Ultra-large city carbon emission reduction multi-stage comprehensive optimization method considering economic development

Through system dynamics and ridge regression algorithm combined with logarithmic mean Diesh index method, an economic-energy-power-environmental balance system was constructed, which solved the problem of unclear system impact relationship of carbon emission prediction model in the existing technology, and achieved more Accurate carbon emission forecast and emission reduction path exploration.

CN118966527APending Publication Date: 2024-11-15ELECTRIC POWER SCI & RES INST OF STATE GRID TIANJIN ELECTRIC POWER CO +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410965564.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The impact relationship between existing carbon emission forecasting models in urban internal systems is unclear, making it difficult to achieve a wider range and more accurate carbon emission forecast.

Method used

Using the system dynamics-based ridge regression algorithm, combined with the logarithmic average Diesh index method and data preprocessing technology, an economic-energy-electric power-environmental balance system is constructed, and a multi-stage comprehensive optimization method for carbon emission reduction is determined, taking into account economic development and energy Dynamic balance of structure and environmental systems.

Benefits of technology

A more reasonable multi-scenario forecast of carbon emission reduction potential in megacities has been achieved. Through multi-objective planning and scenario analysis methods, accurate carbon emission reduction paths are explored and effective carbon emission reduction strategies are formulated.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118966527A_ABST
    Figure CN118966527A_ABST
Patent Text Reader

Abstract

The invention discloses an ultra-large city carbon emission reduction multi-stage comprehensive optimization method considering economic development, and the method comprises the steps: selecting an ultra-large city carbon emission influence factor based on an economic-energy-electric power-environment balance system, and obtaining the historical data of the influence factor; performing data preprocessing on the obtained data set to obtain a high-quality data set; according to the method, the influence factor action mechanism and contribution degree of urban carbon emission are effectively analyzed by adopting a logarithmic average Dirse index method, a carbon emission reduction potential multi-scenario prediction model of an ultra-large city is constructed in combination with system dynamics, correlation coefficients are determined by adopting a ridge regression analysis method, and finally a carbon emission reduction time table and a route map which can be used for reference are generated. According to the method, important carbon emission driving factors are analyzed by providing a systematic method, carbon emission trend prediction is realized, and an effective strategy is formulated to reduce the emission.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of carbon emission reduction optimization and analysis, and in particular to a multi-stage comprehensive optimization method for carbon emission reduction in a super-large city taking into account economic development. Background Art

[0002] With global warming and the development of urbanization and industrialization, controlling carbon emissions has long become an important issue faced by all mankind. At present, most carbon emission forecasts are conducted for a certain industry or region. The existing carbon emission forecasting models still have defects, such as unclear influence relationships between various systems within the city. It is more valuable to achieve a wider range and more accurate carbon emission forecast on this basis.

[0003] Therefore, the present invention proposes a multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development, which comprehensively considers that the economy-energy-power-environment system is in dynamic balance. On this basis, it is more valuable to achieve a wider range and more accurate carbon emission forecast. Summary of the invention

[0004] The present invention provides a multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development, which makes full use of various driving factors, fully reflects the relationship between carbon emissions according to system dynamics, and uses ridge regression algorithm to realize model construction. See the following description for details:

[0005] A multi-stage comprehensive optimization method for carbon emission reduction in a megacity taking into account economic development, the method comprising the following steps:

[0006] Step 1: Based on the “economy-energy-electricity-environment” balance system, select the factors affecting urban carbon emissions and obtain historical data on the factors;

[0007] Step 2: Perform data preprocessing on the data set obtained in step 1, identify bad data, clean data, and complete data to obtain a high-quality data set;

[0008] Step 3: Based on the data provided in step 2, the Logarithmic Mean Dirichlet Index (LMDI) and system dynamics methods are used to effectively analyze the mechanism and contribution of factors affecting urban carbon emissions;

[0009] Step 4: Based on the key influencing factors obtained in step 3, a multi-scenario prediction model for carbon emission reduction potential of megacities is constructed;

[0010] Step 5: Based on the prediction model and historical data in step 4, the ridge regression method is used to determine the model parameters and obtain a multi-stage comprehensive optimization method for carbon emission reduction in megacities while satisfying the constraints.

[0011] Furthermore, in step 1, based on the “economy-energy-electricity-environment” balance system, the influencing factors of carbon emissions in megacities are selected, and the historical data of the influencing factors are obtained. The influencing factors are selected as follows:

[0012] The Kaya identity is established. The Kaya identity is mainly used to decompose the factors affecting carbon emissions. Based on carbon emissions (C), gross domestic product (GDP), total population (P), and the proportion of non-fossil energy (E N ) and electrification rate (E R ) decomposes the factors affecting carbon emissions and expresses it as follows:

[0013]

[0014] Where i (i=1,2,L,n) represents each area in a super-large city.

[0015] The influencing factors are defined as:

[0016] Influencing factors meaning unit C Carbon emissions from energy consumption 10,000 tons GDP Regional GDP 100 million yuan C / GDP Carbon emissions per unit of GDP produced Tons / 10,000 yuan <![CDATA[E N ]]> Non-fossil energy share % C / E Carbon emissions per unit of energy consumption Ton / ton standard coal P Permanent population 10,000 people C / P Carbon emissions per capita Tons / person <![CDATA[E R ]]> Electrification rate %

[0017] Moreover, in step 2, the data set obtained in step 1 is preprocessed to identify bad data, clean data, and complete data to obtain a high-quality data set;

[0018] The LOF and GAKNNI methods were used to clean the data, mainly to remove the economic data with obvious data anomalies;

[0019] Moreover, in step 3, the logarithmic mean Dirichlet index method (LMDI) and system dynamics method are used to effectively analyze the mechanism and contribution of the factors affecting urban carbon emissions, specifically:

[0020] The change in carbon emissions is ΔC, which can be decomposed into l according to the carbon emissions identity. i The specific carbon emission calculation is as follows:

[0021]

[0022] The additive form of the LMDI decomposition model is as follows:

[0023]

[0024] Where l1, l2, l3, and l4 are the number of factors decomposed from each type of driving factor using the LMDI decomposition method.

[0025] The various driving factors are decomposed by the additive LMDI decomposition method, and the logarithm is taken at the same time. Then, the time derivative is taken from both sides at the same time. The transformation is shown in the following formula:

[0026]

[0027] The weight function can be rewritten as:

[0028]

[0029] Then q i (t * ) into the above, we can get:

[0030]

[0031] in,

[0032]

[0033] Moreover, in step 4, a multi-scenario prediction model of carbon emission reduction potential of megacities is constructed based on key influencing factors, specifically:

[0034] Based on the principles of system dynamics and the results of LMDI decomposition, a multi-scenario prediction model for carbon emission reduction potential in megacities is constructed.

[0035] It consists of four sub-models: socio-economic model, energy model, electricity model and environmental model.

[0036]

[0037] Taking the carbon emissions of megacities as the explained variable, factors are selected from four aspects: economy, population size, energy structure, and electrification rate. The variables involved are logarithmized to obtain the specific formula:

[0038] LqCy T =a+blnGDP+clnP+dlnE N +elnE R

[0039] Where a, b, c, d, and e are the elastic coefficients corresponding to each variable.

[0040] The specific process is shown in the attached figure.

[0041] Moreover, in step 5, based on the prediction model and historical data, the ridge regression method is used to determine the model parameters, and the multi-stage comprehensive optimization method for carbon emission reduction in super-large cities is obtained under the constraints. The specific algorithm is:

[0042] Assume the above model formula as shown below:

[0043] y=θx

[0044] Using the least squares method, the loss function is defined as the square of the residual and the loss function is minimized:

[0045] ||Xθ-y 2 ||

[0046] θ=(X T X) -1 X T y

[0047] When calculating (X T X) -1 It will produce large errors and have poor stability and reliability.

[0048] In order to solve the above problem, we need to transform the ill-posed problem into a well-posed problem: we add a regularization term to the above loss function, which becomes:

[0049] ||Xθ-y 2 ||+||Γθ 2 ||

[0050] Make a ridge trace graph and observe at what value the graph tends to be stable, so as to determine the value of α. That is, the coefficient values ​​of each parameter in the multi-scenario prediction model of carbon emission reduction potential of super-large cities can be obtained.

[0051] According to the different policy requirements of megacities, the constraints include: energy supply and demand balance constraints and non-fossil energy power generation ratio constraints.

[0052] T ES (t)≥T EC (t)

[0053] Where, T ES (t) is the total energy supply of the megacity in year t, T EC (t) is the total energy consumption of the megacity in year t, in units of 10,000 tons of standard coal.

[0054]

[0055] In the formula, E N (t) is the proportion of non-fossil energy in the total energy consumption of the mega-city in year t, E C (t) is the final consumption of fossil energy in year t, τ t It is the lower limit of the proportion of non-fossil energy power generation in year t.

[0056] In summary, the optimal solution with the lowest total carbon emissions based on economic development is obtained, and the final output is the optimized combination of influencing factors such as carbon emission trends, peak conditions, and economic development that meet various constraints, and a carbon emission reduction timetable and roadmap for reference is generated.

[0057] The beneficial effects of the technical solution provided by the present invention are:

[0058] (1) Compared with the prior art, the present invention comprehensively considers factors such as economic growth and the proportion of non-fossil energy, and considers the interaction and influence of multiple factors, thereby constructing a more reasonable multi-scenario prediction model for the carbon emission reduction potential of super-large cities.

[0059] (2) Compared with the commonly used prediction methods in the prior art, the present invention obtains the carbon emission trend through multi-objective planning and scenario analysis, thereby exploring the accurate carbon emission reduction path. The feasibility and accuracy of the proposed method are verified through actual research data.

[0060] Therefore, the present invention proposes a multi-stage comprehensive optimization method for carbon emission reduction in super-large cities taking into account economic development, which provides a systematic method to analyze important carbon emission drivers, realize carbon emission trend prediction, and formulate effective strategies to reduce these emissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is a flow chart of the present invention;

[0062] Figure 2 It is a causal loop diagram of the carbon emission system;

[0063] Figure 3 This is a coupling mechanism diagram of the carbon emission reduction model for mega cities;

[0064] Figure 4 An example diagram of calculation using the method of the present invention with actual data from Tianjin as an example. DETAILED DESCRIPTION

[0065] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention are described in further detail below.

[0066] In order to solve the problems existing in the background technology, the present invention proposes a multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development. Relevant data, including energy consumption, economic growth rate, electricity production and carbon emission indicators, are collected, and the impact contribution is quantified by the LMDI method, so as to select the influencing factors with strong correlation; the system dynamics method is used to more clearly show the influence relationship between the influencing factors; based on the system dynamics principle, multi-objective planning method and scenario analysis method, a multi-scenario prediction model for carbon emission reduction potential in megacities is constructed, and a multi-stage comprehensive optimization method for carbon emission reduction in megacities is obtained according to the prediction process and result analysis.

[0067] Example

[0068] A multi-stage comprehensive optimization method for carbon emission reduction in a megacity taking into account economic development, the method comprising the following steps:

[0069] 1: Analysis of factors affecting carbon emissions based on LMDI;

[0070] In order to effectively analyze the mechanism and contribution of each driving factor affecting carbon emissions, the index decomposition method is used to quantitatively measure the key factors affecting carbon emissions and analyze their impact. The specific approach is as follows:

[0071] 101: Establishing the Kaya Identity

[0072] The Kaya identity is established. The Kaya identity is mainly used to decompose the factors affecting carbon emissions. Based on carbon emissions (C), gross domestic product (GDP), total population (P), and the proportion of non-fossil energy (E N ) and electrification rate (E R ) decomposes the factors affecting carbon emissions and expresses it as follows:

[0073]

[0074] Where i (i=1,2,L,n) represents each area in a super-large city.

[0075] 102: The influencing factors are defined as:

[0076]

[0077]

[0078] 2: Data cleaning;

[0079] 201: Outlier Detection

[0080] The LOF algorithm is used to perform anomaly detection on the feature data to be detected. The outlier factor value of each data point in the data set is calculated to determine the set of outliers and determine whether all the data to be tested have completed the detection.

[0081] Based on the historical feature data set, the GAKNNI algorithm is used to complete the abnormal and missing data, and it is determined whether all the pending data have been completed. Among them, data completion uses GAKNNI to calculate the K data points closest to the input sample in space and the training data set, and then calculates the similarity or difference between the development trends of the data;

[0082] The LOF algorithm is used to detect anomalies in the feature data to be detected. The method of determining the abnormal point set by calculating the outlier factor value of each data point in the data set is as follows:

[0083] First, the average density of data points in the data set is analyzed, and the number of outliers M1 and the outlier set D1 in the data set are determined according to the distribution of density. Then, M2 (M2=M1) outliers and the outlier set D2 are determined by calculating the outlier factor, and the intersection of D1 and D2 is taken as the final outlier cluster.

[0084] The method for determining the outliers of a data set according to the distribution of density is as follows:

[0085] The k-distance of data object q is defined as the distance from the kth point in the data set that is closest to data object q to q, denoted as k-distance(q). The distance here refers to the Euclidean distance, that is, the straight-line distance.

[0086] The distance between the calculated data set and the data object q is no greater than k;

[0087] Calculate p, q as any two points in the data set;

[0088] Calculate the local reachability density of q, which refers to the average local reachability density from q to all points in its neighborhood;

[0089] Calculate the degree of outliers in the data;

[0090] Identify and remove outliers in your data based on how outlier they are.

[0091] 201: Data completion

[0092] The method of completing abnormal and missing data based on historical feature data sets and using the GAKNNI algorithm is as follows:

[0093] KNN is used to find the K data points that are closest to the missing point in space in the training data set. Then, based on the similarity weights of the K data points based on the grey relational analysis, the completion value of the missing point is the sum of the products of the K data points and their similarity weights.

[0094] The method to obtain the missing point completion value is:

[0095] Establish the time series characteristic data of carbon emissions and their influencing factors, and obtain the sample matrix T without missing items train , missing item sample time series t lack ;

[0096] Calculate the sample matrix T without missing items train The sample time series t with missing items lack The Euclidean distance parameter d;

[0097] From T train Screening and t lack The nearest K similar time series,

[0098] Calculate t lack and Grey correlation coefficient of K similar time series in ;

[0099] calculate The similarity weight ω of the K similarity times k ;

[0100] Compute the missing points' complement values ​​y.

[0101] 3: Selection of key influencing factors

[0102] 301: Establish a carbon emission factor decomposition model

[0103] The weights are calculated based on the logarithmic mean Dimitrov index decomposition method of addition and the logarithmic expected value calculation formula.

[0104] Assume that the change in carbon emissions is ΔC. According to the carbon emissions identity, it can be decomposed into l i The specific carbon emission calculation is as follows:

[0105]

[0106] The additive form of the LMDI decomposition model is as follows:

[0107]

[0108] Where l i The number of factors decomposed using the LMDI decomposition method for each type of driving factor.

[0109] 302: Additive form of LMDI decomposition

[0110] The various driving factors are decomposed by the additive LMDI decomposition method. The specific process is as follows:

[0111] Take the logarithm at the same time, and then take the derivative of both sides with respect to time, and the transformation is as follows:

[0112]

[0113] For the integration of both sides, 0 represents the base period, T represents the end of the current period, and the transformation is as follows:

[0114]

[0115] and,

[0116]

[0117] Perform power operations with base e on both sides at the same time, so that the identity is transformed into the following formula:

[0118]

[0119] The transformation from the mean value theorem of definite integral is as follows:

[0120]

[0121] In the above formula, q i (t * ) is defined in a t * The value of the weight function at time t * ∈[0,T]. i (t * ) is calculated by taking the logarithmic mean function.

[0122] Among them, the inference method of the logarithmic mean function is as follows:

[0123]

[0124] According to the definition of logarithmic mean function:

[0125]

[0126] In the above formula, L(C i,o ,C i,T ) represents the factorial of the logarithmic mean of carbon emissions generated in the i-th region, L(C0,C T ) is regarded as the sum of logarithmic means, so the weight function can be rewritten as:

[0127]

[0128] Then q i (t * ) into the above, we can get:

[0129]

[0130] in,

[0131]

[0132] 4: Construct a multi-scenario prediction model for carbon emission reduction potential in megacities

[0133] 401: Model Building

[0134] Based on the system dynamics principle and LMDI decomposition results, the interaction relationship and dynamic changes between various factors are displayed, which more intuitively reflects the logical relationship between urban carbon emission elements. The specific carbon emission system causal loop diagram in the model construction can be obtained, such as Figure 2 shown.

[0135] Construct a multi-scenario prediction model for the carbon emission reduction potential of megacities, such as Figure 3As shown. It consists of four sub-models: socio-economic model, energy model, electricity model and environmental model. The socio-economic model focuses on the analysis from the perspective of the production behavior of the industrial sector. Economic development promotes the continuous flow of intermediate inputs and final demands of various sectors, following the input-output balance. The energy system meets the balance between energy supply and demand. The demand side mainly includes industrial sectors and residents' lives. Both production activities and residents' lives need to consume energy to form an energy flow. The present invention controls the total energy consumption from the demand side; the supply side focuses on replacing traditional fossil energy with renewable energy, increasing the overall urban electrification rate, thereby reducing environmental pollution caused by energy production and consumption, and reducing negative impacts on the environment.

[0136] The combustion and use of energy produces carbon dioxide emissions, forming the material flow of the environmental system. The environmental system aims to minimize carbon dioxide emissions. Through the three major measures of industrial structure adjustment, energy structure adjustment and energy efficiency improvement, it studies the carbon peak time and carbon emission peak of megacities, and then explores the multi-stage comprehensive optimization method of carbon emission reduction in megacities.

[0137]

[0138] Taking the carbon emissions of megacities as the explained variable, factors are selected from four aspects: economy, population size, energy structure, and electrification rate. The variables involved are logarithmized to obtain the specific formula:

[0139] LqCy T =a+blnGDP+clnP+dlnE N +elnE R

[0140] Where a, b, c, d, and e are the elastic coefficients corresponding to each variable.

[0141] 5: Multi-stage optimization path for carbon emission reduction

[0142] 501: Generate typical scenarios

[0143] According to the historical statistical data of megacities, combined with the size of the carbon emission driving effect and the policy requirements of megacities, based on multi-objective planning methods and scenario analysis methods, different growth rates are set for the influencing factors to produce three typical scenarios.

[0144] 502: Ridge regression analysis to determine the correlation coefficient

[0145] Assume the above model formula as shown below:

[0146] y=θx

[0147] Using the least squares method, the loss function is defined as the square of the residual and the loss function is minimized:

[0148] ||Xθ-y 2 ||

[0149] θ=(X T X) -1 X T y

[0150] If X is not full rank, or if there is a large linear correlation between some columns, X T The determinant of X is close to 0, that is, X T X is close to being a singular matrix, and the above problem becomes an ill-posed problem. At this time, when calculating (X T X) -1 It will produce large errors and have poor stability and reliability.

[0151] In order to solve the above problem, we need to transform the ill-posed problem into a well-posed problem: we add a regularization term to the above loss function, which becomes:

[0152] ||Xθ-y 2 ||+||Γθ 2 ||

[0153] Here, we define Γ=αI, where I is the identity matrix.

[0154] so:

[0155] θ(α)=(X T X+αI) -1 X T y

[0156] As α gradually increases, the absolute values ​​of each element in θ(α) continue to decrease, and the deviation between them and the correct value gradually increases. When α tends to infinity, θ(α) almost tends to 0. Among them, a ridge trace graph is made to observe at which value the graph tends to be stable, so as to determine the value of α. That is, the coefficient values ​​of each parameter in the multi-scenario prediction model of carbon emission reduction potential of super-large cities can be obtained.

[0157] 503: Constraints

[0158] According to the different policy requirements of megacities, the constraints include: energy supply and demand balance constraints and non-fossil energy power generation ratio constraints.

[0159] T ES (t)≥T EC (t)

[0160] Where, T ES (t) is the total energy supply of the megacity in year t, T EC(t) is the total energy consumption of the megacity in year t, in units of 10,000 tons of standard coal.

[0161]

[0162] In the formula, E N (t) is the proportion of non-fossil energy in the total energy consumption of the mega-city in year t, E C (t) is the final consumption of fossil energy in year t, τ t It is the lower limit of the proportion of non-fossil energy power generation in year t.

[0163] In summary, the optimal solution with the lowest total carbon emissions based on economic development is obtained, and the final output is the optimized combination of influencing factors such as carbon emission trends, peak conditions, and economic development that meet various constraints, and a carbon emission reduction timetable and roadmap for reference is generated.

[0164] The overall process is as follows Figure 1 shown.

[0165] The above embodiments are simulated and analyzed by using the actual data of Tianjin. The beneficial effects obtained by the present invention are as follows: Figure 4 As shown. The model proposed in the present invention predicts the carbon emission reduction path of Tianjin. The prediction results show that the carbon peak can be achieved within the constraints of Tianjin's planning time and the proportion of non-fossil energy, so the carbon emission reduction path under this scenario meets the requirements. It can be considered that the prediction model obtained using the method described in this article is reasonable enough and practical.

[0166] In summary, the embodiments of the present invention provide a multi-stage comprehensive optimization method for carbon emission reduction in a super-large city taking into account economic development. The present invention makes full use of historical information to accurately and reliably explore the multi-stage optimization method for carbon emission reduction.

[0167] Unless otherwise specified, the models of the components in the embodiments of the present invention are not limited, and any device that can perform the above functions may be used.

[0168] Those skilled in the art will appreciate that the accompanying drawing is only a schematic diagram of a preferred embodiment, and the serial numbers of the embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development, characterized in that: The method comprises the following steps: Step 1: Based on the "economy-energy-electricity-environment" balance system, select the factors affecting urban carbon emissions and obtain historical data on the factors; Step 2: Perform data preprocessing on the data set obtained in step 1, identify bad data, clean data, and complete data to obtain a high-quality data set; Step 3: Based on the data provided in step 2, the Logarithmic Mean Dirichlet Index (LMDI) and system dynamics methods are used to effectively analyze the mechanism and contribution of factors affecting urban carbon emissions; Step 4: Based on the key influencing factors obtained in step 3, a multi-scenario prediction model for carbon emission reduction potential of megacities is constructed; Step 5: Based on the prediction model and historical data in step 4, the ridge regression method is used to determine the model parameters and obtain a multi-stage comprehensive optimization method for carbon emission reduction in megacities while satisfying the constraints.

2. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: The influencing factors in step 1 include: carbon emissions from energy consumption C, regional real gross domestic product GDP, carbon emissions per unit of GDP C / GDP, and the proportion of non-fossil energy E N , carbon emissions per unit of energy consumption C / E, permanent population P, per capita carbon emissions C / P, electrification rate E R , using the index decomposition method to quantitatively measure the key factors affecting carbon emissions and analyze their impact, The Kaya identity is established to decompose the factors affecting carbon emissions, which can be expressed as: Where i (i=1,2,L,n) represents each area in a super-large city.

3. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: Step 2 uses the LOF algorithm to perform anomaly detection on the feature data to be detected, determines the set of abnormal points by calculating the outlier factor value of each data point in the data set, and determines whether all the data to be tested have completed the detection.

4. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: Step 2 uses the GAKNNI algorithm based on the historical feature data set to complete the abnormal and missing data, and determines whether all the data to be completed has been completed.

5. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: Step 3 uses the logarithmic mean Dirichlet index method (LMDI) and system dynamics method to effectively analyze the mechanism and contribution of factors affecting urban carbon emissions. The analysis method is as follows: Assume that the change in carbon emissions is ΔC. According to the carbon emissions identity, it can be decomposed into l i The specific carbon emission calculation is as follows: The additive form of the LMDI decomposition model is as follows: Where l i is the number of factors decomposed by LMDI decomposition method for each driving factor, C t is the carbon emission in year t, C t-1 is the carbon emissions in year t-1. The various driving factors are decomposed by the additive LMDI decomposition method, and the logarithm is taken at the same time. Then, the time derivative is taken from both sides at the same time. The transformation is shown in the following formula: Rewrite the weight function as: Then q i (t * ) into the above, we can get: in, 6. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: In step 4, a multi-scenario prediction model of carbon emission reduction potential of megacities is constructed based on key influencing factors, specifically: Based on the principles of system dynamics and the results of LMDI decomposition, a multi-scenario prediction model of carbon emission reduction potential in megacities is constructed, which consists of four sub-models: socio-economic model, energy model, power model and environmental model. The objective function considers that the total carbon emissions of megacities should be minimized when China's economy grows at a 5% rate in the absence of major shocks. Taking the carbon emissions of megacities as the explained variable, factors are selected from four aspects: economy, population size, energy structure, and electrification rate. The variables involved are logarithmized to obtain the specific formula: lnC T =a+blnGDP+clnP+dlnE N +elnE R Where a, b, c, d, and e are the elastic coefficients corresponding to each variable.

7. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 1 is characterized by: In step 5, based on the prediction model and historical data, the ridge regression method is used to determine the model parameters, and a multi-stage comprehensive optimization method for carbon emission reduction in megacities is obtained under the constraints. The specific algorithm is: Assume the model formula as shown below: y=θx Using the least squares method, the loss function is defined as the square of the residual and the loss function is minimized: ||Xθ-y 2 || θ=(X T X) -1 X T y The above loss function plus a regularization term becomes: ||Xθ-y 2 ||+||Γθ 2 || Make a ridge trace plot and observe at what value the plot tends to be stable, so as to determine the value of α, that is, the coefficient values ​​of each parameter in the multi-scenario prediction model of carbon emission reduction potential of super-large cities mentioned above.

8. The multi-stage comprehensive optimization method for carbon emission reduction in megacities taking into account economic development according to claim 7 is characterized by: According to the different policy requirements of super-large cities, the constraints include: energy supply and demand balance constraints and non-fossil energy power generation ratio constraints. T ES (t)≥T EC (t) Where, T ES (t) is the total energy supply of the megacity in year t, T EC (t) is the total energy consumption of the mega-city in year t, in units of 10,000 tons of standard coal. In the formula, E N (t) is the proportion of non-fossil energy in the total energy consumption of the mega-city in year t, E C (t) is the final consumption of fossil energy in year t, τ t is the lower limit of the proportion of non-fossil energy power generation in year t; The optimal solution with the lowest comprehensive total carbon emissions based on economic development is obtained, and the ultimate output is the optimized combination of influencing factors such as carbon emission trends, peak conditions, and economic development that meet various constraints, and a carbon emission reduction timetable and roadmap for reference is generated.