Adaptive optimization method for electricity-hydrogen-gas comprehensive energy system oriented to multi-dimensional carbon asset collaborative transaction

Abstract

The invention discloses an electricity-hydrogen-gas comprehensive energy system self-adaptive optimization method oriented to multi-dimensional carbon asset collaborative transaction, and the method comprises the following steps: 1, introducing a dynamic emission factor and a multi-dimensional carbon asset collaborative transaction mechanism, and constructing a carbon emission optimization model of an electricity-hydrogen-gas three-energy coupling system; 2, performing dynamic allocation of carbon emission responsibility by adopting a game equilibrium-deep learning hybrid model; step 3, taking minimization of energy supply cost, carbon emission cost and green certificate-carbon transaction cost as targets, and constructing an electricity-hydrogen-gas coupled comprehensive energy system scheduling model; 4, simulation is carried out, and the effectiveness of the optimization model and method is verified; according to the invention, the cooperative efficiency of the power system, the natural gas network and the hydrogen energy system is effectively improved, so that the energy utilization is optimized, and the carbon emission of the system is reduced.

Description

Technical Field

[0001] This invention relates to the field of integrated energy system optimization and operation technology, and in particular to an adaptive optimization method for an integrated energy system of electricity, hydrogen and gas oriented towards multidimensional carbon asset collaborative trading. Background Technology

[0002] Against the backdrop of global warming and escalating energy security risks, the energy sector faces unprecedented challenges. To address these challenges, global policies have strengthened carbon emission control and the promotion of renewable energy, posing new requirements for the transformation and development of energy systems. In this process, the power industry, as a major contributor to carbon emissions, urgently needs to balance emission reduction with economic viability while ensuring power supply. To achieve this goal, traditional power system optimization methods are increasingly unable to meet the growing complexity of the demands. Therefore, it is essential to develop market-based mechanisms such as green certificate trading and carbon trading to guide the green transformation of the power system. The green certificate system and the carbon trading market have become core mechanisms for promoting the construction of low-carbon power systems. The green certificate system provides an incentive for clean energy investment by confirming the generation of renewable energy. The carbon trading market, on the other hand, uses price mechanisms to constrain corporate carbon emissions and optimize production and operations. The synergistic effect of both not only effectively promotes the consumption of renewable energy but also provides strong market support for carbon emission reduction.

[0003] However, while electro-gas coupling systems have optimized energy supply and use to some extent, current carbon emission management mechanisms still face numerous challenges. Existing carbon trading and green certificate systems lack flexible interaction mechanisms and cannot adequately address carbon price fluctuations and changes in load demand, leading to high costs and risks for enterprises in fulfilling their obligations. Therefore, how to utilize market mechanisms to optimize carbon emission management and improve the overall economic and environmental performance of energy systems has become an urgent technical problem to be solved. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned shortcomings and provide an adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading, thereby improving the synergistic efficiency of the power system, natural gas network and hydrogen energy system, thereby optimizing energy utilization and reducing system carbon emissions.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by this invention is: an adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multi-dimensional carbon asset collaborative trading, comprising the following steps:

[0006] Step 1: Introduce a dynamic emission factor and a multi-dimensional carbon asset collaborative trading mechanism to construct a carbon emission optimization model for an electricity-hydrogen-gas three-energy coupling system;

[0007] Step 2: Use a game equilibrium-deep learning hybrid model to dynamically allocate carbon emission responsibility;

[0008] Step 3: With the goal of minimizing energy supply costs, carbon emission costs, and green certificate-carbon trading costs, construct an integrated energy system scheduling model that couples electricity, hydrogen, and gas.

[0009] Step 4: Perform simulations to verify the effectiveness of the optimization model and method.

[0010] Preferably, step 1 specifically includes the following process:

[0011] By introducing hydrogen energy modules, the synergistic optimization of electricity, natural gas, and hydrogen energy can be achieved. The hydrogen energy modules include:

[0012] The P2H water electrolysis hydrogen production unit utilizes surplus electricity from the power grid to produce hydrogen.

[0013] Hydrogen energy storage unit (H2S): Stores excess hydrogen in a storage tank;

[0014] Hydrogen re-conversion unit (H2P): When electricity demand is high or natural gas is scarce, hydrogen is converted into electricity through fuel cells / hydrogen-blended generator sets or injected into the natural gas pipeline network;

[0015] The power balance relationship of the hydrogen energy module is as follows:

[0016]

[0017] In the formula: P el (t) represents the hydrogen production power from water electrolysis; η el Electrolysis efficiency; H prod (t) represents the hydrogen production rate; H stor (t) represents the hydrogen storage capacity; H cons (t) represents the amount of hydrogen consumed; P fc (t) represents the power generated by the fuel cell; η fc For fuel cell efficiency;

[0018] Through the above mechanism, hydrogen energy modules can play a role in peak shaving and valley filling between peak and off-peak loads in the power system, while providing backup and low-carbon energy for the system.

[0019] Introducing a dynamic emission factor, which changes dynamically with system load levels and carbon asset prices, the dynamic emission factor is defined as follows:

[0020]

[0021] In the formula: DEF(t) is the dynamic emission factor at time t; EF base L is the baseline emission factor, i.e., the standard emission per unit of electricity; L(t) is the total load of the system at time t; L maxP represents the system's maximum load. carbon (t) represents the carbon price at time t; The highest carbon price; α, β, γ are weighting coefficients, satisfying α+β+γ=1;

[0022] Combining dynamic emission factors, the total carbon emissions of the system are expressed as:

[0023]

[0024] In the formula: C total (t) represents the total carbon emissions of the system at time t; P i (t) represents the power generation of unit i; η i is the fuel emission coefficient of the unit; N is the total number of units in the system.

[0025] Preferably, step 2 specifically includes the following process: The game theory model analyzes the behavior of each participant in terms of carbon emissions by defining the payoff and cost functions of the participants. The participants' goal is to minimize their carbon compliance costs while complying with the market's carbon trading rules and constraints. The interaction between the participants is solved using Nash equilibrium; the carbon emissions of the power producer are set as C. i Its carbon emission cost is C cost,i Then the game equilibrium model is expressed as:

[0026] C cost,i =α i ·C i +β i ·P carbon ·max(0,C i -C quota,i );

[0027] C i Carbon emissions of producer i; α i β is the carbon emission factor of producer i. i P is the carbon trading weighting coefficient. carbon For carbon prices in the carbon trading market; C quota,i Carbon allowances for producers;

[0028] In the game theory model, the Nash equilibrium point is found by solving the carbon emission behavior of different participants, so as to minimize the carbon emission cost of the overall system.

[0029] Based on the game equilibrium model, the deep learning model predicts future carbon emission trends by training on carbon emission data, thereby dynamically adjusting the parameters in the game equilibrium; it predicts time-series carbon emission data through the deep learning model, thereby achieving dynamic allocation of carbon emission responsibility; the deep learning model is trained using a multilayer perceptron (MLP) or long short-term memory (LSTM) network, and learns the system's emission patterns by inputting historical carbon emission data, load data, and carbon price data, and outputs the carbon emission amount for each producer;

[0030] The following steps are used to solve the problem:

[0031] A deep learning model is used to train the carbon emission time series data to obtain the carbon emission prediction values ​​at each time point.

[0032] Based on game equilibrium theory, calculate each producer's carbon emission responsibility under different carbon prices and emission levels;

[0033] The objective function is solved using the particle swarm optimization algorithm to obtain the optimal carbon emission responsibility allocation and trading path.

[0034] Preferably, step 3 specifically includes the following process:

[0035] The objective function is to minimize the sum of system energy supply cost, carbon emission cost, and green certificate-carbon-hydrogen trading cost. The objective function is as follows:

[0036] minF1=f e +f g +f g,c,h +f c ;

[0037] In the formula: F1 is the economic scheduling cost; f e The total fuel cost of the power plant; f g Cost of natural gas well extraction; f g,c,h Costs of green certificate-carbon-hydrogen trading; f c This refers to the cost of carbon emissions; the formulas for calculating each cost are as follows:

[0038]

[0039]

[0040] f c =C cost,i ;

[0041] In the formula: P gird,t Let a be the power generation capacity of power plant i at time t; i b i and c i Let be the power generation coefficient of unit i; The unit extraction price of natural gas at time t; Let be the output of the natural gas well at time t; I and J are the total number of power plants and gas wells in the system, respectively. Costs of green certificates and carbon trading; C cost,i Cost of carbon emissions; Cost of hydrogen energy modules.

[0042] Preferably, the hydrogen energy module involves the electrolysis of water to produce hydrogen, storage, and re-conversion processes, and its cost can be expressed as:

[0043]

[0044] Wherein: H prod (t) represents the hydrogen production power; Unit cost of hydrogen production unit; H stor (t) represents the hydrogen storage capacity; For hydrogen storage costs; H fc (t) represents the output power of the fuel cell; The cost of converting hydrogen energy into electricity.

[0045] Preferably, the constraints in step 3 include: power grid flow constraints, natural gas grid constraints, and green certificate-carbon trading mechanism constraints, wherein the green certificate-carbon trading mechanism constraints include carbon trading segment constraints, green certificate segment constraints, interval constraints, and quota balance constraints.

[0046] Preferably, the power flow constraint adopts DC power flow constraint as the power network constraint, and the constraint conditions include node power balance constraint, phase angle constraint, unit output constraint and transmission line constraint.

[0047] Preferably, in the natural gas network constraints, the natural gas network consists of gas wells, pipelines, and gas load; its constraints consider node supply and demand balance constraints, node pressure constraints, gas network pipeline constraints, and gas source constraints, as follows:

[0048] Q out,t +Q P2G,t =Q flow,t +Q need,t +Q load,t ;

[0049] Q out,t,min ≤Q out,t ≤Q out,t,max ;

[0050]

[0051] Q flow,min,t ≤Q flow,i,t ≤Q flow,max,t ;

[0052] In the formula: Q out,t Q represents the outflow of natural gas. P2G,t The natural gas flow rate converted by the P2G (Power-to-Gas) equipment; Q flow,t Q represents the pipeline natural gas flow rate. need,t For natural gas demand; Q load,t Q represents the amount of natural gas required for the gas load; out,t,min Q represents the minimum natural gas flow rate. out,t,max The maximum natural gas flow rate; sign(Q) flow,i,t () represents the direction of pipeline natural gas flow, with 1 for the positive direction and -1 for the negative direction; GP is the gas pressure coefficient of pipeline natural gas. i,t The initial pipeline pressure; GP j,t Q represents the gas pressure at the end of the pipeline; flow,min,t Q represents the minimum pipeline natural gas flow rate. flow,max,t This represents the maximum flow rate of natural gas in the pipeline.

[0053] Preferably, the constraints of the green certificate-carbon trading mechanism specifically include:

[0054] Carbon trading segmentation constraints:

[0055]

[0056] In the formula: z k The variable is a binary variable, determining the stage in which constraints are applied; ΔE1, ΔE2, and ΔE3 represent the excess carbon emissions for stages 1, 2, and 3, respectively; I1, I2, and I3 represent the carbon allowances for stages 1, 2, and 3, respectively; M is a relatively large upper bound; z k ∈{0,1} indicates whether the k-th stage carbon price mechanism is enabled; z1, z2, and z3 are binary variables of stages 1, 2, and 3, respectively, and are either 1 or 0.

[0057] Green certificate segmented constraints:

[0058]

[0059] In the formula: y j It is a binary variable; This represents the maximum number of tradable green certificates in the first phase. The maximum number of tradable green certificates for the second phase; U k The upper bound of the green certificate trading constraint is defined by y1, y2, and y3, which are binary variables for each stage; N1, N2, and N3 are the tradable green certificates.

[0060] Interval constraints:

[0061] To ensure that the allowances obtained through trading or offsetting are sufficient to cover a certain range of emission intervals, the following constraint is established:

[0062]

[0063] In the formula: μ is the conversion coefficient, which can convert a certain amount of green certificate transactions into offset carbon emissions; l represents the starting point of the current emission ladder; h is the higher ladder position that we hope to avoid entering through offsetting measures.

[0064] Quota balance constraints:

[0065]

[0066] Where: E is the total carbon emissions; E0 is the free carbon allowance; ΔE i This represents the carbon emissions at each stage.

[0067] Preferably, step 4 uses an improved IEEE 39-node system and a Belgian 20-node system for simulation to verify the advantages of the proposed method in carbon emission and cost optimization compared with traditional methods in an electric-hydrogen-gas tri-energy coupling system.

[0068] Beneficial effects of this invention:

[0069] 1. Improved system flexibility and reliability: By introducing hydrogen energy modules, electricity, natural gas, and hydrogen energy complement each other. Hydrogen energy has the ability to be produced, stored, and reconverted through electrolysis. When load fluctuations or renewable energy fluctuations are severe, it can act as a buffer and regulation mechanism for the system, effectively improving the system's operational stability and flexibility.

[0070] 2. Dynamic low-carbon optimization is achieved. The dynamic emission factor proposed in this invention can dynamically adjust the carbon emission accounting value according to the load level and the price of multi-dimensional carbon assets (carbon certificates, green certificates, hydrogen certificates), making the system scheduling more in line with real-time market and environmental policies. This avoids the problems of lagging and inaccurate carbon accounting under the traditional fixed emission factor, thereby achieving more scientific low-carbon optimization.

[0071] 3. Fair and efficient carbon emission responsibility allocation: The game equilibrium-deep learning hybrid model is used to dynamically allocate carbon emission responsibility. This not only improves the prediction accuracy of allocation, but also achieves equilibrium through game theory with the participation of multiple parties. This avoids the unfair phenomenon of high-emission units having too low a responsibility and low-emission units having too high a responsibility, thereby incentivizing the priority consumption of clean energy.

[0072] 4. The method of this invention, by introducing a hydrogen energy module and a dynamic emission factor, combined with a multi-certificate collaborative trading mechanism, effectively improves the synergistic efficiency of the power system, natural gas network and hydrogen energy system, thereby optimizing energy utilization and reducing system carbon emissions. Attached Figure Description

[0073] Figure 1This is a flowchart illustrating an adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading.

[0074] Figure 2 This is a step carbon valence diagram in an embodiment of the present invention;

[0075] Figure 3 This is a framework diagram of the integrated electric-gas-hydrogen energy system in an embodiment of the present invention;

[0076] Figure 4 This is a schematic diagram of the IEEE-39 node power system and the Belgian 20 node natural gas system in an embodiment of the present invention;

[0077] Figure 5 This is a diagram showing the carbon quota allocation for the generating units in an embodiment of the present invention;

[0078] Figure 6 This is a diagram showing the output of each unit in an embodiment of the present invention. Detailed Implementation

[0079] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0080] Example 1: This example discloses an adaptive optimization method for an integrated electricity-hydrogen-gas energy system for multidimensional carbon asset collaborative trading. The workflow is shown in the figure below. Figure 1 As shown, the method specifically includes the following steps:

[0081] Step 1: A hydrogen energy module was introduced to achieve synergistic optimization of three energy sources: electricity, natural gas, and hydrogen. Through the electrolysis, storage, and conversion of hydrogen, the system's flexibility was enhanced, especially under conditions of large load fluctuations. Furthermore, a dynamic emission factor was adopted, adjusting carbon emission levels in real time based on changes in multidimensional carbon asset prices and system load to adapt to constantly changing market demands and policy requirements. The hydrogen energy module includes:

[0082] Electrolysis of water to produce hydrogen (P2H): Utilizes surplus electricity from the power grid to produce hydrogen;

[0083] Hydrogen energy storage unit (H2S): Stores excess hydrogen in a storage tank;

[0084] Hydrogen re-conversion unit (H2P): When electricity demand is high or natural gas is scarce, hydrogen is converted into electricity through fuel cells / hydrogen-blended generator sets or injected into the natural gas pipeline network.

[0085] The power balance relationship of the hydrogen energy module is as follows:

[0086]

[0087] In the formula: P el(t) represents the hydrogen production power from water electrolysis; η el Electrolysis efficiency; H prod (t) represents the hydrogen production rate; H stor (t) represents the hydrogen storage capacity; H cons (t) represents the amount of hydrogen consumed; P fc (t) represents the power generated by the fuel cell; η fc This improves fuel cell efficiency. Through the above mechanism, the hydrogen energy module can play a role in peak shaving and valley filling between peak and off-peak loads in the power system, while providing backup and low-carbon energy for the system.

[0088] To make carbon emission accounting more accurate, this invention introduces a dynamic emission factor (DEF), which changes dynamically with system load levels and carbon asset prices. The dynamic emission factor is defined as follows:

[0089]

[0090] In the formula: DEF(t) is the dynamic emission factor at time t; EF base L is the baseline emission factor (standard emission per unit of electricity); L(t) is the total load of the system at time t; L max P represents the system's maximum load. carbon (t) represents the carbon price at time t; The highest carbon price; α, β, γ are weighting coefficients, satisfying α+β+γ=1.

[0091] This formula reflects: a baseline emission level, which ensures that the emission factor has a basic value; the emission factor rises during periods of high load, strengthening the pressure to reduce emissions; and carbon price feedback, where the emission factor increases accordingly when the carbon price rises, incentivizing the system to reduce emissions.

[0092] Combining dynamic emission factors, the total carbon emissions of the system are expressed as:

[0093]

[0094] In the formula: C total (t) represents the total carbon emissions of the system at time t; P i (t) represents the power generation of unit i; η i , where is the fuel emission factor of the unit (e.g., the differentiation factor for coal-fired and gas-fired units); N is the total number of units in the system. By introducing dynamic emission factors, carbon emission accounting can be updated in real time according to system operating conditions and market environment, making the optimized scheduling results closer to reality.

[0095] Step 2: This invention employs a combination of game equilibrium theory and deep learning algorithms to optimize the allocation of carbon emission responsibility. The game equilibrium method analyzes the interactions between different participants (i.e., electricity producers, natural gas suppliers, etc.) to ensure that each participant makes reasonable decisions based on their carbon emission responsibilities, thereby achieving optimal carbon emissions for the entire system.

[0096] Game theory models analyze the carbon emission behavior of participants by defining their payoff and cost functions. Participants aim to minimize their carbon compliance costs while adhering to market carbon trading rules and constraints. The interactions between participants can be solved using Nash equilibrium, a state where, given that the behavior of other participants is fixed, no single participant can achieve a better outcome by unilaterally changing their strategy. Let the carbon emissions of the electricity producer be C. i Its carbon emission cost is C cost,i Then the game equilibrium model can be expressed as:

[0097] C cost,i =α i ·C i +β i ·P carbon ·max(0,C i -C quota,i (4)

[0098] C i Carbon emissions of producer i; α i P is the carbon emission factor of producer i. carbon For carbon prices in the carbon trading market; C quota,i Carbon allowances for producers.

[0099] In the game theory model, the Nash equilibrium point is found by solving the carbon emission behavior of different participants, so as to minimize the carbon emission cost of the overall system.

[0100] Based on game equilibrium models, deep learning models, trained on carbon emission data, predict future carbon emission trends, thereby dynamically adjusting parameters in the game equilibrium. For example, when a rise in carbon prices is predicted, the system can adjust the allocation of carbon emissions based on feedback signals from deep learning, incentivizing high-emission producers to reduce their emissions. This invention uses deep learning models to predict time-series carbon emission data, thereby achieving dynamic allocation of carbon emission responsibility. The deep learning model is trained using a multilayer perceptron (MLP) or long short-term memory (LSTM) network. Inputting historical carbon emission data, load data, and carbon price data, it learns the system's emission patterns and outputs the carbon emissions for each producer.

[0101] The following steps are used to solve the problem:

[0102] 1. Use deep learning models (such as LSTM networks) to train on carbon emission time series data to obtain carbon emission predictions for each time point;

[0103] 2. Based on game equilibrium theory, calculate the carbon emission responsibility of each producer under different carbon prices and emission levels;

[0104] 3. Use optimization algorithms (particle swarm optimization) to solve the objective function in order to obtain the optimal carbon emission responsibility allocation and trading path.

[0105] Step 3: This invention aims to minimize the sum of energy supply costs, carbon emission costs, and green certificate-carbon trading costs. It constructs an integrated electricity-hydrogen-gas energy system scheduling model, the framework diagram of which is shown below. Figure 2 As shown. The economic dispatch model is an economic model of an integrated energy system coupled with electricity, hydrogen, and gas based on a green certificate-carbon trading mechanism. Its objective function is to minimize the sum of the system's energy supply cost, carbon emission cost, and green certificate-carbon-hydrogen trading cost. The objective function is:

[0106] minF1=f e +f g +f g,c,h +f c (5)

[0107] In the formula: F1 is the economic scheduling cost; f e The total fuel cost of the power plant; f g Cost of natural gas well extraction; f g,c,h Costs of green certificate-carbon-hydrogen trading; f c This refers to the cost of carbon emissions. The formulas for calculating each cost are as follows:

[0108]

[0109] f c =C cost,i (9)

[0110] In the formula: P gird,t Let a be the power generation capacity of power plant i at time t; i b i and c i Let be the power generation coefficient of unit i; The unit extraction price of natural gas at time t; Let be the output of the natural gas well at time t; I and J are the total number of power plants and gas wells in the system, respectively. Costs of green certificates and carbon trading; C cost,i Cost of carbon emissions.

[0111] The hydrogen energy module involves water electrolysis to produce hydrogen, storage, and re-conversion, and its cost can be expressed as:

[0112]

[0113] Wherein: H prod (t) represents the hydrogen production power; Unit cost of hydrogen production unit; H stor (t) represents the hydrogen storage capacity; For hydrogen storage costs; H fc (t) represents the output power of the fuel cell; The cost of converting hydrogen energy into electricity.

[0114] The following constraints apply to step 3:

[0115] 1) Power flow constraints

[0116] This invention uses DC power flow constraints as power network constraints, and the constraints include node power balance constraints, phase angle constraints, unit output constraints, and transmission line constraints.

[0117] 2) Natural gas network constraints

[0118] The natural gas network mainly consists of gas wells, pipelines, and gas load. Its constraints primarily consider node supply and demand balance constraints, node pressure constraints, gas network pipeline constraints, and gas source constraints, as follows:

[0119] Q out,t +Q P2G,t =Q flow,t +Q need,t +Q load,t (11)

[0120] Q out,t,min ≤Q out,t ≤Q out,t,max (12)

[0121]

[0122] Q flow,min,t ≤Q flow,i,t ≤Q flow,max,t (14)

[0123] In the formula: Q out,t Q represents the outflow of natural gas. P2G,t The natural gas flow rate converted by the P2G (Power-to-Gas) equipment; Q flow,t Q represents the pipeline natural gas flow rate. need,t For natural gas demand; Q load,t Q represents the amount of natural gas required for the gas load; out,t,min Q represents the minimum natural gas flow rate. out,t,maxThe maximum natural gas flow rate; sign(Q) flow,i,t () represents the direction of pipeline natural gas flow, with 1 for the positive direction and -1 for the negative direction; GP is the gas pressure coefficient of pipeline natural gas. i,t The initial pipeline pressure; GP j,t Q represents the gas pressure at the end of the pipeline; flow,min,t Q represents the minimum pipeline natural gas flow rate. flow,max,t This represents the maximum flow rate of natural gas in the pipeline.

[0124] 3) Constraints of the Green Certificate-Carbon Trading Mechanism

[0125] 3.1 Segmented Constraints on Carbon Trading

[0126]

[0127] In the formula: z k The variable is a binary variable, determining the stage in which constraints are applied; ΔE1, ΔE2, and ΔE3 represent the excess carbon emissions for stages 1, 2, and 3, respectively; I1, I2, and I3 represent the carbon allowances for stages 1, 2, and 3, respectively; M is a relatively large upper bound; z k ∈{0,1} indicates whether the k-th stage carbon price mechanism is enabled; z1, z2, and z3 are binary variables of stages 1, 2, and 3, respectively, and are either 1 or 0;

[0128] 3.2 Segmented Constraints on Green Certificates

[0129]

[0130] In the formula: y j It is a binary variable; This represents the maximum number of tradable green certificates in the first phase. The maximum number of tradable green certificates for the second phase; U k y1, y2, and y3 are the upper bounds of the green certificate trading constraints; y1, y2, and y3 are the binary variables for each stage; N1, N2, and N3 are the tradable green certificates for each stage.

[0131] 3.3 Interval Constraints

[0132] To ensure that the quotas obtained through trading or offsetting are sufficient to cover a certain range of emission intervals, constraints are established as shown in equation (11):

[0133]

[0134] In the formula: μ is the conversion coefficient, which can convert a certain amount of green certificate transactions into offset carbon emissions; l represents the starting point of the current emission ladder; h is the higher ladder position that we hope to avoid entering through offsetting measures.

[0135] 3.4 Quota Balance Constraints

[0136]

[0137] Where: E is the total carbon emissions; E0 is the free carbon allowance; ΔE i This represents the carbon emissions at each stage.

[0138] Step 4: Simulations were performed using the improved IEEE 39-node system and the Belgian 20-node system to verify the effectiveness of the proposed optimization model and method.

[0139] This example uses an improved IEEE-39-node system and a Belgian 20-node system for analysis. The electro-electric coupling system diagram is shown below. Figure 3 As shown, in the improved 39-node power grid, there are 3 gas turbine units, 2 wind turbine units, and the remaining 5 coal-fired units. Gas turbine units G1, G2, and G3 are connected to nodes 30, 33, and 37 of the power grid and nodes 3, 6, and 19 of the natural gas grid, respectively. Wind turbine units W1 and W2 are connected to nodes 35 and 36 of the power grid, respectively. The power grid and natural gas are coupled through gas turbines and P2G (electricity-to-gas) conversion equipment. P2G1 and P2G2 are connected to nodes 10 and 20 of the power grid and nodes 5 and 10 of the natural gas grid, respectively, with maximum power outputs of 500MW and 600MW and a conversion efficiency of 25%. The Belgian 20-node system includes 20 nodes, 19 pipelines, and 6 gas source stations.

[0140] To evaluate the effectiveness of the proposed adaptive optimization method for multidimensional carbon asset collaborative trading of electricity-hydrogen-gas integrated energy systems in terms of carbon emissions and system cost control, this section sets up the following four comparative scenarios to verify the effectiveness of the proposed scheme.

[0141] Option 1: Traditional scheduling (control group);

[0142] Option 2: Consider only tiered carbon trading;

[0143] Option 3: Green Certificate-Carbon Trading Mechanism;

[0144] Option 4: Adopt a game equilibrium-deep learning-based green certificate-carbon-hydrogen joint trading mechanism.

[0145] Table 1 compares the optimization results of different scenarios. The unit carbon quota allocation diagram is shown below. Figure 4 As shown.

[0146] Table 1. Operating costs and carbon emissions for each scenario

[0147]

[0148] The introduction of hydrogen energy modules can further improve carbon emission reduction efficiency. During periods of high carbon emissions, the system will schedule hydrogen storage and conversion modules to convert electricity into hydrogen for storage. During periods of low carbon emissions, hydrogen energy will be used for peak shaving and supplementing grid load, reducing carbon emissions. In this way, the participation of hydrogen energy can not only alleviate power system fluctuations, but also provide flexible emission reduction pathways for high-emission units through combination with green certificates and carbon trading mechanisms.

[0149] In Scenario 4, units W1 and W2 are assigned zero carbon liability, indicating that their operation within the system is clean or low-carbon, thus exempting them from or incentivizing them in carbon emission liability sharing. This strategy encourages the system to use more clean energy units, improving the overall environmental friendliness of the operation. In particular, high-emission units (such as G7 and G8) bear a higher carbon liability, prompting them to reduce output under the pressure of carbon costs, thereby achieving the overall carbon emission reduction target of the system.

[0150] To further illustrate the specific performance of this strategy on the power generation side and the gas grid side, the unit output is as follows: Figure 5 As shown, the output of each unit (G1 to G10) exhibits a clear stratification across different hours: during the day or high-load periods, some coal-fired or gas-fired units assume baseload and peak-shaving functions, while at night or during periods of low carbon emission factors, more renewable or clean energy units have a competitive advantage and achieve higher operational priority. Especially under the green certificate-carbon trading mechanism, high-emission units proactively reduce output or shorten operating time under carbon cost pressures, thereby freeing up more load capacity for clean energy sources such as wind power. The hydrogen energy module, in turn, provides regulation during high-carbon periods, further optimizing the system's low-carbon dispatch strategy.

[0151] By shifting loads that can be moved and interrupting electricity consumption during high-emission periods, the load side further smooths out the peak-valley difference between morning and evening, making the system's scheduling more stable. This process not only ensures a balance between carbon emissions and energy supply, but also improves the overall operating efficiency of the system by reducing carbon emissions and optimizing energy distribution.

[0152] The calculations in this paper were performed on the Matlab R2022b platform, using YALMIP modeling and the CPLEX solver.

[0153] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. An adaptive optimization method for an integrated electricity-hydrogen-gas energy system for multidimensional carbon asset collaborative trading, characterized in that, Includes the following steps: Step 1: Introduce a dynamic emission factor and a multi-dimensional carbon asset collaborative trading mechanism to construct a carbon emission optimization model for an electricity-hydrogen-gas three-energy coupling system; Step 2: Use a game equilibrium-deep learning hybrid model to dynamically allocate carbon emission responsibility; Step 3: With the goal of minimizing energy supply costs, carbon emission costs, and green certificate-carbon trading costs, construct an integrated energy system scheduling model that couples electricity, hydrogen, and gas. Step 4: Perform simulations to verify the effectiveness of the optimization model and method.

2. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 1, characterized in that, Step 1 is specifically... Includes the following processes: By introducing hydrogen energy modules, the synergistic optimization of electricity, natural gas, and hydrogen energy can be achieved. The hydrogen energy modules include: The P2H water electrolysis hydrogen production unit utilizes surplus electricity from the power grid to produce hydrogen. Hydrogen energy storage unit (H2S): Stores excess hydrogen in a storage tank; Hydrogen re-conversion unit (H2P): When electricity demand is high or natural gas is scarce, hydrogen is converted into electricity through fuel cells / hydrogen-blended generator sets or injected into the natural gas pipeline network; The power balance relationship of the hydrogen energy module is as follows: In the formula: P el (t) represents the hydrogen production power from water electrolysis; η el Electrolysis efficiency; H prod (t) represents the hydrogen production rate; H stor (t) represents the hydrogen storage capacity; H cons (t) represents the amount of hydrogen consumed; P fc (t) represents the power generated by the fuel cell; η fc For fuel cell efficiency; Through the above mechanism, the hydrogen energy module can play a role in peak shaving and valley filling between the peak and valley loads of the power system, while providing backup and low-carbon energy for the system. Introducing a dynamic emission factor, which changes dynamically with system load levels and carbon asset prices, the dynamic emission factor is defined as follows: In the formula: DEF(t) is the dynamic emission factor at time t; EF base L is the baseline emission factor, i.e., the standard emission per unit of electricity; L(t) is the total load of the system at time t; L max P represents the system's maximum load. carbon (t) represents the carbon price at time t; The highest carbon price; α, β, γ are weighting coefficients, satisfying α+β+γ=1; Combining dynamic emission factors, the total carbon emissions of the system are expressed as: In the formula: C total (t) represents the total carbon emissions of the system at time t; P i (t) represents the power generation of unit i; η i is the fuel emission coefficient of the unit; N is the total number of units in the system.

3. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 1, characterized in that, Step 2 specifically includes the following process: The game theory model analyzes the behavior of each participant in terms of carbon emissions by defining the payoff and cost functions of the participants. The participants' goal is to minimize their carbon compliance costs while complying with the market's carbon trading rules and constraints. The interaction between the participants is solved using Nash equilibrium; the carbon emissions of the power producer are set as C. i Its carbon emission cost is C cost,i Then the game equilibrium model is expressed as: C cost,i =a i ·C i +b i ·P carbon ·max(0,C i -C quota,i ); C i Carbon emissions of producer i; α i For producer i, the carbon emission factor; β i P is the carbon trading weighting factor; carbon For carbon prices in the carbon trading market; C quota,i Carbon allowances for producers; In the game theory model, the Nash equilibrium point is found by solving the carbon emission behavior of different participants, so as to minimize the carbon emission cost of the overall system. Based on the game equilibrium model, the deep learning model predicts future carbon emission trends by training on carbon emission data, thereby dynamically adjusting the parameters in the game equilibrium. By using deep learning models to predict time-series data on carbon emissions, dynamic allocation of carbon emission responsibility can be achieved. The deep learning model is trained using a multilayer perceptron (MLP) or long short-term memory (LSTM) network. It learns the system's emission patterns by taking historical carbon emission data, load data, and carbon price data as input, and outputs the carbon emissions of each producer. The following steps are used to solve the problem: A deep learning model is used to train the carbon emission time series data to obtain the carbon emission prediction values ​​at each time point. Based on game equilibrium theory, calculate each producer's carbon emission responsibility under different carbon prices and emission levels; The objective function is solved using the particle swarm optimization algorithm to obtain the optimal carbon emission responsibility allocation and trading path.

4. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 1, characterized in that, Step 3 specifically includes the following process: The objective function is to minimize the sum of system energy supply cost, carbon emission cost, and green certificate-carbon-hydrogen trading cost. The objective function is as follows: minF1=f e +f g +f g,c,h +f c ; In the formula: F1 is the economic scheduling cost; f e The total fuel cost of the power plant; f g Cost of natural gas well extraction; f g,c,h Costs of green certificate-carbon-hydrogen trading; f c This refers to the cost of carbon emissions; the formulas for calculating each cost are as follows: In the formula: P gird,t Let a be the power generation capacity of power plant i at time t; i b i and c i Let be the power generation coefficient of unit i; The unit extraction price of natural gas at time t; Let be the output of the natural gas well at time t; I and J are the total number of power plants and gas wells in the system, respectively. Costs of green certificates and carbon trading; C cost,i Cost of carbon emissions; Cost of hydrogen energy modules.

5. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system for multidimensional carbon asset collaborative trading as described in claim 4, characterized in that, The hydrogen energy module involves the electrolysis of water to produce hydrogen, storage, and conversion, and its cost is... It can be represented as: Wherein: H prod (t) represents the hydrogen production power; Unit cost of hydrogen production unit; H stor (t) represents the hydrogen storage capacity; For hydrogen storage costs; H fc (t) represents the output power of the fuel cell; The cost of converting hydrogen energy into electricity.

6. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading, as described in claim 4, is characterized in that... The constraints in step 3 include: power grid flow constraints, natural gas grid constraints, and green certificate-carbon trading mechanism constraints. Among them, the green certificate-carbon trading mechanism constraints include carbon trading segment constraints, green certificate segment constraints, interval constraints, and quota balance constraints.

7. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 6, characterized in that, The power flow constraints adopted are DC power flow constraints as power network constraints, and the constraints include node power balance constraints, phase angle constraints, unit output constraints and transmission line constraints.

8. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 6, characterized in that, In the aforementioned natural gas network constraints, the natural gas network consists of gas wells, pipelines, and gas load; its constraints consider node supply and demand balance constraints, node pressure constraints, gas network pipeline constraints, and gas source constraints, as detailed below: Q out,t +Q P2G,t =Q flow,t +Q need,t +Q load,t ; Q out,t,min ≤Q out,t ≤Q out,t,max ; Q flow,min,t ≤Q flow,i,t ≤Q flow,max,t ; In the formula: Q out,t Q represents the outflow of natural gas. P2G,t The natural gas flow rate converted by the P2G (Power-to-Gas) equipment; Q flow,t Q represents the pipeline natural gas flow rate. need,t For natural gas demand; Q load,t Q represents the amount of natural gas required for the gas load; out,t,min Q represents the minimum natural gas flow rate. out,t,max The maximum natural gas flow rate; sign(Q) flow,i,t () represents the direction of pipeline natural gas flow, with 1 for the positive direction and -1 for the negative direction; GP is the gas pressure coefficient of pipeline natural gas. i,t The initial pipeline pressure; GP j,t Q represents the gas pressure at the end of the pipeline; flow,min,t Q represents the minimum pipeline natural gas flow rate. flow,max,t This represents the maximum flow rate of natural gas in the pipeline.

9. The adaptive optimization method for an integrated electricity-hydrogen-gas energy system oriented towards multidimensional carbon asset collaborative trading as described in claim 6, characterized in that, The specific constraints of the green certificate-carbon trading mechanism include: Carbon trading segmentation constraints: In the formula: z k The binary variable determines the stage in which constraints are applied; ΔE1, 4E2, and ΔE3 represent the excess carbon emissions for stages 1, 2, and 3, respectively; I1, I2, and I3 represent the carbon allowances for stages 1, 2, and 3, respectively; M is a large upper bound; z k ∈{0,1} indicates whether the k-th stage carbon price mechanism is enabled; z1, z2, and z3 are binary variables of stages 1, 2, and 3, respectively, and are either 1 or 0; Green certificate segmented constraints: In the formula: y j It is a binary variable; This represents the maximum number of tradable green certificates in the first phase. The maximum number of tradable green certificates for the second phase; U k The upper bound of the green certificate trading constraint is defined by y1, y2, and y3, which are binary variables for each stage; N1, N2, and N3 represent the number of green certificates that can be traded in each stage. Interval constraints: To ensure that the allowances obtained through trading or offsetting are sufficient to cover a certain range of emission intervals, the following constraint is established: In the formula: μ is the conversion coefficient, which can convert a certain amount of green certificate transactions into offset carbon emissions; l represents the starting point of the current emission ladder; h is the higher ladder position that we hope to avoid entering through offsetting measures. Quota balance constraints: Where: E is the total carbon emissions; E0 is the free carbon allowance; ΔE i This represents the carbon emissions at each stage.

10. An adaptive optimization method for an integrated electricity-hydrogen-gas energy system for multidimensional carbon asset collaborative trading, as described in claim 6, is characterized in that... Step 4 uses an improved IEEE 39-node system and a Belgian 20-node system for simulation examples to verify the advantages of the proposed method in terms of carbon emission and cost optimization compared with traditional methods in an electric-hydrogen-gas tri-energy coupling system.

Images