleted from the cell state. The specific expression is:

[0075] n t =σ(W n ·[h t-1 ,x t ]+b n );

[0076] The input gate is used to determine the new information retained in the cell state, and the specific expression is:

[0077] l t =σ(W1·[h t-1 ,x t ]+b1);

[0078] r t =tanh(W r ·[h t-1 ,x t ]+b r );

[0079] c t =n t *c t-1 +l t *r t ;

[0080] The output gate is used to determine the information in the cell state as output. The specific expression is:

[0081] m t =σ(W m ·[h t-1 ,x t ]+b m );

[0082] h t =m t *tanh(c t );

[0083] Among them, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, and x t represents the sample input at time t, W n , W1, W r , W m Both represent weight matrices, b n , b1, b r , b m Both represent parameter matrices, n t represents the output of the forget gate, l t represents the input of the input gate, m t represents the output of the output gate, r t Indicates the information saved at the current moment, h t 、h t-1 Respectively represent the sample output at time t and time t-1, c t 、c t-1 They represent the sample states at time t and time t-1 respectively.

[0084] S502, determining the optimal parameters of the carbon emission prediction model:

[0085] The parameters of the carbon emission prediction model include hidden layer learning rate, time step, batch processing, number of hidden layers and number of training times, and the value range of the parameters is set.

[0086] In this embodiment, the number of hidden layers ranges from [1, 3], the time step ranges from [2, 4], the batch processing ranges from [20, 60], and the number of training times ranges from [25, 300].

[0087] (1) Hidden layer learning rate and training times

[0088] The hidden layer learning rate is used to indicate the degree to which the network weights are adjusted through the loss function, which affects the speed at which the model converges to the local minimum. When the learning rate is too small, the function converges slowly, takes longer, and overfitting occurs. When the learning rate is too large, the local minimum may be missed and the loss function is prone to oscillation. During the experiment, the value ranged from 0.01 to 0.00001, and was gradually reduced at 10 times the speed. The mean absolute error was selected as the error evaluation standard, and the state of the error curve was observed. When the curve tended to stabilize and converged to a certain value, the order of magnitude of the learning rate was determined, and then the fine-tuning method was used to determine the optimal learning rate.

[0089] Too few training times will lead to a large deviation between the predicted value and the actual value. Different training times will make the loss function converge to different values. Selecting an appropriate number of training times makes the error converge to the optimal value. During the training process, the mean square error is selected as the error evaluation standard.

[0090] (2) Time step

[0091] The time step indicates the number of historical states contained in each training sample. Since the carbon emission prediction model has the characteristic of long-term memory, the sample at the next moment may use the previous historical sample data. The time step is usually determined according to the time characteristics of the data. If the time step is n, it means that the total carbon emissions of the n+1th sample are predicted using the first n samples. In this embodiment, the time step range is set from 2 to 4, and the mean absolute error is used to measure the error change curve of the training samples at different time steps.

[0092] (3) Batch processing and number of hidden layers

[0093] Batch processing is used to define the number of samples to be processed before updating the internal model parameters. It represents the number of training samples each time. Batch processing is regarded as cyclically selecting one or more samples and making predictions, comparing the predicted values ​​with the true values, and calculating the errors, thereby dynamically updating the internal model parameters. Too small a batch size will lead to severe error oscillations and failure to converge within a certain number of training times. Increasing the batch size will make the training convergence smoother and more stable, and the time for training one generation will be shortened. However, too large a batch size will also reduce the generalization ability of the model, and the speed of achieving the same accuracy will be slower, so choose a suitable batch size to improve the prediction accuracy of the model.

[0094] The network structure of the hidden layer has a better representation ability for the data. The data is represented more abstractly through layer-by-layer feature extraction, thereby reducing errors and improving prediction accuracy. However, too many hidden layers may cause overfitting problems, and will also increase the training time of the model, resulting in failure to converge. This embodiment conducts experiments when different numbers of hidden layers and batch sizes are combined, and selects better parameters by comparing the mean absolute error, mean square error, and root mean square error under different parameters. The lower the mean absolute error, mean square error, and root mean square error, the higher the model accuracy.

[0095] S503, extract the historical carbon emission data of one month from the database of the carbon edge computing device as a data set, and divide it into a training set and a test set; the training set contains a large number of samples and corresponding labels, and the prediction logic or decision boundary of the model is established by learning the relationship between the features and targets in these data, so as to achieve the purpose of model training. The carbon emission prediction model is trained using the training set, and the mean square error, root mean square error and mean absolute error are calculated. When the errors reach more than 98%, the training is stopped to obtain a trained carbon emission prediction model.

[0096] The test set is a data set used to test the performance of the model, to evaluate whether the prediction results of the model are accurate, and to measure the accuracy of the prediction value using mean square error, root mean square error, and mean absolute error, thereby evaluating the accuracy of the model. The test set is input into the trained carbon emission prediction model, and the mean square error, root mean square error, and mean absolute error are calculated. When the errors are all above 98%, it indicates that the model can be used to predict carbon emission data and obtain the final carbon emission prediction model. Otherwise, the data set is replaced and retrained and tested.

[0097] S504, normalize the data set using min-max standardization, the specific formula is:

[0098]

[0099] Where x represents the original value of the data; x represents the normalized value, ranging from [0,1]; Tmin and T max Respectively represent the maximum and minimum values ​​of the data in the sample.

[0100] S505: Input the processed data set into the final carbon emission prediction model to obtain the predicted carbon emission data for the day.

[0101] The present invention establishes a coal mine carbon emission prediction system based on digital twin and LSTM through carbon emission management demand design, including four modules: energy consumption management, carbon emission management, carbon emission prediction, and carbon analysis.

[0102] The energy consumption management module is used to count energy consumption data and comprehensively measure, collect, transmit, count and analyze various energy data such as electricity, oil, heat, coal and gas.

[0103] The carbon emission management module includes a carbon report generation module and a data visualization module. The carbon report generation module is used to obtain carbon emission data from the carbon edge database and produce carbon reports for coal mining enterprises; the data visualization module is used to use the digital twin model to combine coal production and auxiliary production virtual scenes to perform real-time visualization of carbon emission data.

[0104] The carbon emission prediction module includes a carbon emission prediction module and an over-standard warning module. The carbon emission prediction module is used to establish a prediction model using the LSTM algorithm to make real-time predictions on carbon emission data. The over-standard warning module is used to send out an alarm when the predicted carbon emission value exceeds the threshold, indicating that the enterprise may have abnormal energy consumption, excessive carbon emissions, etc. This helps the enterprise to check abnormal points in each link and achieve the purpose of emission control.

[0105] The carbon analysis module is used to analyze and mine data in each link of the coal production chain using big data analysis technology, and to establish different carbon emission analysis models for different scenarios in combination with the real-time mapping data of digital twins, and to actively explore the energy-saving and emission reduction space in each link. Specifically, it includes:

[0106] 1) Carbon emission analysis in production links: Collect energy consumption and carbon emission data of each production link subsystem such as coal mining system, transportation system, washing system, etc. Use data mining in big data analysis methods to mine the carbon emission data under the collaborative mode between related equipment of each subsystem, and then compare it with historical data to determine the abnormal carbon emission situation in each production link.

[0107] 2) Unit carbon emission analysis: Taking the teams and work groups in each production link as the unit, the carbon emissions of different teams and work groups are compared through comparative analysis, and the teams or work groups with abnormal energy consumption are analyzed in real time.

[0108] 3) Carbon emission analysis at different stages: Taking the weekly, monthly, and quarterly time periods of each subsystem in the production process as the unit, the indicator analysis method is used to calculate the average carbon emission value in different time periods, the time periods with the highest and lowest carbon emissions, etc., analyze the carbon emission patterns in different time stages, and combine with the production stage plan to provide data support for decision makers.

[0109] 4) Equipment carbon emission analysis: Establish a mathematical model between carbon emissions and different influencing factors, analyze the relationship between carbon emissions and various influencing factors, determine the weight of each influencing factor, monitor the carbon emission data of each sampling point of energy-consuming equipment in real time during actual operation, explore the carbon emission characteristics of each equipment, and promptly discover and report abnormal carbon emissions of equipment.

[0110] It should be noted that each module in the above system corresponds to the specific steps of the method provided in the embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0111] The embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. It should be noted that when the processor executes the computer program, the specific steps of the method provided in the embodiment of the present invention are corresponding to the specific steps of the method provided in the embodiment of the present invention, and the processor has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0112] The embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. It should be noted that when the computer program is executed by the processor, it corresponds to the specific steps of the method provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0113] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A coal mine carbon emission prediction method based on digital twins and LSTM, characterized by: include: S1. Based on the business needs and digital twin model of the mining area, the carbon emission source information coding standard is established in combination with the production links and auxiliary production links of the mining area; S2. Using the coal mine carbon emission accounting method, perform carbon emission edge calculation on the carbon emission energy consumption data of the carbon emission source to obtain carbon emission data; S3. Using the carbon emission source information coding standard in step S1, establish a mapping relationship between the coal mine digital twin model and the database of the carbon edge computing device; S4. According to the mapping relationship obtained in step S3, the carbon emission data of each link is mapped one by one to the digital twin model of the coal mine; S5. Use the LSTM algorithm to establish a carbon emission prediction model, input the historical actual carbon emission data into the model, and predict the carbon emission data for the day.

2. The coal mine carbon emission prediction method based on digital twin and LSTM according to claim 1 is characterized in that: In step S1, establishing the carbon emission source information coding standard includes the following contents: The production links of the mining area include geological support system, tunneling system, coal mining system, main transportation system, auxiliary transportation system, ventilation and compressed air system, water supply and drainage system and washing system; the auxiliary production links of the mining area include power consumption link and fuel energy consumption link; Establish the linkage between carbon emission sources in the production links and auxiliary production links of the mining area, and obtain the carbon emission source information coding standard.

3. The coal mine carbon emission prediction method based on digital twin and LSTM according to claim 1 is characterized in that: In step S2, the carbon emission data obtained includes the following: Set up carbon edge computing equipment at the energy consumption metering instrument end of the carbon emission source side of the mining area, establish a carbon emission factor database and coal mine carbon emission accounting method based on the carbon emission factor method, and build the method into the carbon edge computing equipment; The energy consumption data from the energy consumption meter is collected in real time and stored in the database of the carbon edge computing device. The carbon emissions are edge calculated using the coal mine carbon emission accounting method to obtain the carbon emission data and store it in the database of the carbon edge computing device.

4. The coal mine carbon emission prediction method based on digital twin and LSTM according to claim 1 is characterized in that: In step S3, establishing a mapping relationship includes the following: Use digital twin technology to build a digital twin model of a coal mine, which includes a data acquisition system, a feedback system, an information synchronization system, and a simulation system. Use the data acquisition system to keep the collected energy consumption data synchronized with the model in real time, use the information synchronization system to keep the status between the twin system and the physical system synchronized, and use the feedback system to promptly feed back carbon emission data to management personnel. Establish j groups of mapping relationships between each coded information and the database, where j represents the energy side attribute in the information coding. The mapping relationship is expressed as BPVj={BNi, BVi, PNi, PVi}, BNi represents the physical construction entity of the i-th device side attribute in the j-th group of mapping relationships, BVi represents the virtual model constructed by the i-th device side attribute in the j-th group of mapping relationships, PNi represents the coding information of the i-th device side attribute in the j-th group of mapping relationships, and PVi represents the carbon emission value corresponding to the i-th device side attribute code in the j-th group of mapping relationships.

5. The coal mine carbon emission prediction method based on digital twin and LSTM according to claim 1 is characterized in that: In step S5, the carbon emission data for the day is predicted to include the following: S501, the carbon emission prediction model includes an input gate, an output gate and a forget gate; Among them, the forget gate is used to determine the information to be deleted from the cell state. The specific expression is: n t =σ(W n ·[h t-1 ,x t ]+b n ); The input gate is used to determine the new information retained in the cell state, and the specific expression is: l t =σ(W1·[h t-1 ,x t ]+b1); r t =tanh(W r ·[h t-1 ,x t ]+b r ); c t =n t *c t-1 +l t *r t ; The output gate is used to determine the information in the cell state as output. The specific expression is: m t =σ(W m ·[h t-1 ,x t ]+b m ); h t =m t *tanh(c t ); Among them, σ represents the sigmoid function, tanh represents the hyperbolic tangent function, and x t represents the sample input at time t, W n , W1, W r , W m Both represent weight matrices, b n , b1, b r 、b m Both represent parameter matrices, n t represents the output of the forget gate, l t represents the input of the input gate, m t represents the output of the output gate, r t Indicates the information saved at the current moment, h t 、h t-1 Respectively represent the sample output at time t and time t-1, c t 、c t-1 Represent the sample states at time t and time t-1 respectively; S502, the parameters of the carbon emission prediction model include hidden layer learning rate, time step, batch processing, number of hidden layers and number of training times, and the value range of the parameters is set; S503, extracting historical carbon emission data from the database of the carbon edge computing device as a data set, and dividing it into a training set and a test set; using the training set to train the carbon emission prediction model, and calculating the mean square error, root mean square error and mean absolute error, when the errors reach the set threshold, the training is stopped to obtain a trained carbon emission prediction model; The test set is input into the trained carbon emission prediction model, and the mean square error, root mean square error and mean absolute error are calculated. When the errors reach the set threshold, it indicates that the model can be used to predict carbon emission data and obtain the final carbon emission prediction model. Otherwise, the data set is replaced and retrained and tested. S504, normalize the data set using min-max standardization, the specific formula is: Among them, x represents the original value of the data; x represents the normalized value; T min and T max Respectively represent the maximum and minimum values ​​of the data in the sample; S505: Input the processed data set into the final carbon emission prediction model to obtain the predicted carbon emission data for the day.

6. A system for predicting coal mine carbon emissions based on digital twins and LSTM as claimed in claim 1, characterized in that: include: Energy consumption management module, used for measuring, collecting, transmitting, counting and analyzing energy consumption data; The carbon emission management module includes a carbon report generation module and a data visualization module. The carbon report generation module is used to obtain carbon emission data and produce carbon reports for coal mining enterprises. The data visualization module is used to use the coal mine digital twin model to combine the production links and auxiliary production links of the mining area to perform real-time visualization of carbon emission data. The carbon emission prediction module includes a carbon emission prediction module and an over-standard warning module. The carbon emission prediction module is used to establish a carbon emission prediction model using the LSTM algorithm to make real-time predictions of carbon emission data; the over-standard warning module is used to issue an alarm when the carbon emission prediction value exceeds the threshold; The carbon analysis module is used to use big data analysis technology to conduct data analysis and mining on each link in the coal production chain. It combines the real-time mapping data of the digital twin to establish different carbon emission analysis models for different scenarios and explore the space for energy conservation and emission reduction in each link.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the coal mine carbon emission prediction method based on digital twin and LSTM as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, the coal mine carbon emission prediction method based on digital twin and LSTM described in any one of claims 1 to 5 is executed.

Images