Regional integrated energy system two-stage optimization method considering efficiency and stepped carbon transaction

Abstract

The invention discloses a regional integrated energy system two-stage optimization method considering # imgabs0 # efficiency and stepped carbon transaction. The method comprises the following steps: constructing a regional integrated energy system model based on an energy concentrator model; establishing a multi-objective optimization model considering the efficiency of the equipment # imgabs1 #, and incorporating the multi-objective optimization model into a stepped carbon transaction mechanism; constructing an RIES two-stage optimization model which comprises an upper-layer capacity configuration model and a lower-layer scheduling model; solving the established RIES two-stage optimization model, solving a Pareto leading edge of equipment capacity configuration by an upper-layer capacity configuration model by adopting an NSGA-III algorithm, and generating a plurality of groups of candidate schemes; the lower-layer scheduling model obtains an optimal scheduling strategy corresponding to each candidate scheme through a solver; and objective weighting is carried out based on an entropy weight method to determine index weight coefficients such as economy, carbon emission intensity and # imgabs2 efficiency, a TOPSIS method is adopted to carry out full-dimension quantitative evaluation on a Pareto leading-edge solution set, and an optimal capacity configuration scheme and optimal scheduling data of the system are selected. According to the method, the system energy efficiency can be improved, the carbon emission intensity is reduced, and the equipment configuration redundancy is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of optimal dispatching of regional integrated energy systems, and specifically relates to a two-stage optimization method for regional integrated energy systems considering efficiency and ladder carbon trading. Background Art

[0002] As the core carrier of energy transformation, the integrated energy system still faces the technical bottleneck of the coordinated optimization of energy efficiency improvement and carbon emission reduction. The existing mechanisms take thermal efficiency or energy conversion efficiency as the core indicators, only focusing on the conversion of energy quantity, but ignoring the difference in energy quality. For example, there is an order-of-magnitude difference in availability between electric energy (energy quality coefficient 1.0) and low-grade thermal energy (energy quality coefficient 0.1 - 0.3). And there are significant defects in the current technical system: the traditional carbon accounting uses the static emission factor method, which cannot accurately quantify the dynamic carbon flow distribution in the multi-energy flow coupling scenario; the evaluation of energy utilization efficiency and the ladder carbon trading mechanism are in a fragmented state. Most of the existing studies take efficiency as an independent evaluation index and do not form a linkage optimization model with carbon price elasticity; most of the existing optimization methods focus on a single time scale, and most of the literature only considers the optimization of the operation dispatching layer, ignoring the spatio-temporal coupling relationship between equipment capacity configuration and operation strategy.

[0003] Specifically, the current carbon trading mechanism generally adopts a fixed pricing model, and the lack of coordination with the thermodynamic efficiency index has become a key obstacle restricting low-carbon development. There are two core contradictions in the current technical system: firstly, the traditional carbon trading mechanism only restricts the total carbon emissions through a linear carbon price and does not establish an elastic pricing model that is dynamically linked with the system efficiency, resulting in high efficiency equipment (such as the power-to-gas system) not being able to obtain economic compensation through the carbon trading market; secondly, most of the current studies take efficiency as an independent evaluation index and do not construct a segmented response mechanism between it and the ladder carbon price, resulting in a fragmented state between the energy quality optimization and carbon cost control objectives. Therefore, it is urgent to deeply couple the ladder carbon trading with the efficiency index and develop a dynamic carbon price- efficiency response model to break through the systematic technical bottleneck of the coordinated optimization of energy efficiency improvement and carbon emission reduction.

[0004] In addition, in the current research on the optimization of integrated energy systems, the lack of a coordination mechanism between capacity planning and dispatching operation is particularly prominent. The existing technical system generally adopts a single-time-scale dispatching strategy, resulting in the fragmentation of the spatio-temporal coupling relationship between equipment configuration and operation strategy. This fragmented optimization mode not only exacerbates the phenomenon of redundant equipment configuration.

[0005] Chinese Patent "Considering In the technical solution disclosed in the "Comprehensive Energy Multi-Objective Planning Method and System for Efficiency and Economy" (Application No.: 202310594321.8), by dividing Efficiency is used as the first objective function to truly reflect the comprehensive energy utilization efficiency. However, this patent does not consider carbon emission factors.

[0006] In the technical solution disclosed in the Chinese patent "Scheduling Method for Source-Load-Storage Collaborative Optimization Considering the Ladder Carbon Trading Mechanism" (Application No.: 202411646619.X), a source-load-storage collaborative optimization scheduling method based on the ladder carbon trading mechanism is disclosed. Through multi-scenario simulation, its feasibility and effectiveness in reducing the system's carbon emissions are verified. However, this patent only conducts research from the scheduling level and fails to fully consider the impact of system equipment capacity configuration on the overall performance, which may lead to redundant or insufficient equipment configuration and restrict the further reduction of carbon emissions. Summary of the Invention

[0007] To solve the above technical problems, the present invention proposes a two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading. The present invention proposes a technical system with " -carbon collaborative optimization" as the core. This method takes the maximization of system efficiency and the minimization of carbon emission costs under the ladder carbon price mechanism as the optimization objectives, constructs a multi-objective decision-making model, and considers the impact of carbon quota allocation on system operation in the optimization framework. At the same time, a two-stage optimization architecture including capacity planning and operation scheduling is designed. The NSGA3 multi-objective optimization algorithm is used in the equipment selection stage to achieve efficiency-oriented capacity configuration, and a cross-time-scale coordination mechanism under carbon cost constraints is established in the scheduling operation stage; this method can improve system energy efficiency, reduce carbon emission intensity, and optimize equipment configuration redundancy.

[0008] The technical solution adopted by the present invention is as follows:

[0009] A two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, including the following steps: Step 1: Construct a regional integrated energy system model based on the energy hub model;

[0010] Step 2: Establish a multi-objective optimization model considering equipment efficiency and incorporate the ladder carbon trading mechanism;

[0011] Step 3: Construct a two-stage optimization model for the RIES, including an upper-layer capacity configuration model and a lower-layer scheduling model;

[0012] Step 4: Solve the two-stage optimization model of RIES established in Step 3. For the upper-layer capacity configuration model, the NSGA-III algorithm is used to solve the Pareto front of the device capacity configuration, generating multiple groups of candidate solutions; for the lower-layer scheduling model, the solver is used to obtain the optimal scheduling strategy corresponding to each candidate solution.

[0013] Step 5: Objectively assign weights based on the entropy weight method to determine the weight coefficients of indicators such as economy, carbon emission intensity, efficiency, etc., and use the TOPSIS method to conduct a full-dimensional quantitative evaluation of the Pareto front solution set, and select the optimal capacity configuration plan and optimal scheduling data of the system.

[0014] In the above Step 1, the regional integrated energy system is modeled based on the Energy Hub model, and the energy coupling characteristics are characterized by the device energy conversion efficiency matrix.

[0015] The energy coupling characteristics of electricity, heat, cold, and gas are quantified through the energy conversion efficiency matrix (C matrix) of the Energy Hub model, covering multiple energy inputs: power grid, gas grid, wind and solar power generation, solar collectors; multiple energy conversion devices: CCHP, PtG, absorption refrigerators, electric refrigerators, gas boilers, electric boilers, and multiple energy storages: electricity, heat, gas; the constructed regional integrated energy system RIES is specifically composed of Figure 1 As shown, a multi-energy flow dynamic balance equation is constructed through matrix operations, and the energy flow distribution ratio and conversion efficiency of device input and output are represented by matrix elements. Specifically as follows:

[0016] The regional integrated energy system model specifically includes:

[0017] The Energy Hub model organically combines the multiple energy inputs, multiple energy conversion devices, multiple energy storages and the load end of the regional integrated energy system through a matrix method, constructing a multi-dimensional coupled framework of the regional integrated energy system;

[0018] L = CP + P s (1);

[0019] In formula (1): L is the electricity, heat, and gas load matrix; P is the input power matrix; C is the coupling coefficient matrix; P s is the energy storage matrix;

[0020]

[0021] In formula (2): L ele 、L gas 、L heat 、L cool are the electricity, gas, cold, and heat loads respectively; are the charging and discharging powers of the electricity, gas, heat, and cold energy storage devices respectively; Pele , P gas , P heat are the input quantities of electric energy, natural gas, and heat energy respectively; are the power generation amounts of WT and PV respectively; is the heat provided to STC; are the purchased electricity quantity and natural gas quantity respectively; The coupling coefficient matrix C is the core part of the energy hub model. Its essence is to quantitatively describe the input-output relationship, energy efficiency characteristics, and multi-energy flow coupling mechanism of each conversion device in the regional integrated energy system through a matrix method. The expression is as follows:

[0022]

[0023] In Equation (3): α, β, and γ are the distribution coefficients of the input electric energy allocated to PtG, EB, and EC respectively; ε and δ are the distribution coefficients of the input natural gas allocated to CCHP and GB respectively; is the efficiency of converting electric energy of PtG into natural gas; are the heat production efficiencies of EB and GB respectively; is the refrigeration coefficient of EC; respectively represent the efficiencies of the CCHP device converting into electricity, heat, and cold; H gas is the calorific value of natural gas.

[0024] In step 2, the efficiency quantification model is introduced. Based on the input (Exin) and the output (Exout) ratio, the system efficiency is defined, as shown in Figure 2 . It can reflect the utilization rate of energy quality. Among them, the input covers the chemical of fossil energy and the grade conversion of renewable energy. The stepped carbon trading mechanism is introduced. The carbon emissions are divided into intervals, and the carbon price rises step by step with the increase of emissions, strengthening the economic incentive of low-carbon scheduling. The carbon quota is dynamically associated with the device output. The carbon emission quotas of devices such as CCHP and gas boilers are allocated according to the benchmark coefficient. Through the multi-objective collaborative optimization of efficiency maximization and carbon cost minimization, the balance between energy quality utilization and environmental externality cost is achieved.

[0025] Construct the system efficiency model:

[0026] Ex in = Ex loss + Ex out (4);

[0027]

[0028] In the above formula: Ex in and Ex out respectively represent the input (payment ), output (benefit ); The efficiency η ex is defined as the ratio of the two, which can effectively reflect the energy utilization situation and characterize the energy efficiency.

[0029] Establish a multi-objective optimization model considering the equipment efficiency, specifically including:

[0030] Definition of the objective function:

[0031] Economic optimum: Minimize the total life cycle cost of the system;

[0032]

[0033] In formula (6), f1 is the total life cycle cost of the system, that is, the first objective function; C 投资,j is the initial investment cost of the jth type of equipment; C 运维,t is the operation and maintenance cost in the tth time period; C 购电,t is the externally purchased electricity cost in the tth time period; C 购气,t is the externally purchased gas cost in the tth time period; n 设备 is the number of types of equipment in the system; T is the total number of time periods in the system life cycle;

[0034] Highest efficiency: Reflect the level of energy quality utilization;

[0035]

[0036] In formula (7), the meaning represented by f2 is the complement of the system efficiency, that is, the second objective function;

[0037] Lowest carbon emissions:

[0038] f3 = E lES,t (8);

[0039] In formula (8), the meaning represented by f3 is the system carbon emissions, that is, the third objective function;

[0040] Constraint handling: Apply a penalty function to individuals that violate the scheduling constraints of the lower-level scheduling model to deteriorate their fitness.

[0041]

[0042] In Equation (9), M is a constant much larger than the normal target value.

[0043] In Step 2, the stepped carbon trading mechanism consists of three parts: the initial carbon emission allowance, the actual carbon emissions of the system, and the stepped carbon trading cost;

[0044] 1) Initial carbon emission allowance:

[0045]

[0046] In the above formula, are the free carbon emission allowances of the RIES, the power purchased from the superior power grid, the CCHP unit, and the gas boiler at time t, respectively; P buy,t and P t CCHP and H CCHP,t and H GB,t represent the power purchase, the power generation of the CCHP, the heat generation of the CCHP, and the heat generation of the gas boiler, respectively; are the benchmark quota coefficients of the power purchase, the CCHP electric power, the CCHP heat power, and the GB heat power, respectively.

[0047] 2) Actual carbon emissions of the system:

[0048] The P2G device absorbs a part of the carbon emissions by consuming electric energy. Therefore, the formula for actual carbon emissions is as follows:

[0049]

[0050] In the above formula, represents the total carbon emissions of the system in the t period; represents the carbon emissions corresponding to the power purchase from the power grid in the t period; represents the carbon emissions of the combined heat and power in the t period; represents the carbon emissions of the gas boiler in the t period; represents the carbon emission quota corresponding to the power purchase from the power grid in the t period; P buy,t represents the power purchase from the power grid in the t period (unit: kWh); represents the carbon emission quota of the combined heat and power (CCHP) in the t period; P t CCHP represents the power generation of the combined heat and power (CCHP) in the t period; H CCHP,t represents the heat generation of the combined heat and power (CCHP) in the t period; represents the carbon emission quota of the gas boiler (GB) in the t period; H GB,t represents the heat generation of the gas boiler (GB) in the t period; are the carbon emission coefficients of the power purchase, the CCHP electric power, the CCHP heat power, and the GB heat power, respectively.

[0051] 3) Stepwise carbon trading cost:

[0052] The stepped carbon trading mechanism incorporates the carbon emission cost into the system economic optimization framework through a segmented pricing method, truly reflecting the non-linear relationship between carbon emissions and carbon price. The core feature of this mechanism is that the carbon price increases stepwise with the increase in emissions. The larger the carbon emissions, the higher the carbon price. The calculation formula for stepped carbon trading is as follows:

[0053] E lES,t = E p - E q (18);

[0054]

[0055] In the formula, represents the carbon trading cost at time t; λ represents the carbon trading benchmark value; l represents the length of the carbon emission interval; α is the carbon trading price growth rate; E lES,t represents the carbon emissions of the system participating in the carbon trading market at time t; represents the carbon trading cost at time t; E p represents the carbon emissions of the system; E q represents the carbon emission quota of the system.

[0056] In step 3, the upper-layer capacity configuration model uses the rated capacity of equipment (CCHP, energy storage, renewable energy installed capacity, etc.) as decision variables, and realizes capacity planning through multi-objective optimization of the whole-life cycle cost (investment + operation and maintenance), system efficiency and carbon emission intensity; the lower-layer operation scheduling model uses the real-time output of equipment, the charge and discharge power of energy storage and the purchased energy quantity as decision variables, and generates a scheduling strategy with the goal of minimizing the single-day operation cost (energy consumption cost + carbon trading cost). Through a two-way coupling mechanism, the upper-layer capacity parameters are embedded into the lower-layer constraint conditions (such as equipment output limit, energy storage capacity boundary), and at the same time, the lower-layer operation cost is fed back to the upper-layer economic objective function, forming a planning-scheduling closed-loop optimization architecture to ensure the physical compatibility and economic synergy of the capacity configuration plan and the real-time operation strategy.

[0057] 1) Upper-layer capacity configuration model:

[0058] Considering the economy and comprehensive energy efficiency of the whole RIES. Taking the total cost C total and efficiency η ex and carbon emissions of the RIES as the objective function, and the decision variables are the capacities of various types of equipment;

[0059]

[0060] In the formula, C inDenote the equipment planning cost; C run Denote the operation cost; D represents the total number of operation days; C in,j , S j , C in,k , S k respectively represent the unit capacity investment cost and configuration scale of the energy production equipment j and the energy storage equipment k; r represents the discount rate, and N represents the planning period.

[0061] The efficiency mathematical model of RIES is as follows:

[0062]

[0063] In the formula, E i,in , E i,out respectively represent the input and output energy of the i-th type of energy, Ω in , Ω out respectively represent the sets of input and output energy forms, λ i is the energy quality coefficient of the i-th type of energy, λ re is the energy quality coefficient of renewable energy; represents the electrical load at time t; represents the gas load at time t; represents the heat load at time t; represents the cooling load at time t; λ ele is the energy quality coefficient of electric energy; λ gas is the energy quality coefficient of natural gas; λ heat is the energy quality coefficient of heat energy; λ cool is the energy quality coefficient of cold energy; is the electricity purchase volume at time is the gas purchase volume at time; P re,t represents the output of renewable energy at time t; T represents the length of a day, 24 hours;

[0064] 2) Lower-level scheduling model:

[0065] The lower-level scheduling model considers the operation level of the system, and the objective function is set as the operation cost;

[0066] The operation cost C run includes: the operation and maintenance cost C op , the electricity and gas purchase cost C ec and the carbon trading cost C ce , in order to achieve the best economy and environment. The decision variables are the output conditions of each equipment unit. Its mathematical model is as follows:

[0067] C run = C op + C ec + C ce (25);

[0068]

[0069] In the formula, are the purchased electricity quantity and natural gas quantity respectively; are the unit prices of electric energy and natural gas respectively; C op,j ,, P j,t are the unit operation and maintenance cost and power of the energy production equipment j respectively; C op,k , P k,t are the unit operation and maintenance cost and power of the energy storage equipment k respectively; are the purchased electricity quantity and natural gas quantity respectively; are the unit prices of electric energy and natural gas.

[0070] The lower-level scheduling model includes the following constraint conditions:

[0071] Energy balance constraint model:

[0072] ①: The power balance constraint is the supply-demand balance constraint of the electricity, gas, heat, and cold energy flows in RIES.

[0073]

[0074] In formula (27), represents the natural gas consumption of the CCHP equipment; represent the electric, heat, and cold power outputs of the CCHP equipment respectively; are the electric power inputs of the PtG, EB, and EC equipment respectively; is the natural gas production of PtG; is the natural gas consumption of GB; are the heat power outputs of EB and GB respectively; is the heat power input of AC; are the cold power outputs of AC and EC respectively;

[0075] represents the wind power generation at time t; represents the photovoltaic power generation at time t; represents the electricity purchased for the system at time t; represents the discharge of the electric energy storage equipment at time t; represents the charging of the electric energy storage equipment at time t; represents the gas purchased for the system at time t; represents the discharge of the gas energy storage equipment at time t; represents the charging of the gas energy storage equipment at time t; represents the output of the solar collector at time t; represents the heat release of the thermal energy storage equipment at time t; Indicates the heat storage of the heat energy storage device at time t; Indicates the natural gas consumption of GB; Respectively indicate the heat power outputs of EB and GB; Indicates the heat power input of AC; Respectively indicate the cooling power outputs of AC and EC; Indicates the cooling release of the cold energy storage device at time t; Indicates the cold storage of the cold energy storage device at time t;

[0076] ②: Unit output ramp constraint:

[0077]

[0078] In formula (28), P j,max , P j,min Are the upper and lower limits of the unit operating power; P j,up , P j,down Are the upper and lower limits of the unit ramp-up capacity; P j,t-1 Indicates the actual output power of unit j in the previous time period; P j,t Indicates the output power of unit j at the current time period t; ③: Renewable energy constraint:

[0079]

[0080] In formula (29): P WT,t The actual output power of wind power generation in time period t; P WT,min Indicates the minimum allowable output of the wind turbine; P WT,max Indicates the maximum available power of the wind turbine; P PV,t The actual output power of photovoltaic power generation in time period t; P PV,min Indicates the minimum allowable output of the photovoltaic; P PV,max Indicates the maximum available power of the photovoltaic; P STC,t The actual output power of the solar collector in time period t; P STC,min Indicates the minimum allowable output of the solar collector; P STC,max Indicates the maximum available power of the solar collector;

[0081] ④: Network constraint:

[0082]

[0083] In formula (30), The power purchase from the power grid by the system at time t; The minimum power purchase from the power grid at time t, which is 0; The maximum power purchase from the power grid; The gas purchase power from the gas network by the system at time t The minimum gas purchase power of the system at time t The maximum gas purchase power of the system at time t;

[0084] ⑤: Energy storage constraint:

[0085]

[0086] In Equation (31), respectively represent the charging and discharging powers of the k-th energy storage method at time t; The rates respectively represent the upper and lower limits of the charging and discharging powers of the k-th energy storage method. S k,t+1 The remaining energy of energy storage device k at time t + 1; S k,t The remaining energy of energy storage device k at time t; The energy storage efficiency of energy storage device k; The energy release efficiency of energy storage device k; The initial capacity of energy storage device k at the beginning of a 24-hour cycle; The capacity of energy storage device k at the end of the cycle.

[0087] In step 4, a hierarchical nested solution strategy is adopted. The upper-layer model uses the NSGA-III algorithm to generate uniformly distributed reference points in the objective space (economy, efficiency, carbon emissions), screens the Pareto front solution set through non-dominated sorting and adaptive normalization, and combines the niche reservation strategy to maintain the diversity of the solution set in the high-dimensional objective space; the lower-layer model is based on the mixed-integer linear programming (MILP) framework, uses the CPLEX solver to handle complex constraints including equipment start-stop states (0-1 variables), energy storage charge-discharge logic (integer constraints), and multi-period power balance (continuous variables), adopts a candidate solution substitution verification mechanism, dynamically embeds the capacity configuration parameters generated by the upper layer into the lower-layer constraint condition set, and eliminates infeasible solutions that violate the operation boundaries to achieve a strict feasibility match between the planning scheme and the scheduling strategy.

[0088] Step 4 includes the following steps:

[0089] S4.1: NSGA-III algorithm initialization:

[0090] Set the population size to 30, the number of iterations to 50, the crossover probability to 0.8, and the mutation probability to 0.2 to avoid the search direction deviation caused by a single parameter in the traditional PSO. Set the capacities of CCHP, AC, EC, EB, and GB devices as constraint conditions.

[0091] S4.2: Initial population generation:

[0092] Randomly generate N individuals within the constrained space CS of Equation (32), where each individual represents a switching combination scheme.

[0093] C i,min ≤C i ≤C i,max (32);

[0094] In Equation (25), C i is the capacity of device i. C i,min , C i,max is the minimum and maximum allowable capacity of device i.

[0095] S4.3: Multi-objective fitness calculation, including Equations (6) to (9);

[0096] S4.4: Non-dominated sorting and reference point association:

[0097] According to the Pareto dominance relationship, divide the population into multiple non-dominated fronts (F1, F2,...). F1 represents the set of all individuals in the current population that are not Pareto-dominated by any other solution; F2 represents the new non-dominated solution set after removing F1;

[0098] Give priority to retaining the solutions in the F1 layer; subsequently, perform adaptive normalization to construct the ideal point f1 min represents the minimum value of the first objective function; represents the minimum value of the second objective function; f3 min represents the minimum value of the third objective function; G min The ideal point should be represented as a set of the minimum values of each objective function;

[0099] Normalize the objective function through the extreme point f i max to eliminate the dimension difference.

[0100]

[0101] In Equation (33), the maximum value of the i-th objective in the current non-dominated front; ∈ is a very small constant to prevent division-by-zero error when f i max = 0; represents the normalized objective function; f i (x) represents the i-th objective function;

[0102] Subsequently, calculate the vertical distance from the individual to the reference point and associate the individual with the nearest reference point to ensure the uniform distribution of the solution set in the objective space.

[0103]

[0104] In formula (34), d is the vertical distance, and f k (x) is the k-th objective function value, is the maximum value of the k-th objective, and r jk is the k-th dimensional coordinate of the reference point j.

[0105] S4.5: Environmental Selection and Offspring Generation:

[0106] Preferentially select the solutions within the reference point area with fewer associated individuals to prevent local aggregation, and perform crossover and mutation randomly.

[0107]

[0108] x new = crossover(x a , x b ) + N(0, δ) (36);

[0109] In the above formula: P j is the crossover point selection probability, and n j is the number of individuals associated with the reference point j; x a , x b are the parent individuals (solution vectors), x new is the newly generated offspring individual, crossover() is the crossover operation function, and N(0, δ) is the Gaussian mutation with a standard deviation of σ;

[0110] S4.6: Iteration Termination Judgment:

[0111] The termination condition is that the change rate of the Pareto front is less than 1% for three consecutive generations, and the preset maximum number of generations is 50. Randomly output the Pareto front solution set.

[0112] The lower-level scheduling model obtains the optimal scheduling strategy corresponding to each candidate solution through the CPLEX solver; specifically:

[0113] Bring the equipment population generated by the upper-level NSGA3 into the lower-level scheduling model as the upper bound constraint of the output of each equipment, and solve it using the CPLEX solver on the premise of satisfying the constraints of the lower-level scheduling model.

[0114] In step 5, based on the entropy weight method, objectively assign weights to the dispersion of the original data of economic efficiency, carbon emission intensity, and efficiency indicators to generate weight coefficients reflecting the internal correlation of each indicator; construct a standardized evaluation matrix through the TOPSIS method, and calculate the distances from each candidate solution to the positive ideal solution (optimal economy, highest efficiency, lowest carbon emissions) and the negative ideal solution (worst economy, The Euclidean distance with the lowest efficiency and the highest carbon emissions is used to fully sort the Pareto front solution set according to the relative closeness. Finally, the solution with the largest relative closeness is selected as the optimal capacity configuration. Finally, a set of data including equipment configuration parameters and whole-life cycle performance data is generated, covering data such as efficiency, investment cost, operation cost, and carbon emissions.

[0115] The TOPSIS method is used to conduct a full-dimensional quantitative evaluation of the Pareto front solution set, select the optimal capacity configuration plan for the system, generate a system scheduling operation plan based on the calculation results, and obtain data such as efficiency, whole-life cycle economic cost (including equipment investment and operation and maintenance costs), and carbon emissions.

[0116] Step 5 includes the following steps:

[0117] S5.1: Data standardization:

[0118] The original data matrix X = {x ij} m×n is processed to be positive, eliminating the dimensional difference. Here, x ij represents the value of the j-th index of the i-th sample, m represents the total number of samples (the number of rows of the matrix), and n represents the number of evaluation indices (the number of columns of the matrix);

[0119] S5.1.1: Conversion of extremely small indices to extremely large ones:

[0120] Either of the following two conditions can be satisfied:

[0121]

[0122] In the above formula, x′ ij represents the value of the j-th extremely small index of the i-th sample in the original data matrix; x ij represents the converted extremely large index value; min(x j ) represents the minimum value of the j-th index among all samples; max(x j ) represents the maximum value of the j-th index among all samples.

[0123] S5.1.2: Conversion of intermediate / interval indices to extremely large ones:

[0124] Through piecewise function adjustment, the data is concentrated towards the ideal value. After standardization, the matrix X′ = {x′ ij} m×n is obtained, where {x′ ij} m×n represents the standardized extremely large index value;

[0125] And the standard ratio is calculated:

[0126]

[0127] In formula (39), p ij represents the standardized ratio of the i-th sample in the j-th index, and ∑p ij = 1;

[0128] S5.2: Calculate the information entropy:

[0129] The information entropy e of the j-th index j reflects the degree of data dispersion:

[0130]

[0131] In formula (40), the entropy value e j ∈[0, 1]. The larger the entropy, the more convergent the index data and the less information;

[0132] S5.3: Calculate the weight:

[0133] Calculate the coefficient of variation g through the information entropy j = 1 - e j , and normalize it to obtain the index weight w j :

[0134]

[0135] In formula (41), g j represents the coefficient of variation of the j-th index; e j represents the information entropy of the j-th index, and k is the index number, k = 1, 2,..., n;

[0136] S5.4: Construct the weighted standardized matrix:

[0137] Combine the standardized data with the weights to generate the weighted matrix Z = {z ij} m×n , where {z ij} m×n represents the weighted standardized matrix, with m rows and n columns;

[0138] z ij = w j ·x' ij (42);

[0139] In formula (42), z ij represents the weighted standardized value of the i-th sample on the j-th index;

[0140] S5.5: Determine the positive and negative ideal solutions:

[0141] Positive ideal solution Z +and the negative ideal solution Z - They are respectively composed of the optimal values and the worst values of each index:

[0142] Z + =(max(z 1j ), max(z 2j ),..., max(z nj ))

[0143] Z - =(min(z 1j ), min(z 2j ), …, min(z nj )) (43);

[0144] In formula (43), max(z 1j ) is the maximum value of all samples on the first index; max(z 2j ) is the maximum value of all samples on the second index; max(z nj ) is the maximum value of all samples on the nth index; min(z 1j ) is the minimum value of all samples on the first index; min(z 2j ) is the minimum value of all samples on the first index; min(z nj ) is the minimum value of all samples on the nth index;

[0145] S5.6: Calculate the Euclidean distances from the samples to the positive and negative ideal solutions:

[0146]

[0147] In formula (44), represents the Euclidean distance from the ith sample to the positive ideal solution; represents the Euclidean distance from the ith sample to the negative ideal solution; is the value of the positive ideal solution on the jth index; is the value of the negative ideal solution on the jth index; z ij is the value of the negative ideal solution on the jth index;

[0148] S5.7: Calculate the relative closeness:

[0149] The comprehensive score C of sample i i represents its closeness to the ideal solution:

[0150]

[0151] In formula (45), C i ∈[0, 1], and the larger the value, the better.

[0152] One kind of the present invention considering Two-stage optimization method for regional integrated energy system with efficiency and ladder carbon trading, and the technical effects are as follows: 1) In step 1 of the present invention, the EnergyHub model is adopted, and the input-output relationship of four-dimensional energy of electricity, heat, cold and gas is abstracted into a coupling coefficient matrix through matrix modeling method, and a collaborative energy supply framework for multi-energy dynamic matching is constructed.

[0153] 2) In step 2 of the present invention, the innovative " -carbon collaborative optimization" multi-objective system is proposed, and the system efficiency maximization and ladder carbon emission minimization are taken as optimization objectives. Compared with the traditional single economic optimization method, this model quantifies the energy quality loss through efficiency, innovatively constructs a ladder carbon trading cost model, dynamically associates the carbon emission with the carbon price through a piecewise pricing function, and internalizes the environmental externality cost into the system operation cost.

[0154] 3) The two-stage optimization architecture constructed in step 3 of the present invention has the upper-layer planning model with efficiency maximization, carbon emission intensity minimization and optimal life-cycle cost as multi-objective functions, and the lower-layer scheduling model constructs the objective of the lowest operation cost based on mixed-integer linear programming, and realizes the two-way coupling mechanism of the double-layer economic indexes through the carbon trading cost constraint.

[0155] 4) The hybrid solution strategy designed in step 4 of the present invention adopts the NSGA-III algorithm to generate the non-dominated solution set in the upper-layer capacity planning stage, and ensures the uniform distribution of the Pareto front in the high-dimensional objective space ( efficiency, carbon emission, economy) through the reference point generation and normalization correlation operation; in the lower-layer scheduling stage, the CPLEX solver is used to process the mixed-integer linear programming problem including the equipment start-stop state (0-1 variable), energy storage charge-discharge logic (integer constraint) and multi-period power balance (continuous variable), and embeds the equipment capacity parameters of the upper-layer planning scheme into the scheduling layer constraint condition set through the dynamic constraint transfer mechanism, strictly ensuring the spatial feasibility and time series continuity of the solution.

[0156] 5) Based on the entropy weight method, the objective weights of indexes such as economy, carbon emission intensity, efficiency, etc. are determined objectively, and the TOPSIS method is used to conduct a full-dimensional quantitative evaluation of the Pareto front solution set, and the optimal capacity configuration scheme of the system and the system scheduling data are selected. The combination of the two not only retains the objectivity of data-driven, but also strengthens the robustness of the evaluation model through algorithm improvement, so that the finally selected capacity configuration scheme reaches the optimal balance in the three objective dimensions at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0157] The present invention will be further described below in conjunction with the drawings and embodiments:

[0158] Figure 1 It is the framework diagram of the regional integrated energy system.

[0159] Figure 2 It is for the system transfer diagram.

[0160] Figure 3 It is the two-stage optimization model of the regional integrated energy system.

[0161] Figure 4 It is the flow chart for solving the two-stage optimization model of the regional integrated energy system.

[0162] Figure 5 It is the three-objective Pareto solution set of the system.

[0163] Figure 6(a) is the dispatching diagram of the system's electric energy operation;

[0164] Figure 6(b) is the dispatching diagram of the system's thermal energy operation;

[0165] Figure 6(c) is the dispatching diagram of the system's cold energy operation; Specific implementation manners

[0166] A two-layer optimization method for capacity configuration and dispatching of a regional integrated energy system (RIES) that comprehensively considers system efficiency, economy, and carbon emissions. This method takes multi-objective collaborative optimization as the core. First, a mathematical model covering efficiency, full-life-cycle economic cost, and stepped carbon trading mechanism is constructed to quantify the impact of energy quality loss and carbon emission constraints on system performance; secondly, through a two-layer optimization framework, collaborative decision-making of capacity planning and operation dispatching is realized. The upper layer determines the capacity configuration of key equipment such as photovoltaic and energy storage with equipment investment cost, efficiency, and carbon trading cost as optimization objectives, and the lower layer optimizes the unit output strategy and multi-energy complementary dispatching scheme based on energy demand; finally, the NSGA-III algorithm is used to solve the Pareto front solution set, and the subjective and objective weights are fused by combining the TOPSIS entropy weight method to screen out the optimal solution that takes into account energy efficient utilization, low-carbon economic operation, and equipment investment benefits.

[0167] The specific steps are as Figure 4 shown, including:

[0168] Step 1: Model the system based on the Energy Hub model, characterize the energy coupling characteristics through the equipment energy conversion efficiency matrix, and construct the multi-energy flow operation constraint conditions;

[0169] Step 2: Establish a multi-objective optimization model considering equipment efficiency and incorporate the stepped carbon trading mechanism;

[0170] Step 3: Construct the RIES two-stage optimization model according to the content of Step 1 and Step 2, which is divided into an upper-layer capacity configuration model and a lower-layer scheduling model;

[0171] Step 4: Solve the two-layer optimization model established in Step 3. The upper-layer model uses the NSGA-III algorithm to solve the Pareto front of the device capacity configuration, generating multiple groups of candidate solutions, and the lower-layer model uses the CPLEX solver to obtain the optimal scheduling strategy corresponding to each candidate solution;

[0172] Step 5: Objectively assign weights based on the entropy weight method to determine the weight coefficients of indicators such as economy, carbon emission intensity, efficiency, etc., and use the TOPSIS method to conduct a full-dimensional quantitative evaluation of the Pareto front solution set, and select the optimal capacity configuration plan and the optimal scheduling data of the system;

[0173] In the said Step 1, the integrated energy system based on the energy hub model integrates multiple energy inputs (power grid, gas network, photovoltaic / wind power, and solar collector), configures multi-energy conversion devices (combined cooling, heating and power unit CCHP, power-to-gas device PtG, absorption chiller AC, electric chiller EC, gas boiler GB, electric boiler EB) and multiple energy storages (electric / thermal / gas), and realizes the dynamic matching of electric, thermal, cooling, and gas four-dimensional loads through multi-energy flow coupling, and constructs a collaborative energy supply system.

[0174] 1.1 Regional integrated energy model based on energy hub:

[0175] The energy hub model organically combines the input end, energy conversion end, storage section, and load end of the energy system through a matrix method, and constructs a multi-dimensional coupled integrated energy system framework.

[0176] L = CP + P s ;

[0177]

[0178] In the formula, L is the electric, thermal, and gas load matrix; P is the input power matrix; C is the coupling coefficient matrix; Ps is the energy storage matrix; L ele 、L gas 、L heat 、L cool are the electric, gas, cooling, and thermal loads respectively; P ele 、P gas 、P heat are the input amounts of electric energy, natural gas, and thermal energy respectively; are the power generations of WT and PV respectively; is the heat provided by STC; are the purchased electricity and natural gas amounts respectively; They are the charging and discharging powers of the electrical, gas, heat, and cold energy storage devices respectively. 1.2 Coupling coefficient matrix:

[0179] The coupling coefficient matrix is the core part of the energy hub model. Its essence is to quantitatively describe the input-output relationships, energy efficiency characteristics, and multi-energy flow coupling mechanisms of various conversion devices in the energy system through matrix methods. The expression is as follows:

[0180]

[0181] In the formula, α, β, and γ are the distribution coefficients of the input electric energy allocated to PtG, EB, and EC respectively; ε and δ are the distribution coefficients of the input natural gas allocated to CCHP and GB respectively; is the efficiency of converting electric energy of PtG into natural gas; are the heat production efficiencies of EB and GB; is the refrigeration coefficient of EC; respectively represent the efficiencies of the CCHP device converted into electricity, heat, and cold; H gas is the calorific value of natural gas.

[0182] In step 2, establishing a multi-objective optimization model considering equipment efficiency and incorporating a stepped carbon trading mechanism includes:

[0183] System construction of the efficiency model and integration of the stepped carbon trading mechanism;

[0184] 2.1. System efficiency model:

[0185] Ex in = Ex loss + Ex out ;

[0186]

[0187] In the formula: Ex in , Ex out respectively represent the input (payment ), output (benefit ); The efficiency η ex is defined as the ratio of the two, which can effectively reflect the energy utilization situation and characterize the energy efficiency.

[0188] 2.2. Stepped carbon trading model:

[0189] The stepped carbon trading mechanism mainly includes three parts: the initial quota of carbon emission rights, the actual carbon emissions of the system, and the stepped carbon trading cost.

[0190] 2.2.1 Initial carbon emission allowances:

[0191]

[0192] In the formula, are the free carbon emission allowances of RIES, power purchased from the superior power grid, CCHP units, and gas boilers at time t, respectively. P buy,t , P t CCHP , H CCHP,t , H GB,t , respectively represent the purchased power, the power generated by CCHP, the heat generated by CCHP, and the heat generated by the gas boiler. respectively represent the benchmark quota coefficients of the purchased electricity, CCHP electric power, CCHP heat power, and GB heat power.

[0193] 2.2.2 System actual carbon emissions:

[0194] The P2G device absorbs a part of carbon emissions by consuming electric energy. Therefore, the actual carbon emission formula is as follows:

[0195]

[0196] In the formula, respectively represent the carbon emission coefficients of the purchased electricity, CCHP electric power, CCHP heat power, and GB heat power.

[0197] 2.2.3 Stepwise carbon trading cost:

[0198] The stepped carbon trading mechanism incorporates the carbon emission cost into the system economic optimization framework through a segmented pricing method. It truly reflects the non-linear relationship between carbon emissions and carbon price. The core feature of this mechanism is that the carbon price increases stepwise with the increase in emissions. The larger the carbon emissions, the higher the carbon price. The calculation formula for stepped carbon trading is as follows:

[0199] E lES,t = E p - E q ;

[0200]

[0201] In the formula, is the carbon trading cost at time t; λ is the carbon trading benchmark value; l is the length of the carbon emission interval; α is the carbon trading price growth rate;

[0202] In step 3, the improved two-layer power grid model is as Figure 1 shown.

[0203] 3.1 Upper-layer capacity planning model:

[0204] The upper layer of the model, i.e., the planning layer, considers the overall economy and comprehensive energy efficiency of the RIES. The total cost C of the RIES total and efficiency η ex and carbon emissions are used as the objective function, and the decision variables are the capacities of various devices.

[0205] C total = C in + C run ;

[0206]

[0207] In the formula, C in,j , S j , C in,k , S k represent the unit capacity investment cost and configuration scale of the energy production device j and the energy storage device k respectively; the discount rate r; the planning period N.

[0208] The mathematical model of the efficiency of the RIES is as follows: In the formula, E

[0209]

[0210] , E i,in , E i,out represent the input and output energy of the i-th type of energy respectively, Ω in , Ω out represent the sets of input and output energy forms respectively, λ i is the energy quality coefficient of the i-th type of energy, and λ re is the energy quality coefficient of renewable energy. The device parameters are shown in Table 1 below.

[0211] Table 1 Device Parameter Table

[0212]

[0213] 3.2. Lower-layer operation scheduling model:

[0214] The operation scheduling model mainly considers the operation level of the system, and the objective function is set as the operation cost. The price parameters are as described in Table 2 below:

[0215] Table 2 Price Parameter Table

[0216]

[0217] The operation cost C run includes two items of energy consumption cost C op (operation and maintenance cost), C ec (cost of purchasing electricity and gas), and C ce(Carbon trading cost) to achieve the best economy and environment. The decision variables are the output of each equipment unit. Its mathematical model is as follows:

[0218] C run = C op + C ec + C ce ;

[0219]

[0220] In the formula, C op,j , P j,t , C op,k , P k,t are the unit operation and maintenance costs and powers of the energy production equipment j and the energy storage equipment k respectively; are the externally purchased electricity and natural gas quantities respectively; are the unit prices of electricity and natural gas.

[0221] 3.3 System constraint conditions:

[0222] 3.3.1 Energy balance constraint model:

[0223] The power balance constraint is the supply-demand balance constraint of the power, gas, heat, and cold energy flows in the RIES.

[0224]

[0225] In the formula, is the natural gas consumption of the CCHP equipment and its power outputs of electricity, heat, and cold; is the electric power input of the PtG, EB, and EC equipment; is the natural gas production of the PtG; is the natural gas consumption of the GB; is the heat power output of the EB and GB; is the heat power input of the AC; is the cold power output of the AC and EC.

[0226] 3.3.2: Unit output ramp constraint:

[0227]

[0228] In the formula, P j,max , P j,min are the upper and lower limits of the unit operation power; P j,up , P j,down are the upper and lower limits of the unit ramp-up capacity.

[0229] 3.3.3: Renewable energy constraint:

[0230]

[0231] The above formulas respectively represent the upper and lower limit constraints of the power of the fan, photovoltaic, and solar collector.

[0232] 3.3.4: Network constraint:

[0233]

[0234] The above formulas respectively represent the upper and lower limits of the power purchased by the RIES from the upstream power grid and the gas grid.

[0235] 3.3.5 Energy storage constraint:

[0236]

[0237] In the formula, respectively represent the charging and discharging power of the k-th energy storage method at time t, The rates respectively represent the upper and lower limits of the charging and discharging power of the k-th energy storage method.

[0238] In step 4, NSGA-III is used to solve the capacity configuration model of the two-stage optimization of the regional integrated energy, including the following steps:

[0239] S401: Algorithm initialization:

[0240] Set the population size to 30, the number of iterations to 50, the crossover probability to 0.8, and the mutation probability to 0.15. To avoid the search direction deviation caused by a single parameter in the traditional PSO, set the device capacity as a constraint condition.

[0241] S402: Generation of the initial population:

[0242] Randomly generate N individuals within the constraint space CS, and each individual represents a switching combination scheme.

[0243] S403: Calculation of multi-objective fitness:

[0244] Definition of the objective function:

[0245] Optimal economy: Minimize the total life cycle cost of the system,

[0246]

[0247] Highest efficiency: Reflect the utilization level of energy quality.

[0248]

[0249] Lowest carbon emissions:

[0250] f3 = E lES,t ;

[0251] Constraint handling: Apply a penalty function to individuals that violate the scheduling constraints of the lower-level model to deteriorate their fitness.

[0252] S404: Non-dominated sorting and reference point association:

[0253] Divide the population into multiple non-dominated fronts (F1, F2,...) according to the Pareto dominance relationship, and preferentially retain the solutions in the F1 layer. Then perform adaptive normalization to construct the ideal point. Through the extreme point f i max Normalize the objective function to eliminate the dimension difference. Then calculate the vertical distance from the individual to the reference point, and associate the individual to the nearest reference point to ensure the uniform distribution of the solution set in the objective space.

[0254] S405: Environmental selection and offspring generation:

[0255] Preferentially select the solutions in the reference point area with fewer associated individuals to prevent local aggregation, and perform crossover and mutation randomly.

[0256] S406: Iterative termination judgment:

[0257] The termination condition is that the change rate of the Pareto front is less than 1% for three consecutive generations, and the preset maximum number of generations is 50. Randomly output the Pareto front solution set.

[0258] Figure 5 It is the Pareto solution set of the three objectives of the system. Figure 5 The three-dimensional Pareto front graph shown intuitively demonstrates the complex trade-off relationship among the three optimization objectives (economic cost, efficiency, carbon emissions). The discrete points on the edge of the graph represent the extreme solutions with a single objective prioritized. The three objectives have strong conflicts. Reducing carbon emissions (Z-axis) requires increasing the economic cost (X-axis) and sacrificing efficiency (Y-axis) at the same time.

[0259] In the said step 5:

[0260] Based on the entropy weight method, objectively assign weights to determine the weight coefficients of indicators such as economy, carbon emission intensity, efficiency, etc., and use the TOPSIS method to solve the Pareto front solution set.

[0261] 5.1 Data standardization:

[0262] Perform positive processing on the original data matrix X = {x ij} m×n to eliminate the dimension difference:

[0263] 5.1.1 Convert the extremely small index to an extremely large one;

[0264] Either of the following two conditions can be satisfied:

[0265]

[0266] 5.1.2 Transformation of Intermediate / Interval-Type Indicators into Positive-Type Indicators:

[0267] Through piecewise function adjustment, the data is concentrated towards the ideal value. After standardization, the matrix X′ = {x′ ij} m×n is obtained, and the standard ratio is calculated.

[0268]

[0269] In the formula, ∑p ij = 1;

[0270] 5.2 Calculation of Information Entropy

[0271] The information entropy e of the jth indicator j reflects the degree of data dispersion:

[0272]

[0273] The entropy value e j ∈[0, 1]. The larger the entropy, the more convergent the indicator data and the less information.

[0274] 5.3 Calculation of Weights

[0275] The difference coefficient g is calculated through information entropy j = 1 - e j , and after normalization, the indicator weight w j is obtained:

[0276]

[0277] 5.4 Construction of Weighted Standardized Matrix

[0278] The standardized data is combined with the weights to generate the weighted matrix Z = {z ij} m×n :

[0279] z ij = w j ·x′ ij

[0280] 5.5 Determination of Positive and Negative Ideal Solutions

[0281] The positive ideal solution Z + and the negative ideal solution Z - are respectively composed of the optimal values and the worst values of each indicator:

[0282] Z+ = (max(z 1j ), max(z 2j ), …, max(z nj ))

[0283] Z - = (min(z 1j ), min(z 2j ), …, min(z nj ))

[0284] 5.6 Calculate the Euclidean distance from the sample to the positive and negative ideal solutions

[0285]

[0286] 5.7 Calculate the relative closeness

[0287] The comprehensive score C of sample i i indicates its closeness to the ideal solution

[0288]

[0289] where C i ∈ [0, 1], and the larger the value, the better

[0290] Generate the system scheduling operation plan as shown in Figures 6(a), 6(b), and 6(c). It can be concluded that this plan exhibits good operating characteristics in multi-energy flow collaborative scheduling: the output curves of each energy unit match the load demand, dynamic balance is achieved among the electric / thermal / cooling energy flows, the energy storage system effectively suppresses the fluctuations of intermittent energy sources, and the overall plan reflects the optimized scheduling strategy of multi-energy complementarity. These diagrams verify the rationality of the system capacity configuration. Through the coordinated cooperation of different energy forms, both the reliability of energy supply and the overall energy efficiency level are ensured

[0291] Table 3 Data table of the system's capacity configuration

[0292]

[0293] The capacity configuration data of the system is shown in Table 3. It can be seen from Table 3 that the capacity configuration plan of this system forms a thermoelectric cooling collaborative supply system with CCHP (15468 kW) as the core energy supply unit, in cooperation with GB (5779 kW) and AC (18759 kW), supplemented by EB (1662 kW) and EC (16789 kW) to achieve efficient conversion of electric energy - thermal energy - cooling energy. At the same time, PTG (313 kW) provides the ability to adjust power to gas. The overall configuration fully reflects the optimized design concept of multi-energy complementarity

[0294] Table 4 System economy, carbon emissions and efficiency table

[0295]

[0296] The economy, carbon emissions and efficiency are shown in Table 4. It can be seen from Table 4 that the total cost is 430,724 yuan, of which the energy purchase cost is only 249,019 yuan, accounting for about 57.8%, indicating that the system has significant advantages in energy procurement economy. At the same time, the system's exergy efficiency reaches 0.669 and the energy utilization efficiency is as high as 0.856, fully reflecting the optimization effect of energy cascade utilization and system integration. The carbon trading cost is controlled within 39,910 yuan. Together with the investment and operation and maintenance cost of 141,795 yuan, it presents a virtuous cycle characteristic of "low energy purchase cost - high energy utilization efficiency" as a whole, verifying the comprehensive advantages of the system in energy conservation and low carbon emissions. This balanced design of economy and energy efficiency reflects the effectiveness of the optimization method proposed in this paper.

Claims

1. A method for two-stage optimization of a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that It includes the following steps: Step 1: Build a regional integrated energy system model based on the energy hub model; Step 2: Establish a multi-objective optimization model considering the equipment efficiency and incorporate a stepped carbon trading mechanism; Step 3: Build a two-stage optimization model for the RIES, including an upper-layer capacity configuration model and a lower-layer scheduling model; Step 4: Solve the two-stage optimization model for the RIES established in Step 3. The upper-layer capacity configuration model uses the NSGA-III algorithm to solve the Pareto front of the device capacity configuration and generate multiple groups of candidate solutions; The lower-layer scheduling model obtains the optimal scheduling strategy corresponding to each candidate solution through a solver; Step 5: Based on the objective weighting of the entropy weight method, determine the weight coefficients of economic efficiency, carbon emission intensity, efficiency indicators, and use the TOPSIS method to conduct a full-dimensional quantitative evaluation of the Pareto front solution set, and select the optimal capacity configuration plan and optimal scheduling data of the system.

2. The two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading according to claim 1, characterized in that: In the said Step 1, the regional integrated energy system model specifically includes: L = CP + P s (1); In Equation (1): L is the electricity, heat, and gas load matrix; P is the input power matrix; C is the coupling coefficient matrix; P s is the energy storage matrix; In formula (2): L ele , L gas , L heat , L cool are the electric, gas, cooling, and heating loads respectively; are the charging and discharging powers of the electric, gas, heat, and cold energy storage devices respectively; P ele , P gas , P heat are the input amounts of electric energy, natural gas, and heat energy respectively; are the power generations of WT and PV respectively; is the heat provided by STC; are the purchased electricity amount and natural gas amount respectively; The coupling coefficient matrix is the core part of the energy hub model. Through the matrix method, the input-output relationships, energy efficiency characteristics, and multi-energy flow coupling mechanisms of various conversion devices in the regional integrated energy system are quantitatively described. The expression is as follows: In Equation (3): α, β, and γ are the distribution coefficients of the input electric energy allocated to PtG, EB, and EC, respectively; ε and δ are the distribution coefficients of the input natural gas allocated to CCHP and GB, respectively; is the efficiency of converting the electric energy of PtG into natural gas; are the heat production efficiencies of EB and GB, respectively; is the refrigeration coefficient of EC; respectively represent the efficiencies of the CCHP device converted into electricity, heat, and cold; H gas is the calorific value of natural gas.

3. A two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that: In step 2, construct the system Efficiency model: Ex in = Ex loss + Ex out (4); In the above formula: Ex in and Ex out respectively represent the input output efficiency η ex is defined as the ratio of the two, which can effectively reflect the energy utilization situation and characterize the energy efficiency.

4. A two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading according to claim 3, characterized in that: Establish a multi-objective optimization model considering equipment efficiency, specifically including: Definition of the objective function: Optimal economy: Minimize the total life-cycle cost of the system; In formula (6), f1 is the total cost of the system's entire life cycle, that is, the first objective function; C 投资,j The initial investment cost of the jth type of equipment; C 运维,t The operation and maintenance cost in the tth time period; C 购电,t The cost of purchased electricity in the tth time period; C 购气,t The cost of purchased gas in the tth time period; n 设备 The number of types of equipment in the system; T is the total number of time periods in the system's entire life cycle; Highest efficiency: reflecting the level of energy quality utilization; In formula (7), the meaning represented by f2 is the system complement of the efficiency, that is, the second objective function; Lowest carbon emissions: f3 = E lES,t (8); In Equation (8), the meaning represented by f3 is the system carbon emissions, that is, the third objective function.

5. The two-stage optimization method for a regional integrated energy system considering efficiency and stepped carbon trading according to claim 4, characterized in that: In Step 2, the stepped carbon trading mechanism includes three parts: the initial carbon emission quota, the actual carbon emissions of the system, and the stepped carbon trading cost; 1) Initial carbon emission quota: In the above formula, are the free carbon emission allowances of the RIES, the electricity purchased from the upstream power grid, the CCHP unit, and the gas boiler at time t, respectively; P buy,t , P t CCHP , H CCHP,t , H GB,t represent the power purchase, the power generation of the CCHP, the heat generation power of the CCHP, and the heat generation power of the gas boiler, respectively; represent the benchmark quota coefficients of the electricity purchase, the electric power of the CCHP, the heat power of the CCHP, and the heat power of the GB, respectively; 2) Actual carbon emissions of the system: The P2G device absorbs a part of the carbon emissions by consuming electric energy. Therefore, the formula for the actual carbon emissions is as follows: In the above formula, represents the total carbon emissions of the system in period t; represents the carbon emissions corresponding to the electricity purchased from the power grid in period t; represents the carbon emissions of the cogeneration in period t; represents the carbon emissions of the gas boiler in period t; represents the carbon emission quota corresponding to the electricity purchased from the power grid in period t; P buy,t represents the electricity purchase volume from the power grid in period t; represents the carbon emission quota of the cogeneration in period t; P t CCHP represents the power generation power of the cogeneration in period t; H CCHP,t represents the heat generation power of the cogeneration in period t; represents the carbon emission quota of the gas boiler in period t; H GB,t represents the heat generation power of the gas boiler in period t; respectively represent the carbon emission factors of electricity purchase volume, CCHP electric power, CCHP heat power, and GB heat power; 3) Stepped carbon trading cost: The core feature of the stepped carbon trading mechanism is that the carbon price rises step by step as the emissions increase. The larger the carbon emissions, the higher the carbon price. The calculation formula for the stepped carbon trading is as follows: E lES,t = E p - E q (18); In the formula, represents the carbon trading cost at time t; λ represents the carbon trading benchmark value; l represents the length of the carbon emission interval; α is the carbon trading price growth rate; E lES,t represents the carbon emissions of the system participating in the carbon trading market at time t; represents the carbon trading cost at time t; E p represents the carbon emissions of the system; E q represents the carbon emission quota of the system.

6. The two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading according to claim 1, characterized in that: In the said Step 3, the upper-layer capacity configuration model specifically includes: Take the total cost C of RIES total and efficiency η ex and carbon emissions as the objective function, and the decision variables are the capacities of various types of equipment; C total = C in + C run (20); Where, C in represents the equipment planning cost; C run represents the operating cost; D represents the total number of operating days; C in,j , S j , C in,k , S k respectively represent the unit capacity investment cost and configuration scale of the energy production equipment j and the energy storage equipment k; r represents the discount rate, and N represents the planning period; of RIES The efficiency mathematical model is as follows: where, E i,in , E i,out respectively represent the input and output energies of the i-th type of energy, Ω in , Ω out respectively represent the sets of input and output energy forms, λ i is the energy quality coefficient of the i-th type of energy, λ re is the energy quality coefficient of renewable energy; represents the electrical load at time t; represents the gas load at time t; represents the heat load at time t; represents the cooling load at time t; λ ele is the energy quality coefficient of electric energy; λ gas is the energy quality coefficient of natural gas; λ heat is the energy quality coefficient of heat energy; λ cool is the energy quality coefficient of cold energy; is the electricity purchase volume at time t; is the gas purchase volume at time t; P re,t represents the output of renewable energy at time t; T represents the duration of a day, 24 hours.

7. A two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that: The lower-layer scheduling model considers the operation level of the system, and the objective function is set as the operation cost; Operating cost C run including: operation and maintenance cost C op , power and gas purchase cost C ec and carbon trading cost C ce , so as to achieve the best economy and environment; the decision variable is the output of each equipment unit; its mathematical model is as follows: C run = C op + C ec + C ce (25); In the formula, are the purchased electricity quantity and natural gas quantity respectively; are the unit prices of electricity and natural gas respectively; C op,j ,, P j,t are the unit operation and maintenance cost and power of energy production equipment j respectively; C op,k , P k,t are the unit operation and maintenance cost and power of energy storage equipment k respectively; are the purchased electricity quantity and natural gas quantity respectively; is the unit price of electricity and natural gas.

8. According to claim 7, a two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that: The lower-layer scheduling model includes constraint conditions: ①: The power balance constraint is the supply-demand balance constraint for the four types of energy flows of electricity, gas, heat, and cold in the RIES; In Equation (27), represents the natural gas consumption of the CCHP device; respectively represent the electrical, thermal, and cooling power outputs of the CCHP device; are the electrical power inputs of the PtG, EB, and EC devices respectively; is the natural gas production of PtG; is the natural gas consumption of GB; are the thermal power outputs of EB and GB respectively; is the thermal power input of AC; are the cooling power outputs of AC and EC respectively; Indicates the power generation of the wind turbine at time t; Indicates the photovoltaic power generation at time t; Indicates the electricity purchase volume of the system at time t; Indicates the discharge of the electrical energy storage device at time t; Indicates the charging of the electrical energy storage device at time t; Indicates the gas purchase volume of the system at time t; Indicates the discharge of the gas energy storage device at time t; Indicates the charging of the gas energy storage device at time t; Indicates the output of the solar collector at time t; Indicates the heat release of the thermal energy storage device at time t; Indicates the heat storage of the thermal energy storage device at time t; Indicates the natural gas consumption of GB; Respectively indicate the thermal power outputs of EB and GB; Indicates the thermal power input of AC; Respectively indicate the cooling power outputs of AC and EC; Indicates the cold release of the cold storage device at time t; Indicates the cold storage of the cold storage device at time t; ②: Unit output ramp-up constraint: In formula (28), P j,max , P j,min are the upper and lower limits of the operating power of the unit; P j,up , P j,down are the upper and lower limits of the ramping capacity of the unit; P j,t-1 represents the actual output power of unit j in the previous time period; P j,t represents the output power of unit j at the current time period t; ③: Renewable energy constraint: In formula (29): P WT,t The actual output power of wind power generation at time t; P WT,min Represents the minimum allowable output of the wind turbine; P WT,max Represents the maximum available power of the wind turbine; P PV,t Actual output power of photovoltaic power generation at time t; P PV,min Indicates the minimum allowable output of photovoltaic; P PV,max Indicates the maximum available power of photovoltaic; P STC,t Actual output power of solar collector at time t; P STC,min represents the minimum allowable output of the solar collector; P STC,max represents the maximum available power of the solar collector; ④: Network constraint: In formula (30), The electric power purchased by the system from the power grid at time t; The minimum power of purchasing electricity from the power grid at time t, which is 0; The maximum power of purchasing electricity from the power grid; The gas purchasing power of the system from the gas grid at time t The minimum gas purchasing power of the system from the gas grid at time t The maximum gas purchasing power of the system from the gas grid at time t; ⑤: Energy storage constraint: In formula (31), respectively represent the charging and discharging powers of the k-th energy storage method at time t; The upper and lower limits of the charging and discharging powers of the k-th energy storage method are respectively represented; S k,t+1 The remaining energy of energy storage device k at time period t + 1; S k,t The remaining energy of energy storage device k at time period t; The energy storage efficiency of energy storage device k; The energy release efficiency of energy storage device k; The initial capacity of energy storage device k at the start of a cycle with a 24-hour period; The capacity of energy storage device k at the end of the cycle.

9. According to claim 7, a two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that: The said Step 4 includes the following steps: S4.1: NSGA-III algorithm initialization: It includes setting the capacities of CCHP, AC, EC, EB, and GB devices as constraint conditions; S4.2: Generation of the initial population: Randomly generate N individuals within the constraint space of Equation (32). Each individual represents a switching combination scheme; C i,min ≤C i ≤C i,max (32); In Equation (25), C i is the capacity of device i; C i,min , C i,max are the minimum and maximum allowable capacities of device i; S4.3: Multi-objective fitness calculation; S4.4: Non-dominated sorting and reference point association: According to the Pareto dominance relationship, divide the population into multiple non-dominated fronts (F1, F2,...). F1 represents the set of all individuals in the current population that are not Pareto-dominated by any other solution; F2 represents the new non-dominated solution set after removing F1; Prioritize retaining the solution of layer F1; subsequently, perform adaptive normalization to construct the ideal point G min =(f1 min , f2 min , f3 min ), where f1 min represents the minimum value of the first objective function; f2 min represents the minimum value of the second objective function; f3 min represents the minimum value of the third objective function; G min represents that the ideal point should be expressed as a set of the minimum values of each objective function; Through the extreme point f i max Normalize the objective function to eliminate the dimension difference; In Equation (33), it is the maximum value of the $i$-th objective in the current non-dominated front; $\epsilon$ is a very small constant to prevent division-by-zero error when $f$ i max = 0; represents the normalized objective function; $f$ i (x) represents the $i$-th objective function; Subsequently, calculate the vertical distance from the individual to the reference point, and associate the individual with the nearest reference point to ensure that the solution set is evenly distributed in the objective space; In Equation (34), d is the vertical distance, and f k (x) is the k-th objective function value, is the maximum value of the k-th objective, r jk is the k-th dimensional coordinate of the reference point j; S4.5: Environmental selection and offspring generation: Prioritize selecting the solutions in the reference point area with fewer associated individuals to prevent local aggregation, and randomly perform crossover and mutation; x new = crossover(x a , x b ) + N(0, δ) (36); In the above formula: P j is the intersection point selection probability, and n j is the number of individuals associated with the reference point j; x a , x b is the parent individual (solution vector), x new is the newly generated offspring individual, crossover() is the crossover operation function, and N(0, δ) is a Gaussian mutation with a standard deviation of σ; S4.6: Iterative termination judgment: The termination condition is that the change rate of the Pareto front is less than 1% for three consecutive generations, reaching the preset maximum number of generations of 50; randomly output the Pareto front solution set.

10. A two-stage optimization method for a regional integrated energy system considering efficiency and ladder carbon trading, characterized in that: Step 5 includes the following steps: S5.1: Data standardization: For the original data matrix X = {x ij}, m×n perform a positive transformation to eliminate the dimensional difference, where x ij represents the j-th index value of the i-th sample, m represents the total number of samples, and n represents the number of evaluation indicators; S5.1.1: Conversion of extremely small indicators to extremely large indicators: Either of the following two conditions is satisfied: In the above formula, x′ ij represents the value of the j-th extremely small type index of the i-th sample in the original data matrix; x ij represents the converted extremely large type index value; min(x j ) represents the minimum value of the j-th index among all samples; max(x j ) represents the maximum value of the j-th index among all samples; S5.1.2: Transformation of Intermediate / Interval-Type Indexes into Positive-Type Indexes: Through piecewise function adjustment, the data is concentrated towards the ideal value. After standardization, the matrix X′={x i ′ j} m×n , where {x i ′ j} m×n represents the extremely large type index value after standardization; And calculate the standard ratio: In formula (39), pi j represents the standardized proportion of the i-th sample in the j-th index, and ∑pi j = 1; S5.2: Calculate the information entropy: The information entropy e of the j-th index j Reflecting the degree of data dispersion: In formula (40), the entropy value e j ∈[0,1]. The larger the entropy, the more the index data converges, and the less the information content is; S5.3: Calculate the weights: Calculate the difference coefficient g through information entropy j = 1 - e j , and normalize it to obtain the index weight w j : In formula (41), g j represents the coefficient of variation of the j-th index; e j represents the information entropy of the j-th index, k is the index number, k = 1, 2,..., n; S5.4: Construct a weighted standardized matrix: Combine the standardized data with the weights to generate a weighted matrix Z = {z ij}, where {z m×n} ij} m×n represents the weighted standardized matrix, with m rows and n columns; z ij = w j ·x' ij (42); In formula (42), z ij represents the weighted standardized value of the i-th sample on the j-th index; S5.5: Determine the positive and negative ideal solutions: Positive ideal solution Z + and negative ideal solution Z - are respectively composed of the optimal values and the worst values of each index: Z + = (max(z 1j ), max(z 2j ), …, max(z nj )) Z - = (min(z 1j ), min(z 2j ), …, min(z nj )) (43); In formula (43), max(z 1j ) is the maximum value of all samples on the first index; max(z 2j ) is the maximum value of all samples on the second index; max(z nj ) is the maximum value of all samples on the nth index; min(z 1j ) is the minimum value of all samples on the first index; min(z 2j ) is the minimum value of all samples on the first index; min(z nj ) is the minimum value of all samples on the nth index; S5.6: Calculate the Euclidean distances from the samples to the positive and negative ideal solutions: In formula (44), represents the Euclidean distance from the i-th sample to the positive ideal solution; represents the Euclidean distance from the i-th sample to the negative ideal solution; the value of the positive ideal solution on the j-th index; the value of the negative ideal solution on the j-th index; z ij the value of the negative ideal solution on the j-th index; S5.7: Calculate the relative closeness: The comprehensive score C of sample i i indicating its closeness to the ideal solution: In formula (45), C i ∈ [0, 1], and the larger the value, the better.

Images