Carbon emission prediction, regulation and control system and method based on multi-scale dynamic optimization

Abstract

The invention discloses a carbon emission prediction, regulation and control system and method based on multi-scale dynamic optimization. The system comprises a data acquisition layer, a five-dimensional coupling model layer and an optimization regulation and control layer. The data acquisition layer acquires multi-source heterogeneous data in real time by deploying an Internet of Things sensor network; a five-dimensional dynamic coupling model is arranged in the five-dimensional coupling model layer and comprises a multi-scale carbon emission kinetic equation, an industrial association network model, a space-time coupling prediction model and a dynamic optimization control model; an improved particle swarm optimization algorithm and a non-dominated sorting genetic algorithm are integrated in the optimization regulation and control layer, and parameter dynamic identification and multi-target optimization solution in the five-dimensional dynamic coupling model are realized based on multi-source heterogeneous data acquired by the data acquisition layer. The method can be applied to a regional carbon emission management platform, can be embedded into a high-energy-consumption enterprise production system, optimizes energy distribution and technical investment in real time, can also be used for carbon trading market auxiliary decision making, and provides data support for enterprise carbon quota distribution and emission reduction strategy making.

Description

Technical Field

[0001] The present invention belongs to the technical field of carbon emission prediction, and relates to a carbon emission prediction and control system and method based on multi-scale dynamic optimization. Background Art

[0002] Current carbon emission forecasting and control technologies primarily rely on single-scale static models and extensive control strategies. Traditional methods employ a single spatial and temporal scale analysis framework (e.g., annual or provincial), processing data at specific dimensions in isolation. This approach struggles to capture the dynamic, coupled evolutionary characteristics of multiple scales, from minute-to-year to interannual, and from industrial parks to national levels (e.g., the interaction between short-term climate fluctuations and long-term industrial upgrading). Consequently, forecasts fail to reflect the true dynamics of the system.

[0003] In terms of modeling nonlinear mechanisms, existing technologies (such as traditional regression models and simple time series analysis) can only handle linear or weakly nonlinear relationships. They inadequately capture complex nonlinear characteristics such as the diminishing marginal effect of technological progress and delayed transmission of industrial linkages. This results in weak model generalization and insufficient prediction accuracy in complex scenarios, making them inaccurate for regulatory control. Regarding spatiotemporal adaptability, existing strategies ignore regional differences in industrial structure, energy endowment, and climate conditions. They adopt a "one-size-fits-all" approach, such as fixed-ratio energy structure adjustments and uniform intensity emission reduction targets. These strategies are unable to dynamically adapt the characteristic parameters of different scenarios, such as high-energy-consuming industrial parks and clean energy bases, resulting in significant spatial variation in the effectiveness of regulatory solutions. In the field of multi-objective coordinated optimization, existing technologies lack a coupled framework for economic, energy, and environmental objectives. They focus solely on a single carbon emission indicator and fail to incorporate key factors such as GDP growth rate, energy transition costs, and the benefits of technological investment. As a result, emission reduction strategies often sacrifice economic growth, making it difficult to achieve a Pareto optimal balance between peak carbon emission control, peak timing optimization, and economic costs.

[0004] That is, the existing field of carbon emission prediction technology generally has the following defects and deficiencies:

[0005] 1. Limitations of single-scale static modeling: It only supports analysis at a single time / space scale and cannot reflect the dynamic changes of the system under multi-scale coupling. For example, it cannot capture the immediate impact of short-term climate fluctuations on energy consumption and the cumulative effect of long-term industrial upgrading.

[0006] 2. Insufficient modeling of nonlinear relationships: Traditional regression methods have difficulty describing complex mechanisms such as the marginal effects of technological progress and the delayed transmission of industrial linkages, resulting in weak generalization capabilities of the model in complex scenarios.

[0007] 3. Lack of coordinated optimization: A multi-objective optimization framework has not been established, and it is impossible to take into account the carbon emission peak, peak time and economic cost at the same time, which often leads to an imbalance between the economic and environmental benefits of emission reduction strategies.

[0008] 4. Poor temporal and spatial adaptability: Ignoring differences in regional industrial structure, energy endowment, and climatic conditions, the strategy design lacks specificity. For example, the regulatory strategies for high-energy-consuming industrial parks and clean energy bases do not reflect differentiation.

[0009] In summary, existing technologies are limited by single-scale isolated modeling, lack of nonlinear mechanism characterization, extensive spatiotemporal adaptation, and lack of multi-objective coordination. They have core bottlenecks of "insufficient prediction accuracy - low regulation efficiency - unbalanced development goals". There is an urgent need to build an innovative solution that integrates multi-dimensional dynamic coupling, spatiotemporal differentiated adaptation, and multi-objective intelligent optimization. Summary of the Invention

[0010] The purpose of the present invention is to address the deficiencies of the existing technology and provide a carbon emission prediction and control system and method based on multi-scale dynamic optimization.

[0011] The technical solution adopted in the present invention is:

[0012] A carbon emission prediction and control system based on multi-scale dynamic optimization, including

[0013] Data acquisition layer, five-dimensional coupling model layer and optimization control layer; among them:

[0014] The data collection layer collects multi-source heterogeneous data in real time by deploying an IoT sensor network;

[0015] The five-dimensional coupling model layer has a five-dimensional dynamic coupling model, including a multi-scale carbon emission dynamics equation, an industrial association network model, a spatiotemporal coupling prediction model, and a dynamic optimization control model, to achieve deep fusion modeling of multi-dimensional data;

[0016] The improved particle swarm optimization (IPSO) algorithm and the non-dominated sorting genetic algorithm (NSGA-Ⅲ) are integrated in the optimization and control layer. Based on the multi-source heterogeneous data collected by the data acquisition layer, dynamic identification of parameters and multi-objective optimization solutions in the five-dimensional dynamic coupling model are realized.

[0017] In the above technical solution, further, the multi-source heterogeneous data includes energy data by category, economic indicators including GDP growth rate and industrial output value, and environmental parameters including climate factors and carbon emission monitoring data.

[0018] Furthermore, the system also includes a decision support layer, which generates visual control strategies based on the solution results of the optimization control layer, provides an interactive decision interface, and supports the customized needs of users at different levels.

[0019] Furthermore, the five-dimensional dynamic coupling model first calculates the carbon emission intensity of each industry through an industrial linkage network model. This model uses the industrial linkage matrix to track the transfer of carbon emissions between industries and introduces Gaussian terms to capture the short-term effects of policy interventions. It uses industrial emission data as input to drive a multi-scale carbon emission dynamics equation, which integrates four factors: energy consumption, economic growth, technological emission reduction, and climate impact to generate a time series of cumulative carbon emissions.

[0020] The accumulated carbon emissions are then fed into a spatiotemporal coupled prediction model. This model combines the spatial diffusion equation with the source-sink dynamic function to quantify the impact of geographic factors on the propagation of carbon emissions through weighted diffusion coefficients. Ultimately, it outputs the distribution of carbon emission concentrations for different regional grids in the future. These predictions are captured by a dynamic optimization control model. This model solves for the optimal combination of energy structure adjustment rate, industrial upgrading rate, and technological input intensity, while satisfying the constraints of system dynamics equations, policy boundaries, and emission reduction targets, to minimize the sum of long-term emission penalties and regulation costs.

[0021] The optimized regulatory instructions are fed back to the industrial linkage network in real time, forming a closed loop by adjusting the carbon emission transfer coefficient and benchmark emission intensity among industries. The entire process is continuously iterated to ensure the economy and reliability of the emission reduction path under climate change and policy disturbances.

[0022] Furthermore, the industrial linkage network model is specifically as follows: a linkage matrix is ​​established based on the carbon emission transfer relationship between industries, the emission intensity between industries is quantified through the carbon emission conduction equation, and a Gaussian function is introduced to simulate the short-term effects of policy intervention, thereby generating dynamic industrial emission source intensity data.

[0023] Furthermore, the multi-scale carbon emission dynamics equation couples energy consumption, economic growth, technological emission reduction, and climate impact to obtain the cumulative carbon emissions in the time dimension, specifically:

[0024]

[0025] in: Indicates the cumulative carbon emissions at time t (10,000 tons), Indicates the energy consumption of category i (10,000 tons of standard coal), Indicates GDP growth rate (%), Indicates emission reduction technology index (0-1), Indicates climate impact factors, represents the dynamic coupling coefficient.

[0026] Furthermore, the spatiotemporal coupling prediction model discretizes geographical areas into grids, uses diffusion coefficients to quantify the spatial propagation characteristics of carbon emissions, and diffuses faster in high-weight areas. It combines the source-sink dynamic function to generate a three-dimensional distribution map of carbon emission concentrations in future time periods, and accurately locates regional emission hotspots. The source-sink dynamic function is composed of an emission source term for representing the carbon emission intensity in the region, a carbon sink absorption term for quantifying the absorption of carbon by natural or artificial systems, and an economic activity modulation term for reflecting the spatiotemporal modulation effect of economic activities on emissions.

[0027] Furthermore, the dynamic optimization control model aims to minimize emission penalty costs and regulatory economic costs, and solves the optimal regulation plan under the premise of satisfying system dynamics constraints, policy boundary conditions and initial emission status, including a real-time combination strategy of energy structure adjustment rate, industrial upgrading rate and technology investment intensity.

[0028] Furthermore, the improved particle swarm optimization (IPSO) algorithm adds a gradient term to the particle swarm optimization algorithm to guide particles to move in the direction where the fitness function decreases fastest, thereby performing a more refined search in a local area.

[0029] The present invention also provides a carbon emission prediction and control method based on multi-scale dynamic optimization, constructs a five-dimensional dynamic coupling model in the system as described in any of the above items, and optimizes and solves it.

[0030] The present invention further provides an electronic device, comprising:

[0031] one or more processors;

[0032] a memory for storing one or more programs;

[0033] When the one or more programs are executed by the one or more processors, the one or more processors implement any of the above methods.

[0034] A computer-readable storage medium stores computer-executable instructions, wherein the instructions are used to implement any of the above methods when executed.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] This invention integrates time, space, industry, energy, and technology dimensions for the first time, constructing multiscale dynamic equations and spatiotemporal diffusion models. Through five-dimensional dynamic coupling modeling, combined with the IPSO method with gradient-guided terms and parameter identification and multi-objective optimization using NSGA-III, it achieves multiscale carbon emission prediction. The method of this invention can be applied to regional carbon emission management platforms and can be expanded to city, provincial, and municipal scales. It can be integrated with the Industrial Internet energy-saving module and embedded in the production systems of high-energy-consuming enterprises to optimize energy allocation and technology investment in real time. It can also be used to assist in decision-making in the carbon trading market, providing data support for corporate carbon quota allocation and emission reduction strategy formulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a structural block diagram of the system of the present invention. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more clearly understood, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. The technical features in the various embodiments of the present invention can be combined accordingly without conflicting with each other.

[0039] The present invention provides a carbon emission prediction and control system based on multi-scale dynamic optimization. According to a specific embodiment of the present invention, the system adopts a four-layer five-dimensional modeling framework:

[0040] It adopts a four-layer technical architecture design, including data collection layer, five-dimensional dynamic model layer, optimization and control layer, and decision support layer; specifically:

[0041] Data collection layer: Deploy an IoT sensor network to collect multi-source heterogeneous data in real time, including energy consumption (energy data by category), economic indicators (GDP growth rate, industrial output value), and environmental parameters (climate factors, carbon emission monitoring data), supporting minute-level data updates.

[0042] Five-dimensional dynamic model layer: A five-dimensional dynamic coupling model is constructed, including multi-scale carbon emission kinetic equations, industrial linkage network models, spatiotemporal coupling prediction models and dynamic optimization control models, to achieve deep fusion modeling of multi-dimensional data.

[0043] Optimization and control layer: Integrates the improved particle swarm optimization (IPSO) algorithm and the non-dominated sorting genetic algorithm (NSGA-III) to achieve dynamic identification of model parameters and multi-objective optimization solutions, and supports real-time iterative optimization of strategies.

[0044] Decision support layer: Based on the results obtained, it generates visual control strategies and provides an interactive decision-making interface to support the customized needs of users at different levels (government, enterprises, and parks).

[0045] The five-dimensional coupled modeling system (time, space, industry, energy, and technology) in this invention innovatively integrates the five dimensions of time (minutes to years), space (park to country), industry (industry classification), energy (energy category), and technology (emission reduction technology index) to construct a multi-scale coupled model, enabling dynamic simulation of carbon emissions from micro-enterprises to macro-regions. Details are as follows:

[0046] 1. Industry linkage network model

[0047] Based on the carbon emission transfer relationship between industries, a correlation matrix is ​​established, and the industry correlation matrix is ​​defined as follows:

[0048]

[0049] The elements is the ratio of carbon emissions transferred from industry i to industry j to the total carbon emissions output of industry i, indicating the proportion of carbon emissions transferred from industry i to industry j, that is:

[0050]

[0051] The carbon emission conduction equation is used to quantify the emission intensity among industries, and a Gaussian function is introduced to simulate the short-term effects of policy interventions, thereby generating dynamic industrial emission source intensity data. The carbon emission conduction equation is:

[0052]

[0053] Where, is all input carbon flows to industry i, is the output carbon flow of industry j, is the baseline emission intensity of industry j, i.e., the carbon emissions per unit output without intervention, is a Gaussian term, which represents the short-term impact of the policy (such as production restriction order) introduced at time t0. Controls how long the effect lasts.

[0054] After deducting the carbon emissions transferred to downstream, the net emissions generated by the industry’s own production activities are the emission source intensity.

[0055] 2. Multiscale carbon emission dynamics equation

[0056] By coupling energy consumption, economic growth, technological emission reduction and climate impact, a dynamic equation is constructed to describe the evolution mechanism of cumulative carbon emissions:

[0057]

[0058] in: Indicates the cumulative carbon emissions at time t (10,000 tons), Indicates the energy consumption of category i (10,000 tons of standard coal), Indicates GDP growth rate (%), Indicates emission reduction technology index (0-1), Indicates climate impact factors, represents the dynamic coupling coefficient.

[0059] 3. Spatiotemporal coupling prediction model

[0060] The geographical area is discretized into grids, and the diffusion coefficient is used to quantify the spatial propagation characteristics of carbon emissions. High-weight areas diffuse faster. Combined with the source-sink dynamic function, a three-dimensional distribution map of carbon emission concentrations in the future period is generated to accurately locate regional emission hotspots.

[0061] Spatial discretization:

[0062]

[0063] The diffusion coefficient is:

[0064]

[0065] Where, is the geographic grid weight (e.g. industrial area weight > farmland), 、 is the grid space step size, is the carbon emission concentration field in the spatiotemporal dimension, where the source-sink dynamic function It consists of the emission source term C, which is used to represent the carbon emission intensity in the region, the carbon sink absorption term E, which is used to quantify the absorption of carbon by natural or artificial systems, and the economic activity modulation term G, which is used to reflect the spatiotemporal modulation effect of economic activities on emissions.

[0066] 4. Dynamic optimization control model

[0067] With the goal of minimizing emission penalty costs and regulatory economic costs, the objective function is:

[0068]

[0069] The constraints are system dynamics, policy boundaries, emission reduction targets, and initial states:

[0070]

[0071] The control variables are: They represent the energy structure adjustment rate, industrial upgrading rate and technological investment intensity respectively.

[0072] In this paper, gradient-guided particle swarm optimization and NSGA-III are used to solve multi-objective optimization problems. The gradient-guided term is introduced into IPSO to improve the parameter identification accuracy, and NSGA-III handles high-dimensional multi-objective optimization problems.

[0073] The workflow of the carbon emission forecasting and dynamic control system in this invention begins with the construction of an industry-linked network model. First, a correlation matrix is ​​established based on the carbon emission transfer relationships between industries. The carbon emission conduction equation quantifies the emission intensity between industries, and a Gaussian function is introduced to simulate the short-term effects of policy interventions, thereby generating dynamic industry emission source intensity data. This emission data is then input into a multi-scale carbon emission dynamics equation, which simultaneously integrates economic growth rate, emission reduction technology index, and climate impact factors. The dynamic coupling coefficient is used to calculate the evolution trend of cumulative carbon emissions.

[0074] The resulting cumulative carbon emissions then drive a spatiotemporal coupled prediction model: the system discretizes the geographic region into a grid and uses diffusion coefficients to quantify the spatial propagation characteristics of carbon emissions (high-weight areas such as industrial zones and transportation hubs experience faster diffusion). Combined with source-sink dynamic functions, it generates a three-dimensional distribution map of carbon emission concentrations over future time periods, pinpointing regional emission hotspots. The prediction results immediately trigger a dynamic optimization control model, which aims to minimize "emission penalty costs" and "control economic costs." While satisfying system dynamics constraints, policy boundary conditions, and initial emission states, it determines the optimal control instructions—a real-time combination strategy that considers the energy structure adjustment rate, industrial upgrade rate, and technological investment intensity.

[0075] Ultimately, these regulatory instructions are fed back to the industrial linkage network in a closed loop: by adjusting the carbon emission transfer coefficient between industries, reducing the weight of high-energy-consuming industries or increasing the linkage strength of clean technology industries, the emission transmission path is reconstructed, thereby updating the industrial emission source intensity data.

[0076] The entire system forms a continuously iterative closed loop of "industrial network → dynamic equation → spatiotemporal prediction → optimization and regulation → industrial network", dynamically maintaining the economy and sustainability of the emission reduction path under climate change and policy disturbances.

[0077] Taking an industrial park in a certain province in 2025 as an example, we conduct prediction and carbon regulation. The system deployment is as follows:

[0078] IoT monitoring nodes: 50, covering the park's main energy-consuming equipment; data collection frequency: 1Hz (real-time monitoring of energy consumption and emissions data); model update cycle: 1 hour (edge ​​nodes process data in real time, and the cloud regularly optimizes model parameters); strategy adjustment cycle: 24 hours (generating the optimal control plan daily).

[0079] After optimizing and regulating using the solution of the present invention, the proportion of renewable energy has been increased through energy structure adjustment. Combined with technological investment and industrial upgrading, the emission intensity per unit of output value has been reduced. Intelligent regulation and optimization of energy distribution have also been achieved, reducing equipment downtime, thereby achieving a reduction in carbon emissions while improving production efficiency and reducing energy consumption.

[0080] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0081] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0082] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0083] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0084] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A carbon emission prediction and control system based on multi-scale dynamic optimization, characterized in that: include Data acquisition layer, five-dimensional coupling model layer and optimization control layer; among them: The data collection layer collects multi-source heterogeneous data in real time by deploying an IoT sensor network; The five-dimensional coupling model layer has a five-dimensional dynamic coupling model, including a multi-scale carbon emission dynamics equation, an industrial association network model, a spatiotemporal coupling prediction model, and a dynamic optimization control model, to achieve deep fusion modeling of multi-dimensional data; The improved particle swarm optimization (IPSO) algorithm and the non-dominated sorting genetic algorithm (NSGA-Ⅲ) are integrated in the optimization and control layer. Based on the multi-source heterogeneous data collected by the data acquisition layer, dynamic identification of parameters and multi-objective optimization solutions in the five-dimensional dynamic coupling model are realized.

2. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 1 is characterized in that: The multi-source heterogeneous data include energy data by category, economic indicators including GDP growth rate and industrial output value, and environmental parameters including climate factors and carbon emission monitoring data.

3. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 1 is characterized in that: The system also includes a decision support layer, which generates visual control strategies based on the solution results of the optimization control layer, provides an interactive decision interface, and supports the customized needs of users at different levels.

4. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 1 is characterized in that: The five-dimensional dynamic coupling model first calculates the carbon emission intensity of each industry through an industrial linkage network model. The model uses the industrial linkage matrix to track the transfer of carbon emissions between industries and introduces Gaussian terms to capture the short-term effects of policy interventions. Using industrial emission data as input, a multi-scale carbon emission dynamics equation is driven. This equation integrates four factors: energy consumption, economic growth, technological emission reduction, and climate impact to generate a time series of cumulative carbon emissions. The accumulated carbon emissions are then fed into a spatiotemporal coupled prediction model. This model combines the spatial diffusion equation with the source-sink dynamic function to quantify the impact of geographic factors on the propagation of carbon emissions through weighted diffusion coefficients. Ultimately, it outputs the distribution of carbon emission concentrations for different regional grids in the future. These predictions are captured by a dynamic optimization control model. This model solves for the optimal combination of energy structure adjustment rate, industrial upgrading rate, and technological input intensity, while satisfying the constraints of system dynamics equations, policy boundaries, and emission reduction targets, to minimize the sum of long-term emission penalties and regulation costs. The optimized regulatory instructions are fed back to the industrial linkage network in real time, forming a closed loop by adjusting the carbon emission transfer coefficient and benchmark emission intensity among industries. The entire process is continuously iterated to ensure the economy and reliability of the emission reduction path under climate change and policy disturbances.

5. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 4 is characterized in that: The industrial linkage network model is specifically as follows: a linkage matrix is ​​established based on the carbon emission transfer relationship between industries, the emission intensity between industries is quantified through the carbon emission conduction equation, and a Gaussian function is introduced to simulate the short-term effects of policy intervention, thereby generating dynamic industrial emission source intensity data.

6. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 4 is characterized in that: The multi-scale carbon emission dynamics equation couples energy consumption, economic growth, technological emission reduction, and climate impact to obtain the cumulative carbon emissions in the time dimension, specifically: in: represents the cumulative carbon emissions at time t, represents the energy consumption of category i, represents the GDP growth rate, represents the emission reduction technology index, Indicates climate impact factors, represents the dynamic coupling coefficient.

7. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 4 is characterized in that: The spatiotemporal coupling prediction model discretizes geographic areas into grids and uses diffusion coefficients to quantify the spatial propagation characteristics of carbon emissions. High-weight areas diffuse faster, and combines source-sink dynamic functions to generate a three-dimensional distribution map of carbon emission concentrations in future time periods, accurately locating regional emission hotspots. The source-sink dynamic function is composed of an emission source term used to represent the carbon emission intensity within the region, a carbon sink absorption term used to quantify the absorption of carbon by natural or artificial systems, and an economic activity modulation term used to reflect the spatiotemporal modulation effect of economic activities on emissions.

8. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 4 is characterized in that: The dynamic optimization control model aims to minimize emission penalty costs and regulatory economic costs. Under the premise of satisfying system dynamics constraints, policy boundary conditions and initial emission status, it solves the optimal regulation plan, including a real-time combination strategy of energy structure adjustment rate, industrial upgrading rate and technology investment intensity.

9. The carbon emission prediction and control system based on multi-scale dynamic optimization according to claim 1 is characterized in that: The improved particle swarm optimization (IPSO) algorithm adds a gradient term to the particle swarm optimization algorithm to guide particles to move in the direction where the fitness function decreases fastest, thereby performing a more refined search in a local area.

10. A carbon emission prediction and control method based on multi-scale dynamic optimization, characterized in that: Construct a five-dimensional dynamic coupling model in the system as described in any one of claims 1 to 9, and optimize and solve it.

Images