An artificial intelligence-based factory carbon emission monitoring and prediction system

Abstract

A factory carbon emission monitoring and prediction system based on artificial intelligence relates to the technical field of carbon emission monitoring. The system divides the factory's overall production process into different production sub-processes, obtains basic information and equipment data of industrial equipment corresponding to each production sub-process, constructs a digital twin model of each production sub-process, obtains the difference set of operation data and emission data of adjacent production sub-processes, constructs a first evaluation model for each production sub-process and a second evaluation model for adjacent production sub-processes, obtains real-time equipment data of each production sub-process, and judges whether the corresponding production sub-process is in an abnormal state by combining the first evaluation model and the second evaluation model. The system can analyze the carbon emission impact relationship between adjacent production sub-processes, effectively judge whether each current production sub-process is in an abnormal state and what kind of abnormal state exists, and provide timely feedback.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission monitoring, and in particular to an artificial intelligence-based factory carbon emission monitoring and prediction system. Background Art

[0002] Using artificial intelligence to monitor and predict factory carbon emissions is an emerging technology. It aims to utilize advanced sensor technology, data acquisition and transmission technology, and artificial intelligence algorithms to achieve real-time monitoring, precise analysis, and scientific prediction of factory carbon emissions. This can help factory managers fully understand carbon emissions and enhance the company's sustainable development capabilities.

[0003] Existing technologies often monitor carbon emissions by considering the entire production process as a whole. This monitoring method is often inaccurate and cannot understand the carbon emissions of each process step. Due to the connection between different equipment, the quality of products processed by the front equipment will directly affect the carbon emissions of the back equipment during processing.

[0004] For example, if the former is not fully reacted or the processing is not thorough, the latter's processing process will inevitably become more energy-consuming, thereby making the carbon emissions of each process link more complicated. The existing technology lacks the technical means to analyze the carbon emission impact between adjacent process links and monitor their abnormalities. In response to the shortcomings of the existing technology, the present invention provides a factory carbon emission monitoring and prediction system based on artificial intelligence. Summary of the Invention

[0005] The purpose of the present invention is to provide a factory carbon emission monitoring and prediction system based on artificial intelligence.

[0006] The purpose of the present invention can be achieved through the following technical solution: A factory carbon emission monitoring and prediction system based on artificial intelligence, including the following modules:

[0007] The process division module is used to obtain the overall production process of the factory and divide it into different production sub-processes, and obtain the basic information and equipment data of the industrial equipment corresponding to each production sub-process;

[0008] The data simulation module is used to build a digital twin model of the corresponding production sub-process based on the equipment data of each industrial equipment, and obtain the difference sets of operation data and corresponding emission data between adjacent production sub-processes;

[0009] A model building module is used to build a first evaluation model for each production sub-process based on equipment data, and to build a second evaluation model for adjacent production sub-processes based on a difference set of operation data and a difference set of emission data;

[0010] The data prediction module is used to obtain real-time equipment data of each production sub-process, and combine the first evaluation model and the second evaluation model to determine whether the corresponding production sub-process has an abnormal state and provide feedback.

[0011] Furthermore, the process of obtaining the overall production process of the factory and dividing it into different production sub-processes includes:

[0012] The overall production process refers to the entire process from raw material processing to finished products in the factory, including different process links and their corresponding industrial equipment, and the production processes of industrial equipment corresponding to different process links are respectively regarded as production sub-processes.

[0013] Furthermore, the process of obtaining basic information and equipment data of industrial equipment corresponding to each production sub-process includes:

[0014] The basic information refers to the specifications of the industrial equipment corresponding to each production sub-process, and a three-dimensional modeling tool is used to construct a physical model of the corresponding production sub-process according to the specifications of each industrial equipment;

[0015] In actual application scenarios, an operation collection unit and an emission collection unit are respectively set for each industrial equipment, and corresponding equipment operation data and equipment emission data are respectively obtained, wherein the equipment data includes equipment operation data and equipment emission data;

[0016] The equipment operation data refers to the pressure, load rate, temperature, vibration amplitude, current, voltage, input rate, and output rate of the industrial equipment, and the equipment emission data refers to the energy consumption parameters and emission gas parameters of the industrial equipment.

[0017] Furthermore, the process of constructing a digital twin model of the corresponding production sub-process based on the equipment data of each industrial equipment includes:

[0018] Using simulation software to simulate the working process of industrial equipment based on the physical model of a single production sub-process to obtain a corresponding simulation model, and uploading the equipment operation data of the industrial equipment to the simulation model for synchronization;

[0019] Continuously adjusting the values ​​of various simulation parameters during the simulation process, including inertia parameters, electromagnetic parameters, thermodynamic parameters, environmental parameters, and performance degradation parameters, to obtain energy consumption parameters and emission gas parameters in the simulation model under different values ​​of simulation parameters;

[0020] When the energy consumption parameters and emission gas parameters obtained are the same as the equipment emission data at the corresponding moment of the synchronized equipment operation data, the simulation model at this time is used as the digital twin model of the production sub-process, and the simulation parameters with corresponding values ​​are used as digital twin parameters to obtain the digital twin model and digital twin parameters of each production sub-process respectively.

[0021] Furthermore, the process of respectively obtaining the difference sets of operation data and the corresponding difference sets of emission data between adjacent production sub-processes includes:

[0022] The digital twin models of each production sub-process are integrated in sequence according to the order of the process links to obtain the digital twin model of the overall production process. Any two adjacent production sub-processes in the process links are marked as a group of adjacent production sub-processes.

[0023] The digital twin parameters in any adjacent production sub-process and the equipment operation data of the subsequent production sub-process remain unchanged, and the equipment operation data of the previous production sub-process is continuously adjusted. The difference between the various parameters in the equipment operation data of the previous production sub-process before and after each adjustment is used as the operation data difference set;

[0024] In the digital twin overall model, the difference in the various parameters in the equipment emission data of the latter production sub-process before and after each adjustment is used as the emission data difference set, and the different operation data difference sets and emission data difference sets between each group of adjacent production sub-processes are obtained respectively.

[0025] Furthermore, the process of constructing the first evaluation model of each production sub-process based on the equipment data includes:

[0026] Based on the different equipment operation data in the equipment data of a single production sub-process and the equipment emission data at the corresponding time, combined with the digital twin parameters of the production sub-process, a first evaluation set is generated, and the first evaluation set is divided into a first training set and a first test set;

[0027] Construct a first convolutional neural network, use the digital twin parameters and different equipment operation data in the first training set as input data of the first convolutional neural network, and use the corresponding equipment emission data in the first training set as output data of the first convolutional neural network;

[0028] The first convolutional neural network is trained to obtain an initial first convolutional neural network, and the initial first convolutional neural network is model verified using the first test set. The initial first convolutional neural network with a first test error threshold that is less than or equal to a preset first test error threshold is output as the first evaluation model, and the first evaluation model of each production sub-process is constructed respectively.

[0029] Furthermore, the process of constructing a second evaluation model of the adjacent production sub-process based on the difference set of operation data and the difference set of emission data includes:

[0030] According to the different operation data difference sets and emission data difference sets between any adjacent production sub-processes, combined with the equipment operation data of the latter production sub-process, a second evaluation set is generated, and the second evaluation set is divided into a second training set and a second test set;

[0031] Constructing a second convolutional neural network, using the equipment operation data in the second training set and different operation data difference sets as input data of the second convolutional neural network, and using the corresponding emission data difference sets in the second training set as output data of the second convolutional neural network;

[0032] The second convolutional neural network is trained to obtain an initial second convolutional neural network, and the initial second convolutional neural network is verified using the second test set. The initial second convolutional neural network with a second test error threshold that is less than or equal to a preset second test error threshold is output as the second evaluation model, and the second evaluation model of each group of adjacent production sub-processes is constructed respectively.

[0033] Furthermore, the process of obtaining real-time equipment data of each production sub-process, combining the first evaluation model and the second evaluation model to determine whether the corresponding production sub-process is in an abnormal state and providing feedback includes:

[0034] The real-time equipment data includes real-time operation data and real-time emission data. The current equipment operation data and equipment emission data of each production sub-process are used as its real-time operation data and real-time emission data respectively. The digital twin parameters and real-time operation data of a single production sub-process are input into the first evaluation model to obtain theoretical emission data of the production sub-process;

[0035] Comparing the theoretical emission data of the production sub-process with various parameters in its real-time emission data to obtain parameters in a first abnormal state, and generating a first abnormal signal for feedback, respectively obtaining the theoretical emission data of each production sub-process and determining whether it is in the first abnormal state;

[0036] The difference between the current equipment operation data of the previous production sub-process and the parameters of the equipment operation data at the previous moment in any adjacent production sub-process is used as its real-time operation data difference set, and combined with the real-time operation data of the next production sub-process, they are input into the second evaluation model to obtain the theoretical emission data difference set of the next production sub-process;

[0037] The parameters in the second abnormal state in the theoretical emission data difference set of the subsequent production sub-process are obtained, and a second abnormal signal is generated for feedback. The theoretical emission data difference set of the subsequent production sub-process in each group of adjacent production sub-processes is obtained respectively, and it is determined whether it is in the second abnormal state.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] By dividing the entire production process into different production sub-processes, the present invention can conduct targeted monitoring of the carbon emissions of each production sub-process and analyze the carbon emission impact relationship between adjacent production sub-processes. By constructing a digital twin model and obtaining digital twin parameters, it can effectively obtain the aging status of industrial equipment corresponding to each production sub-process in actual application scenarios, and can continuously simulate the data basis for subsequent model construction;

[0040] By constructing the first evaluation module and the second evaluation module respectively, the equipment emission data of a single production sub-process under different equipment operation data can be effectively obtained, and the impact of the change in the operating parameters of the previous production sub-process on the emission parameters of the next production sub-process can be effectively obtained. Combined with the real-time operation data of each production sub-process, it is possible to effectively determine whether each production sub-process has an abnormal state and what kind of abnormal state exists and provide timely feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a module schematic diagram of the present invention. DETAILED DESCRIPTION

[0042] like Figure 1 As shown in the figure, an artificial intelligence-based factory carbon emission monitoring and prediction system includes the following modules:

[0043] The process division module is used to obtain the overall production process of the factory and divide it into different production sub-processes, and obtain the basic information and equipment data of the industrial equipment corresponding to each production sub-process;

[0044] The data simulation module is used to build a digital twin model of the corresponding production sub-process based on the equipment data of each industrial equipment, and obtain the difference sets of operation data and corresponding emission data between adjacent production sub-processes;

[0045] A model building module is used to build a first evaluation model for each production sub-process based on equipment data, and to build a second evaluation model for adjacent production sub-processes based on a difference set of operation data and a difference set of emission data;

[0046] The data prediction module is used to obtain real-time equipment data of each production sub-process, and combine the first evaluation model and the second evaluation model to determine whether the corresponding production sub-process has an abnormal state and provide feedback.

[0047] It should be further explained that, in the specific implementation process, the process of obtaining the overall production process of the factory and dividing it into different production sub-processes includes:

[0048] The overall production process refers to the entire process from raw materials to finished products in a factory, which involves different process steps, and different process steps correspond to different industrial equipment. For example, the steel production process includes ironmaking, steelmaking, continuous casting, hot rolling, cold rolling, heat treatment, surface treatment, and finishing.

[0049] The above-mentioned process links correspond to the blast furnace, converter, continuous casting machine, hot rolling mill, cold rolling mill, continuous annealing furnace, pickling line, and laser cutting machine respectively. The production process of the industrial equipment corresponding to the different process links is regarded as a production sub-process. This method can divide the complete production process into several different production sub-processes.

[0050] It should be further explained that, in the specific implementation process, the process of obtaining the basic information and equipment data of the industrial equipment corresponding to each production sub-process includes:

[0051] The basic information refers to the specifications of the industrial equipment corresponding to each production sub-process, including various relevant data required for building a physical model. A 3D modeling tool is used to build a physical model of each production sub-process based on the specifications of each industrial equipment.

[0052] In actual application scenarios, a corresponding operation collection unit and an emission collection unit are respectively set for each industrial equipment, and the equipment operation data of the industrial equipment is collected in real time by the operation collection unit, and the equipment emission data of the industrial equipment is collected in real time by the emission collection unit. The equipment data includes equipment operation data and equipment emission data;

[0053] The equipment operation data refers to the pressure, load rate, temperature, vibration amplitude, current, voltage, input rate, and output rate of industrial equipment. The equipment operation data of different industrial equipment are not the same. The equipment emission data refers to the energy consumption parameters and emission gas parameters of industrial equipment. The energy consumption parameters include electricity consumption and fuel consumption. The emission gas parameters include the concentration and flow rate of gases such as CO2, CH4, N2O, SF6, SO2, etc. The equipment emission data of different industrial equipment are not the same.

[0054] It should be further explained that, in the specific implementation process, the process of building a digital twin model of the corresponding production sub-process based on the equipment data of each industrial equipment includes:

[0055] Taking the physical model of a single production sub-process as an example, simulation software is used to simulate the working process of the industrial equipment based on the physical model to obtain the corresponding simulation model. The equipment operation data of the industrial equipment is uploaded to the simulation model for synchronization, and the values ​​of various simulation parameters in the simulation process are adjusted;

[0056] The simulation parameters refer to the inertia parameters, electromagnetic parameters, thermodynamic parameters, environmental parameters, and performance degradation parameters corresponding to industrial equipment in actual application scenarios. In short, the simulation parameters include all adjustable parameters except equipment operation data and equipment emission data, and the energy consumption parameters and emission gas parameters in the simulation model under different simulation parameter values ​​are obtained;

[0057] When the energy consumption parameters and emission gas parameters obtained are the same as the equipment emission data at the corresponding moment of the synchronized equipment operation data, the simulation model at this time is used as the digital twin model of the production sub-process, and the simulation parameters of the corresponding values ​​are used as the digital twin parameters of the digital twin model. The same method is used to obtain the digital twin model and digital twin parameters of each production sub-process respectively.

[0058] It should be further explained that, in the specific implementation process, the process of obtaining the difference sets of operation data and the corresponding difference sets of emission data between adjacent production sub-processes includes:

[0059] The digital twin models of each production sub-process are integrated in sequence according to the order of the process links to obtain the digital twin model of the overall production process. Any two adjacent production sub-processes in the process links are marked as a group of adjacent production sub-processes.

[0060] Taking any adjacent production sub-processes as an example, the digital twin parameters of two of the production sub-processes and the equipment operation data of the latter production sub-process are kept fixed, and the equipment operation data of the previous production sub-process is continuously adjusted. The difference in the various parameters in the equipment operation data of the previous production sub-process before and after each adjustment is used as the corresponding operation data difference set;

[0061] In the digital twin overall model, the difference in various parameters in the equipment emission data of the subsequent production sub-process before and after each adjustment is obtained, and used as the corresponding emission data difference set. The operation data difference set and the emission data difference set corresponding to each adjustment are bound together, and the same method is used to obtain the different operation data difference sets and their corresponding emission data difference sets between each group of adjacent production sub-processes.

[0062] It should be further explained that, in the specific implementation process, the process of constructing the first evaluation model of each production sub-process based on the equipment data includes:

[0063] Taking the equipment data of a single production sub-process as an example, based on the different equipment operation data and the equipment emission data at the corresponding time in the equipment data, combined with the digital twin parameters of the production sub-process, a corresponding first evaluation set is generated and divided into a first training set and a first test set;

[0064] Construct a first convolutional neural network, use the digital twin parameters and different equipment operation data in the first training set as input data of the first convolutional neural network, and use the corresponding equipment emission data in the first training set as output data of the first convolutional neural network;

[0065] The first convolutional neural network is trained to obtain an initial first convolutional neural network, the initial first convolutional neural network is model verified using a first test set, and the initial first convolutional neural network that is less than or equal to a preset first test error threshold is output as the corresponding first evaluation model.

[0066] It should be further explained that, in the specific implementation process, the process of constructing the second evaluation model of the adjacent production sub-process based on the difference set of operation data and the difference set of emission data includes:

[0067] Taking any adjacent production sub-processes as an example, based on the different operation data difference sets and emission data difference sets between the adjacent production sub-processes, combined with the equipment operation data of the latter production sub-process, a corresponding second evaluation set is generated, and the second evaluation set is divided into a second training set and a second test set;

[0068] Constructing a second convolutional neural network, using the equipment operation data in the second training set and different operation data difference sets as input data of the second convolutional neural network, and using the corresponding emission data difference sets in the second training set as output data of the second convolutional neural network;

[0069] The second convolutional neural network is trained to obtain an initial second convolutional neural network, the initial second convolutional neural network is model verified using the second test set, and the initial second convolutional neural network that is less than or equal to a preset second test error threshold is output as the corresponding second evaluation model.

[0070] It should be further explained that, in the specific implementation process, the process of obtaining real-time equipment data of each production sub-process, combining the first evaluation model and the second evaluation model to determine whether the corresponding production sub-process has an abnormal state and providing feedback includes:

[0071] The current equipment operation data and equipment emission data of each production sub-process are used as its real-time operation data and real-time emission data, respectively. The real-time equipment data includes real-time operation data and real-time emission data. The digital twin parameters and real-time operation data of a single production sub-process are input into the corresponding first evaluation model to obtain the theoretical emission data of the production sub-process;

[0072] Compare the theoretical emission data of the production sub-process with its real-time emission data. If the difference between the parameters of the two is greater than the corresponding first preset threshold, mark the corresponding parameter as a first abnormal state, generate a first abnormal signal and feed it back to relevant personnel. The first preset threshold varies for different parameters.

[0073] The first abnormal signal is used to reflect that the carbon emissions of the corresponding production sub-process are significantly abnormal in the actual application scenario, and is used to notify relevant personnel to inspect and repair the production sub-process. The same method is adopted to obtain the theoretical emission data of each production sub-process and determine whether it is in the first abnormal state;

[0074] The difference between the current equipment operation data of the previous production sub-process and the parameters of the equipment operation data at the previous moment in any adjacent production sub-process is used as its real-time operation data difference set, and combined with the real-time operation data of the next production sub-process, they are input into the second evaluation model to obtain the theoretical emission data difference set of the next production sub-process;

[0075] Compare each parameter in the theoretical emission data difference set of the subsequent production sub-process with the corresponding second preset threshold value. If there is a parameter that is greater than the second preset threshold value, the corresponding parameter is marked as a second abnormal state, and a second abnormal signal is generated and fed back to relevant personnel. The second preset threshold value varies for different parameters.

[0076] The second abnormal signal is used to reflect that there is an obvious abnormal change in the carbon emissions of the subsequent production sub-process in the actual application scenario, and is used to notify relevant personnel to adjust the equipment operation data of the previous production sub-process. The same method is used to obtain the difference set of theoretical emission data of the subsequent production sub-process in each group of adjacent production sub-processes, and to determine whether there is a second abnormal state.

[0077] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An artificial intelligence-based factory carbon emission monitoring and prediction system, characterized by: Includes the following modules: The process division module is used to obtain the overall production process of the factory and divide it into different production sub-processes, and obtain the basic information and equipment data of the industrial equipment corresponding to each production sub-process, including equipment operation data and equipment emission data; The data simulation module is used to build a digital twin model of the corresponding production sub-process based on the equipment data of each industrial equipment and obtain the digital twin parameters, and obtain the operation data difference set and the corresponding emission data difference set between adjacent production sub-processes respectively; A model building module is used to build a first evaluation model for each production sub-process based on equipment data, and to build a second evaluation model for adjacent production sub-processes based on a difference set of operation data and a difference set of emission data; The data prediction module is used to obtain real-time equipment data of each production sub-process, and combine the first evaluation model and the second evaluation model to determine whether the corresponding production sub-process has an abnormal state and provide feedback; The process of constructing the first evaluation model for each production sub-process includes: Based on the different equipment operation data in the equipment data of a single production sub-process and the equipment emission data at the corresponding time, combined with the digital twin parameters of the production sub-process, a first evaluation set is generated, and the first evaluation set is divided into a first training set and a first test set; Construct a first convolutional neural network, use the digital twin parameters and different equipment operation data in the first training set as input data of the first convolutional neural network, and use the corresponding equipment emission data in the first training set as output data of the first convolutional neural network; The first convolutional neural network is trained to obtain an initial first convolutional neural network, and the initial first convolutional neural network is model verified using the first test set. The initial first convolutional neural network with a first test error threshold that is less than or equal to a preset first test error threshold is output as the first evaluation model, and the first evaluation model of each production sub-process is constructed respectively.

2. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 1 is characterized in that: The process of dividing the overall production process into production sub-processes includes: The overall production process refers to the entire process from raw material processing to finished products in the factory, including different process links and their corresponding industrial equipment, and the production processes of industrial equipment corresponding to different process links are respectively regarded as production sub-processes.

3. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 2 is characterized in that: The process of obtaining basic information and equipment data for each production sub-process includes: The basic information refers to the specifications of the industrial equipment corresponding to each production sub-process, and a three-dimensional modeling tool is used to construct a physical model of the corresponding production sub-process according to the specifications of each industrial equipment; In actual application scenarios, an operation collection unit and an emission collection unit are set up for each industrial equipment, and the corresponding equipment operation data and equipment emission data are obtained respectively; The equipment operation data refers to the pressure, load rate, temperature, vibration amplitude, current, voltage, input rate, and output rate of the industrial equipment, and the equipment emission data refers to the energy consumption parameters and emission gas parameters of the industrial equipment.

4. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 3 is characterized in that: The process of building a digital twin model of each production sub-process includes: Using simulation software to simulate the working process of industrial equipment based on the physical model of a single production sub-process to obtain a corresponding simulation model, and uploading the equipment operation data of the industrial equipment to the simulation model for synchronization; Continuously adjusting the values ​​of various simulation parameters during the simulation process, including inertia parameters, electromagnetic parameters, thermodynamic parameters, environmental parameters, and performance degradation parameters, to obtain energy consumption parameters and emission gas parameters in the simulation model under different values ​​of simulation parameters; When the energy consumption parameters and emission gas parameters obtained are the same as the equipment emission data at the corresponding moment of the synchronized equipment operation data, the simulation model at this time is used as the digital twin model of the production sub-process, and the simulation parameters with corresponding values ​​are used as digital twin parameters to obtain the digital twin model and digital twin parameters of each production sub-process respectively.

5. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 4 is characterized in that: The process of obtaining the difference sets of operation data and emission data of adjacent production sub-processes includes: The digital twin models of each production sub-process are integrated in sequence according to the order of the process links to obtain the digital twin model of the overall production process. Any two adjacent production sub-processes in the process links are marked as a group of adjacent production sub-processes. The digital twin parameters in any adjacent production sub-process and the equipment operation data of the subsequent production sub-process remain unchanged, and the equipment operation data of the previous production sub-process is continuously adjusted. The difference between the various parameters in the equipment operation data of the previous production sub-process before and after each adjustment is used as the operation data difference set; In the digital twin overall model, the difference in the various parameters in the equipment emission data of the latter production sub-process before and after each adjustment is used as the emission data difference set, and the different operation data difference sets and emission data difference sets between each group of adjacent production sub-processes are obtained respectively.

6. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 5 is characterized in that: The process of constructing the second evaluation model of the adjacent production sub-process includes: According to the different operation data difference sets and emission data difference sets between any adjacent production sub-processes, combined with the equipment operation data of the latter production sub-process, a second evaluation set is generated, and the second evaluation set is divided into a second training set and a second test set; Constructing a second convolutional neural network, using the equipment operation data in the second training set and different operation data difference sets as input data of the second convolutional neural network, and using the corresponding emission data difference sets in the second training set as output data of the second convolutional neural network; The second convolutional neural network is trained to obtain an initial second convolutional neural network, and the initial second convolutional neural network is verified using the second test set. The initial second convolutional neural network with a second test error threshold that is less than or equal to a preset second test error threshold is output as the second evaluation model, and the second evaluation model of each group of adjacent production sub-processes is constructed respectively.

7. The artificial intelligence-based factory carbon emission monitoring and prediction system according to claim 6 is characterized in that: The process of obtaining real-time device data, determining whether there is an abnormal state, and providing feedback includes: The real-time equipment data includes real-time operation data and real-time emission data. The current equipment operation data and equipment emission data of each production sub-process are used as its real-time operation data and real-time emission data respectively. The digital twin parameters and real-time operation data of a single production sub-process are input into the first evaluation model to obtain theoretical emission data of the production sub-process; Comparing the theoretical emission data of the production sub-process with various parameters in its real-time emission data to obtain parameters in a first abnormal state, and generating a first abnormal signal for feedback, respectively obtaining the theoretical emission data of each production sub-process and determining whether it is in the first abnormal state; The difference between the current equipment operation data of the previous production sub-process and the parameters of the equipment operation data at the previous moment in any adjacent production sub-process is used as its real-time operation data difference set, and combined with the real-time operation data of the next production sub-process, they are input into the second evaluation model to obtain the theoretical emission data difference set of the next production sub-process; The parameters in the second abnormal state in the theoretical emission data difference set of the subsequent production sub-process are obtained, and a second abnormal signal is generated for feedback. The theoretical emission data difference set of the subsequent production sub-process in each group of adjacent production sub-processes is obtained respectively, and it is determined whether it is in the second abnormal state.

Images