A method, device, medium and equipment for synergistic optimization of electric carbon resources

Abstract

The application discloses a kind of electric carbon resources'synergic optimization method, device, medium and equipment, belong to electric carbon field, comprising: obtaining multiple microgrid system real-time operation data, the operation data is input in the electric carbon resources synergic optimization model based on the two-layer game of preset;Using the reconstruction decomposition technique based on feasible region recovery to the electric carbon resources synergic optimization model is comprehensively handled, and the optimization problem easy to solve is obtained;By improved minimum constraint set identification technology, identify and delete redundant constraints, realize accelerated solution.This process accurately formulates electric carbon resources synergic optimization scheduling instruction, and then real-time adjusts the electric energy transaction in multiple microgrid system, carbon quota allocation and equipment operating parameter, to ensure that electric carbon resources is efficiently optimized configuration and utilization in multiple microgrid system, the present application effectively solves the problem that prior art cannot accurately and efficiently optimize the electric carbon resources of multiple microgrid.

Description

Technical Field

[0001] This invention relates to the field of electrocarbon, and more particularly to a method, apparatus, medium, and equipment for the synergistic optimization of electrocarbon resources. Background Technology

[0002] With increasing global awareness of climate change and environmental protection, and the introduction of the "dual carbon" goals (i.e., carbon peaking and carbon neutrality), higher demands are being placed on the low-carbon operation of energy systems, especially power systems. Multi-microgrid systems, as small-scale power systems integrating various distributed energy sources, have become a key link in achieving energy transition and low-carbon operation due to their flexibility and high proportion of renewable energy.

[0003] In multi-microgrid systems, electricity trading and carbon emission rights allocation are two interconnected issues. On the one hand, electricity trading needs to consider the balance between electricity supply and demand and cost-effectiveness; on the other hand, carbon emission rights allocation involves environmental benefits and emission reduction responsibilities. Therefore, the problem of synergistic optimization of electricity and carbon resources not only needs to consider economic efficiency but also environmental impact, making it a typical multi-objective, multi-constraint, and dynamic optimization problem.

[0004] Existing optimization methods for multi-microgrid systems typically focus on optimizing a single objective, such as considering only economic or environmental benefits. These methods often overlook the inherent link between electricity generation and carbon emissions, resulting in optimization outcomes that may not simultaneously meet both economic and environmental requirements. Furthermore, as the proportion of renewable energy increases, the operational uncertainties of multi-microgrid systems also increase, further complicating the optimization problem.

[0005] These shortcomings prevent existing technologies from accurately and efficiently optimizing the carbon resources of multi-microgrids. Summary of the Invention

[0006] This invention provides a method, apparatus, medium, and equipment for the collaborative optimization of electric carbon resources, in order to solve the problem that existing technologies cannot accurately and efficiently optimize the electric carbon resources of multi-microgrids.

[0007] In a first aspect, this application provides a method for the synergistic optimization of electrical carbon resources, comprising:

[0008] Acquire real-time operational data of the multi-microgrid system, including price data for electricity carbon trading, renewable energy output data, and load data;

[0009] The reconstruction decomposition technique based on feasible domain recovery is adopted, and combined with the real-time operation data, the preset electric carbon resource collaborative optimization model is processed to obtain the first optimization problem; through the improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system is obtained by solving it.

[0010] Based on the aforementioned collaborative optimization scheduling instructions for electricity carbon resources, the electricity energy trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system are adjusted in real time to optimize the electricity carbon resources of the multi-microgrid system.

[0011] The proposed method for coordinated optimization of electricity carbon resources acquires real-time operational data from multi-microgrid systems, including key information such as electricity carbon trading prices, renewable energy output, and load demand, providing accurate input for the optimization process. Utilizing a pre-defined two-level game model, this method simulates the complex interaction between distribution network operators and microgrid operators, ensuring the model reflects the real-world operating environment. Combining feasible region recovery-based reconfiguration decomposition technology and minimum constraint set identification technology, this method effectively processes real-time data, resolves the complexity caused by integer variables, and accelerates the solution process, resulting in a set of coordinated optimization scheduling instructions for electricity carbon resources. These instructions are used to adjust electricity trading, carbon quota allocation, and equipment operating parameters in the multi-microgrid system in real time, achieving optimal allocation of electricity carbon resources. This method not only improves the economic operating efficiency of multi-microgrid systems by reducing unnecessary electricity waste and carbon emissions but also enhances the system's flexibility and reliability, thereby supporting the achievement of sustainable energy strategies and environmental goals. The implementation of this method may lead to a reduction in energy costs, a reduction in carbon emissions, and an optimization of the overall energy structure. It provides multi-microgrid system operators with an innovative and efficient means of co-optimizing electric carbon resources. This application effectively solves the problem that existing technologies cannot accurately and efficiently optimize the electric carbon resources of multi-microgrids.

[0012] Furthermore, the acquisition of real-time operational data of the multi-microgrid system specifically includes:

[0013] Price data for electricity carbon trading is obtained from a pre-set power grid dispatch system. The price data includes the price of abandoned load, the purchase and sale price of electricity between the distribution network and the external power grid, and the carbon price of the distribution network participating in external carbon market transactions.

[0014] Renewable energy output data is obtained from the monitoring system of each microgrid, including the maximum active power output value that renewable energy units in each microgrid can obtain;

[0015] Load data is obtained from the load monitoring system of each microgrid, and the load data includes the non-transferable load and the transferable load of each microgrid.

[0016] This application provides a precise snapshot of the operational status of microgrid systems by acquiring real-time operational data, including key information such as electricity carbon trading prices, renewable energy output, and load demand. This data encompasses load curtailment prices, electricity purchase and sale prices, and carbon market trading prices, providing comprehensive economic parameters for the optimization model. Simultaneously, renewable energy output and load data provide technical constraints for the model. Combining this information, this method utilizes a two-level game model and reconstruction decomposition techniques to effectively process and optimize the allocation of electricity carbon resources, thereby achieving a dual improvement in economic and environmental benefits.

[0017] Furthermore, the reconstruction decomposition technique based on feasible region recovery, combined with the real-time operating data, is used to process the preset electric carbon resource collaborative optimization model to obtain the first optimization problem; through the improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system is obtained by solving it, specifically:

[0018] The real-time operating data is input into a preset two-layer game-based collaborative optimization model for electric carbon resources. This model determines the decision variables in the upper-level distribution network optimization model based on the real-time operating data, and processes the integer variables in the lower-level microgrid response model using a feasible region recovery-based reconstruction decomposition technique. The electric carbon resource collaborative optimization model is then reconstructed and decomposed into a main problem and sub-problems. The electric carbon resource collaborative optimization model includes an upper-level distribution network optimization model and a lower-level microgrid response model. The first optimization problem is the main problem and sub-problems obtained after applying the feasible region recovery-based reconstruction decomposition technique.

[0019] Based on the improved minimum constraint set identification technology, redundant constraints in the decomposed main problem are identified and deleted to obtain a simplified main problem.

[0020] The simplified main problem and the decomposed subproblems are solved iteratively to update the decision variables in the upper-level distribution network optimization model and the decision variables in the lower-level microgrid response model until the preset convergence condition is met, thereby obtaining the collaborative optimization scheduling instruction for the carbon resources of the multi-microgrid system.

[0021] This application acquires and inputs real-time operational data from a multi-microgrid system into a pre-defined two-layer game-theoretic co-optimization model for electricity-carbon resources. This model integrates upper-layer optimization of the distribution network and lower-layer response of the microgrid. Utilizing a reconfiguration decomposition technique based on feasible region recovery, integer variables in the lower-layer model are effectively handled, simplifying the problem's complexity. Furthermore, a minimum constraint set identification technique further simplifies the model's size by identifying and eliminating redundant constraints. The combination of these two techniques not only improves the efficiency of model solving but also ensures the accuracy of the solution process. The simplified main problem and decomposed subproblems are iteratively solved until convergence conditions are met, thereby obtaining accurate co-optimization results for electricity-carbon resources.

[0022] Furthermore, the reconstruction decomposition technique based on feasible region recovery processes the integer variables in the lower-level microgrid response model, reconstructing and decomposing the co-optimization model of electricity carbon resources into a main problem and sub-problems, specifically as follows:

[0023] Identify all integer variables in the lower-level microgrid response model;

[0024] After replacing all the identified integer variables with a preset set of discrete values, non-negative relaxation variables are introduced into the lower-level microgrid response model, and a penalty term corresponding to the non-negative relaxation variables is added to the objective function of the lower-level microgrid response model to reconstruct the lower-level microgrid response model.

[0025] The reconstructed response model of the lower-level microgrid is as follows:

[0026]

[0027] z′ i,t,j ∈F i,t j = 1, ..., n i,t

[0028] In the formula, y i,t include and g i,t include 0 elements and w i,t This represents a continuous variable vector of microgrid i over time period t, including... and y′ i，t，j With w′ i,t,j Let z′ be a continuous variable representing the replication of the lower-level problem in the j-th substitution constraint; i,t,j Let b3 and d be constant vectors, representing the j-th feasible charge / discharge state of the energy storage unit in microgrid i; i P i Q i R i Si M i and T i Let ξ be a constant parameter vector or matrix of appropriate dimension. i,t,j F represents the non-negative slack variable in the j-th substitution constraint; i,t For the feasible state space; n i,t For F i,t The number of states in the system;

[0029] After combining the reconstructed lower-level microgrid response model with the upper-level distribution network optimization model, the main problem and sub-problems are obtained by further decomposition.

[0030] The main problem and sub-problems obtained from the decomposition are:

[0031]

[0032] stAx t ≤b1

[0033] Bx t +Cy i,t +Dg i,t ≤b2

[0034] P i g i,t +Q i y i,t +R i w i,t +S i z i,t ≤b3

[0035]

[0036] In the formula, To optimize the set of time periods; Ω MG For microgrid collection; x t This represents a vector of continuous variables in the distribution network during time period t, including... V n,t and θ n,t ;z i,t Represents the vector of integer variables of follower i during time period t, including and K i,t For z i,t The number of elements in the middle; c1, c 2,i b1, b2, A, B, C, and D are constant parameter vectors or matrices with a preset dimension; F i,t,k This represents z′ after the k-th iteration. i,t,j The set of values ​​for F i,t,k The element is z′ in the first k iterations. i,t,j All provisional solutions; n i,t,k For Fi,t,k The number of elements in; This is a provisional solution to the higher-level problem.

[0037] This application employs a reconstruction decomposition technique based on feasible region recovery to effectively process integer variables in the response model of the lower-level microgrid in a multi-microgrid system. First, all integer variables in the lower-level model are identified. Then, non-negative relaxation variables are introduced into these variables, and a penalty term is designed for each variable to reconstruct the lower-level model, eliminating the solution difficulties caused by integer variables. Next, the reconstructed lower-level model is combined with the upper-level distribution network optimization model to form a main problem and sub-problems. The main problem focuses on the distribution network's decision-making, while the sub-problems focus on the microgrid's response. Through this reconstruction and decomposition, the model not only simplifies the solution process but also improves the solution efficiency, enabling the co-optimization model of electricity and carbon resources to quickly adapt to real-time data changes and achieve dynamic optimization.

[0038] Furthermore, the improved minimum constraint set identification technique identifies and removes redundant constraints in the decomposed main problem, resulting in a simplified main problem, specifically:

[0039] Identify all constraints of the lower-level model in the decomposed main problem and obtain the index set of all constraints;

[0040] Based on the improved minimum constraint set identification technology, an index is randomly selected from the index set, and a redundancy check is performed on the constraints corresponding to the index. If the check fails, the index is deleted from the index set.

[0041] If the test is passed, the first hyperplane through which the ray passes is identified using a preset auxiliary ray algorithm. The constraint that restricts the feasible region the most in the current direction is obtained, and the index of the constraint that restricts the feasible region the most is deleted from the index set. The index of the constraint that restricts the feasible region the most is a non-redundant constraint index.

[0042] The redundancy check is completed and the minimum constraint set is obtained when the index set becomes an empty set.

[0043] The first hyperplane is:

[0044]

[0045] In the formula, l * The index that imposes the greatest constraint on the feasible region;

[0046] The redundancy check is determined by solving an optimization problem with relaxed constraints, which is:

[0047]

[0048] P i,l g i,t +Q i,l y′ i,t,j +R i,l w′ i,t,j +S i,l z′ i,t,j -T i,l ξ i,t,j ≤b 3,l +1

[0049] In the formula, M is the set of indices for all constraints. i Where l is the number of constraints, and l is the index of the constraint being tested;

[0050] The main problem is updated based on the minimum constraint set, resulting in a simplified main problem.

[0051] This application first identifies and lists all constraints of the lower-level model in the decomposed main problem, constructing an index set of all constraints. Next, using an improved minimum constraint set identification technique, each constraint in the index set is checked for redundancy. Constraints that fail the check are deleted, thus gradually narrowing down the range of constraints considered. For constraints that pass the redundancy check, an auxiliary ray algorithm is further applied to determine the constraint that imposes the greatest restriction on the feasible region in the current direction, i.e., the first hyperplane, and this constraint is retained, while others are deleted from the index set. This process is repeated until the index set is empty, meaning all redundant constraints have been identified and deleted, resulting in a minimum constraint set. Finally, the main problem is updated based on this minimum constraint set, resulting in a simplified main problem that reduces model complexity, computational cost, and speeds up the solution process. This process not only improves the efficiency of the optimization model but also ensures the accuracy and reliability of the solution results, providing a more efficient strategy for managing and optimizing carbon resources in multi-microgrid systems.

[0052] Furthermore, the step of adjusting the electricity trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system in real time according to the electricity carbon resource collaborative optimization scheduling instruction specifically includes:

[0053] The electricity carbon resource collaborative optimization scheduling instructions include electricity trading strategies, carbon quota allocation instructions, and equipment operating parameter adjustment instructions.

[0054] Adjust the active power exchange between the distribution network and the external power grid, and between the distribution network and each microgrid, according to the electricity trading strategy;

[0055] In accordance with the carbon quota allocation directive, adjust the amount of carbon quota transferred from the distribution network to each microgrid and the redistribution ratio of carbon quotas within the microgrid;

[0056] Based on the equipment operating parameter adjustment instructions, the output of traditional units, the charging and discharging power of energy storage devices, the time period allocation of transferable loads, and the output of reactive power compensation devices within the multi-microgrid system are controlled in real time.

[0057] This application first guides effective power trading between distribution networks and microgrids based on optimization instructions generated from real-time data, ensuring optimal power allocation across different networks according to demand and supply capacity. Second, carbon quota allocation instructions help optimize the allocation of carbon emission rights, rationally adjusting the transfer volume and redistribution ratio within the microgrid, thereby reducing the overall carbon footprint. Finally, equipment operation parameter adjustment instructions allow for real-time control of energy equipment within the microgrid, such as traditional generating units, energy storage devices, transferable loads, and reactive power compensation devices, to adapt to constantly changing energy demand and supply conditions. Overall, these optimization measures improve energy efficiency, reduce operating costs, promote the use of renewable energy, support environmental sustainability goals, and achieve a win-win situation for both economic and environmental benefits.

[0058] Secondly, this application provides a synergistic optimization device for electric carbon resources, the synergistic optimization device for electric carbon resources comprising:

[0059] The acquisition module is used to acquire real-time operating data of the multi-microgrid system, including price data of electricity carbon trading, renewable energy output data, and load data.

[0060] The processing module is used to process the preset electric carbon resource collaborative optimization model by using the reconstruction decomposition technology based on feasible domain recovery and combining the real-time running data to obtain the first optimization problem; and to identify and delete redundant constraints in the first optimization problem by using the improved minimum constraint set identification technology, and solve to obtain the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system.

[0061] The optimization module is used to adjust the electricity trading, carbon quota allocation and equipment operating parameters of the multi-microgrid system in real time according to the collaborative optimization scheduling instruction for electricity carbon resources, thereby optimizing the electricity carbon resources of the multi-microgrid system.

[0062] The device described in this application comprises an acquisition module, a processing module, and an optimization module, which together achieve real-time optimization management of a multi-microgrid system. First, the acquisition module collects key real-time operational data, including electricity carbon trading prices, renewable energy output, and load demand, providing data support for system decision-making. Next, the processing module utilizes this data, employing a pre-defined two-layer game-theoretic electricity carbon resource collaborative optimization model, combined with feasible region recovery reconstruction decomposition technology and minimum constraint set identification technology, to perform in-depth analysis and processing of the data, generating electricity carbon resource collaborative optimization scheduling instructions. Finally, the optimization module adjusts the electricity trading, carbon quota allocation, and equipment operating parameters in the multi-microgrid system in real time according to these instructions, ensuring the system operates in the most economical and environmentally friendly manner. Through this modular and automated processing flow, this application not only improves the operating efficiency and economy of the multi-microgrid system but also enhances the system's flexibility and capacity to absorb renewable energy, thereby promoting the green transformation of the energy structure and sustainable environmental development.

[0063] Furthermore, the processing module utilizes a reconstruction decomposition technique based on feasible region recovery, combined with the real-time operating data, to process the preset electric carbon resource collaborative optimization model to obtain a first optimization problem; through an improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction for the multi-microgrid system is obtained by solving the problem, specifically:

[0064] The processing module inputs the real-time operating data into a preset two-layer game-based collaborative optimization model for electric carbon resources. This model determines the decision variables in the upper-level distribution network optimization model based on the real-time operating data, and processes the integer variables in the lower-level microgrid response model using a feasible region recovery-based reconstruction decomposition technique. The electric carbon resource collaborative optimization model is then reconstructed and decomposed into a main problem and sub-problems. The electric carbon resource collaborative optimization model includes an upper-level distribution network optimization model and a lower-level microgrid response model. The first optimization problem is the main problem and sub-problems obtained after applying the feasible region recovery-based reconstruction decomposition technique.

[0065] Based on the improved minimum constraint set identification technology, redundant constraints in the decomposed main problem are identified and deleted to obtain a simplified main problem.

[0066] The simplified main problem and the decomposed subproblems are solved iteratively to update the decision variables in the upper-level distribution network optimization model and the decision variables in the lower-level microgrid response model until the preset convergence condition is met, thereby obtaining the collaborative optimization scheduling instruction for the carbon resources of the multi-microgrid system.

[0067] The processing module of this application integrates real-time operational data into a pre-defined two-layer game-theoretic collaborative optimization model for electricity and carbon resources. This model encompasses an upper-layer distribution network optimization model and a lower-layer microgrid response model to achieve data-driven decision optimization. Utilizing a feasible region recovery-based reconstruction decomposition technique, the module can transform integer variables in the lower-layer model into a more easily processed form, thereby simplifying the computation process. Furthermore, by employing a minimum constraint set identification technique, the module can accurately identify and eliminate redundant constraints in the decomposed main problem, resulting in a more concise and efficient main problem model. This process not only improves the efficiency of model solving but also ensures the accuracy of the solution. By iteratively solving the simplified main problem and the decomposed subproblems until the pre-defined convergence conditions are met, an optimized electricity-carbon collaborative scheduling scheme is finally obtained.

[0068] Thirdly, this application provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform a collaborative optimization method for electric carbon resources as described above. Its beneficial effects are the same as those of the collaborative optimization method for electric carbon resources provided in the first aspect of this application.

[0069] Fourthly, this application provides a terminal device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement any of the collaborative optimization methods for electric carbon resources as described in the first aspect. Attached Figure Description

[0070] Figure 1 : A schematic flowchart of an embodiment of the collaborative optimization method for carbon resources provided in this application;

[0071] Figure 2 : A schematic flowchart of an embodiment of the optimized logic framework provided in this application;

[0072] Figure 3 : A schematic diagram of an embodiment of the synergistic optimization device for electric carbon resources provided in this application. Detailed Implementation

[0073] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0074] Example 1

[0075] Please refer to Figure 1 In order to solve the problem that existing technologies cannot accurately and efficiently optimize the carbon resources of multi-microgrids, this invention provides a collaborative optimization method for carbon resources, including steps S01-S03.

[0076] S01: Obtain real-time operating data of the multi-microgrid system, including price data for electricity carbon trading, renewable energy output data, and load data.

[0077] In a preferred embodiment of this invention, the acquisition of real-time operating data of the multi-microgrid system, including price data for electricity carbon trading, renewable energy output data, and load data, specifically involves:

[0078] In the collaborative optimization of carbon resources in multi-microgrid systems, the acquisition of real-time operational data is a fundamental step, and its accuracy and comprehensiveness directly affect the effectiveness of subsequent optimization decisions. Specifically, the acquisition methods and content of three core data types are as follows:

[0079] Price data for electricity carbon trading is primarily obtained by connecting to a pre-defined power grid dispatch system and regional carbon trading platform. Electricity trading prices include peak-hour, normal-hour, and off-peak electricity purchase and sale prices between the distribution network and the external power grid (e.g., peak-hour purchase price of RMB 1.2 / kWh and off-peak sales price of RMB 0.5 / kWh), time-of-use settlement prices for electricity purchases or sales from the distribution network to microgrids, and prices for load shedding due to insufficient load capacity (e.g., RMB 0.8 / kWh). Carbon trading prices encompass the purchase and sale prices of carbon allowances by the distribution network participating in external carbon market transactions (e.g., RMB 60 / ton CO2), as well as the carbon allowance transfer prices between the distribution network and microgrids based on internal negotiations.

[0080] Renewable energy output data is collected in real time through monitoring systems deployed in each microgrid, including the current active power output of photovoltaic (PV) and wind turbine units (e.g., 200 kW PV output and 150 kW wind power output in a microgrid). Simultaneously, short-term output forecasts for the next hour are also obtained (for forward-looking adjustments to support optimization decisions). This data directly reflects the power supply capacity of distributed energy resources and is a key basis for balancing carbon resources.

[0081] Load data is collected by the load monitoring systems of each microgrid and is subdivided into two categories: non-transferable loads and transferable loads. Non-transferable loads refer to rigid loads whose power supply periods cannot be adjusted (such as the power consumption of medical equipment and critical production equipment; the real-time value of non-transferable load in a residential microgrid is 300 kW). Transferable loads refer to flexible loads whose power consumption periods can be adjusted within a certain time range (such as electric vehicle charging loads and commercial building air conditioning loads; the total transferable load of a commercial microgrid is 500 kW, and it is allowed to be redistributed between 10:00 and 15:00). By distinguishing between these two types of loads, data support can be provided for load transfer strategies in subsequent optimized scheduling.

[0082] The above three types of data are aggregated into the multi-microgrid collaborative optimization system through a standardized interface to form a real-time updated dataset, providing initial input for the collaborative optimization of electric carbon resources based on a two-level game model.

[0083] S02: Using the reconstruction decomposition technology based on feasible domain recovery, combined with the real-time operating data, the preset electric carbon resource collaborative optimization model is processed to obtain the first optimization problem; through the improved minimum constraint set identification technology, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system is obtained by solving it.

[0084] In a preferred embodiment of this invention, the reconstruction decomposition technique based on feasible domain recovery, combined with the real-time operating data, processes the preset electric carbon resource collaborative optimization model to obtain a first optimization problem; through an improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system is obtained by solving the problem, specifically:

[0085] In the collaborative optimization of carbon resources in a multi-microgrid system, the core step in achieving precise scheduling is to process real-time operational data based on a pre-defined two-layer game-theoretic collaborative optimization model for carbon resources, combined with feasible region recovery reconstruction decomposition technology and minimum constraint set identification technology. The specific process is as follows:

[0086] First, the collected real-time operational data (including carbon trading prices, renewable energy output, load data, etc.) is input into a two-layer game model. In this model, the upper-layer distribution network optimization model acts as the "leader," aiming to minimize its own operating costs. Based on real-time price data and load demand, it initially determines decision variables such as the power purchased and sold with the external grid, the carbon trading prices to each microgrid, and the initial carbon quota allocation. The lower-layer microgrid response model acts as the "follower," aiming to minimize its own costs based on the price signals set by the upper layer. It makes decisions regarding the amount of carbon trading with the distribution network, the output of internal traditional units, and the energy storage charging and discharging plan. However, this includes integer variables such as the energy storage charging and discharging status, making direct solution difficult.

[0087] More specifically, the leader problem is as follows:

[0088] As the leader in the game, the distribution network operator's objective function is to minimize the total operating cost of the distribution system, including the cost of traditional generating units, penalties for load shedding, electricity purchase and sale fees, and carbon trading fees, specifically defined as follows:

[0089]

[0090] In the formula, To optimize the set of time periods; A collection of traditional generating units in a power distribution network; For the set of distribution network nodes; Ω MG For microgrid collection; The cost coefficient of traditional generating unit g in the distribution network; The active power output of a traditional generating unit g in the distribution network during time period t; and These represent the load abandonment price and active power reduction at node n during time period t, respectively. and These represent the electricity purchase and sale price and active power exchange between the distribution network and the external power grid during time period t. and These represent the electricity purchase and sale price and active power exchange between the distribution network and microgrid i during time period t, respectively. The carbon price when the distribution network participates in external carbon market trading during time period t; The carbon price for transactions between the distribution network and the microgrid i; and Let i represent the remaining carbon allowances for the distribution network and microgrid i in time period t, respectively, and their calculation methods are as follows:

[0091]

[0092]

[0093] In the formula, and The initial free carbon allowances allocated to the distribution network and microgrid i, respectively; E is the collection of traditional generating units in microgrid i; g The carbon emission intensity of a conventional unit (g); This refers to the active power output of a traditional generator unit g in a microgrid during time period t.

[0094] The decision variables of a leader need to satisfy the following constraints:

[0095] (a) Network flow constraints:

[0096]

[0097] In the formula, Let n be the set of traditional generating units at node n of the distribution network; and These represent the active and reactive loads at node n during time period t; For the reduction of reactive load at node n in time period t; G nm and B nm V represents the real and imaginary parts of the element at position (n, m) in the nodal admittance matrix, respectively; m,t and θ m,t These represent the voltage magnitude and phase angle at node m in the distribution network, respectively. For reactive power exchange between the distribution network and the external power grid; For time period t, the reactive power exchange between the distribution network and microgrid i is taken. Let n be the set of reactive power compensation devices at node n; The reactive power output of the reactive power compensation device s during time period t.

[0098] (b) Constraints of traditional units:

[0099]

[0100] In the formula, and These represent the upper and lower limits of the active power output of traditional generating units g in the power distribution network; and These represent the downhill and uphill rates of a traditional generator unit g in a power distribution network, respectively.

[0101] (c) Electricity carbon price constraint:

[0102]

[0103]

[0104] In the formula, p m and These are the lower and upper limits of the electricity price, respectively; p cm and These represent the lower and upper limits of the carbon price, respectively.

[0105] (d) Node voltage constraints and line transmission capacity constraints:

[0106]

[0107] In the formula, V n and These are the lower and upper limits of the voltage amplitude at node n, respectively; θ n and These are the lower and upper limits of the voltage phase angle at node n, respectively; This represents the upper limit of the transmission power of transmission line mn.

[0108] (e) Switching power constraints:

[0109]

[0110] In the formula, P gd and These represent the lower and upper limits of active power exchange between the distribution network and the external power grid, respectively; Q gd and These represent the lower and upper limits of reactive power exchange between the distribution network and the external power grid, respectively.

[0111] (f) Reactive power compensation and load shedding constraints:

[0112]

[0113] In the formula, and Ω represents the lower and upper limits of the reactive power output of the reactive power compensation device s; S It is a collection of reactive power compensation devices.

[0114] The follower problem specifically refers to:

[0115] Microgrid operator i, acting as a follower, has the objective function of minimizing the microgrid's... The total operating cost includes the cost of generating electricity from traditional generating units, the cost of operating and maintaining energy storage, the cost of purchasing and selling electricity, and the cost of carbon trading, which are defined as follows:

[0116]

[0117] In the formula, The cost coefficient of traditional generating unit g in the distribution network; Let i be the collection of energy storage units in microgrid i; and The cost coefficient for energy storage unit e; The charging / discharging efficiency of energy storage unit e; and Let t be the charging and discharging power of energy storage unit e during time period t.

[0118] The decision variables of followers need to satisfy the following constraints:

[0119] (a) Power balance constraints:

[0120]

[0121] In the formula, and These represent the set of traditional generating units, renewable energy generating units, and energy storage units at node α in microgrid i; Let r be the predicted active power output of renewable energy unit r during time period t; and These represent the non-transferable and transferable loads at node α during time period t; Let i be the set of nodes in the microgrid i.

[0122] (b) Constraints of traditional units:

[0123]

[0124] In the formula, and These represent the lower and upper limits of the active power output of a traditional generator unit g in a microgrid; and These represent the downhill and uphill ramp rates of the traditional generator unit g in the microgrid, respectively.

[0125] (c) Energy storage unit constraints:

[0126]

[0127] In the formula, and These are the upper limits of the charging power and the upper limit of the discharging power of energy storage unit e, respectively; and For the state of charge / discharge of energy storage unit e; The state of charge of energy storage unit e during time period t; and These represent the lower and upper limits of the energy storage capacity of energy storage unit e.

[0128] (d) Node voltage constraints:

[0129]

[0130] In the formula, and These represent the voltage amplitude and phase angle at node α of the microgrid during time period t; and These are the lower and upper limits of the voltage amplitude at node α, respectively. and These are the lower and upper limits of the phase angle of the node α voltage, respectively.

[0131] (e) Switching power constraints:

[0132]

[0133] In the formula, and These represent the lower and upper limits of active power exchange between the distribution network and microgrid i, respectively. and These represent the lower and upper limits of reactive power exchange between the distribution network and the microgrid i, respectively.

[0134] (f) Transferable load constraints:

[0135]

[0136] In the formula, The set of time ranges for transferable loads; Time range The total amount of internally transferable load; This represents the upper limit of the load that can be transferred during time period t.

[0137] To handle integer variables in the lower-level model, a reconstructed decomposition technique for recovering the feasible region is introduced: First, integer variables in the lower-level model are identified (such as the identifiers for energy storage charging state 1 and discharging state 0). By replicating these variables and their corresponding constraints, non-negative relaxation variables and penalty terms are introduced to reconstruct the lower-level model. Relaxation variables are used to recover the feasible region that is reduced due to the discreteness of integer variables, while penalty terms force the relaxation variables to take values ​​close to zero, ensuring that the reconstructed model is equivalent to the original model. Subsequently, the reconstructed overall model is decomposed into a main problem and sub-problems. The main problem includes upper-level decision variables and reconstructed lower-level constraints, while the sub-problems are the response models of each microgrid based on the upper-level provisional strategy.

[0138] More specifically, the reconstruction and decomposition technique based on feasible domain recovery includes:

[0139] The above model is essentially a multi-stage, bi-level nonconvex optimization problem. This is because the leader's objective function contains a large number of nonlinear terms (i.e., the bilinear cost in equation (1)). and And there are integer variables in follower decisions (i.e.) and This problem, stemming from the limitations of traditional algorithms, makes it difficult to solve. To address this issue, this application proposes a reconstruction and decomposition technique based on feasible region recovery to accurately solve the electric carbon coordinated scheduling problem in multi-microgrid systems.

[0140] Based on equations (1)-(26), the collaborative optimization model of electric carbon resources in a multi-microgrid system can be simply described as follows:

[0141]

[0142] stAx t ≤b1 (28)

[0143] Bx t +Cy i,t +Dgi,t ≤b2 (29)

[0144]

[0145]

[0146] In the formula, x t This represents a vector of continuous variables representing the leader during time period t, including... V n,t and θ n,t ;y i,t include and g i,t include (and Multiplication), zero element (and) (multiplication) and (and (multiplication); w i,t This represents a continuous variable vector of follower i during time period t, including... and z i,t Represents the vector of integer variables of follower i during time period t, including and K i,t For z i,t The number of elements in F; i,t (g i,t ) represents a given g i,t The solution set of follower i (i.e., the response of microgrid i to the distribution network decision); c1, c 2,i b1, b2, b3, d i A, B, C, D, P i Q i R i and S i It is a constant parameter vector or matrix with appropriate dimensions.

[0147] (a) Restructuring process:

[0148] First, by copying the original variable y of follower i. i,t w i,t z i,t Reconstruct the upper-level problems (27)-(29) with the original constraints (31), and represent the replicated variables and replicated constraints as y′ respectively. i,t w′ i,t 、z′ i,t First, constraints (36) are applied; second, inequality constraints (37) that are easy to handle are used to replace the original constraints (30) to achieve the initial reconstruction of the original model.

[0149]

[0150] stAx t ≤b1 (34)

[0151] Bx t +Cy i,t +Dg i,t ≤b2 (35)

[0152] P i g i,t +Q i y i,t +R i w i,t +S i z i,t ≤b3 (36)

[0153]

[0154]

[0155] It should be noted that, due to the joint constraints of (36) and (37), the reconstructed models (33)-(39) are equivalent to the original models (27)-(32). Although the reconstructed models are more complex in form than the original models, this representation method is convenient for deriving the non-trivial boundaries of game problems and can effectively reflect the complex coupling relationship between the leader's decision and the follower's decision.

[0156] As can be seen from (38), the lower-level problem of the reconstructed model still contains a binary integer variable z′. i,t Directly solving for z′ presents significant challenges. i,t Let represent the charging and discharging states of each energy storage unit in microgrid i. Since the number of feasible state combinations is finite, a discrete set of values ​​can be used to replace it, thereby eliminating integer variables in the lower-level problem. At this point, (37)-(38) can be reconstructed as:

[0157]

[0158] In the formula, y′ i,t,j With w′ i,t,j Let z′ be a continuous variable representing the replication of the lower-level problem in the j-th substitution constraint; i,t,j is a constant vector, representing the j-th feasible charge / discharge state of the energy storage unit in microgrid i; In the distribution network decision g i,t Under these conditions, the energy storage charge / discharge state z′ i,t The state space; for The number of states in the system.

[0159] It should be noted that in constraints (40)-(42) z′ i,t state space Possibly with g i,t The constraints change with the changes in the upper levels; this means that when the upper-level decisions change, the lower-level constraints need to be redefined, which will significantly increase the computational burden of solving the problem. However, if directly in each... union G is g i,t Reconstructing (37)-(38) under the feasible region) will reduce the feasible region of the original optimization problem by increasing the constraints, resulting in the simplified model being unable to be equivalent to the original model;

[0160] To address this issue, the present invention introduces a nonnegative relaxation variable ξ in (38). i,t and its corresponding penalty item Introducing (37). At this point, (37)-(38) can be transformed into:

[0161]

[0162]

[0163] In the formula, M i T is a constant parameter vector of appropriate dimension, where each component is a sufficiently large positive number; i This is a constant parameter matrix used to add slack variables to constraints;

[0164] Clearly, in the case of the nonnegative slack variable ξ i,t Under the influence of this, the reduced feasible region will be restored. Simultaneously, due to the introduction of the relaxation penalty term... To minimize the objective function, ξ i,t The result will tend to 0, at which point (43)-(44) is equivalent to (37)-(38). Based on this, the integer variable z′ in the lower-level problem is eliminated. i,t (43)-(44) can be equivalently transformed into:

[0165]

[0166] z′ i,t,j ∈F i,t j = 1, ..., n i,t (47)

[0167] In the formula, ξ i,t,j Let n be the non-negative slack variable in the j-th substitution constraint; i,t For F i,t The number of states in the system;

[0168] (b) Decomposition process:

[0169] Although the reconstruction process can eliminate integer variables in the lower-level problem, the number of substitution constraints (45)-(47) will increase exponentially with the increase of microgrids and energy storage units. Existing solvers are unable to solve such a large-scale two-level NLP problem in a finite time. In addition, since constructing equivalent substitution constraints requires knowing all feasible states of energy storage units, model reconstruction is very difficult. To address the above problems, this invention decomposes the reconstruction model into easily solvable main problems (48)-(54) and sub-problems (55)-(56), and obtains the optimal solution iteratively.

[0170]

[0171] stAx t ≤b1 (49)

[0172] Bx t +Cy i,t +Dg i,t ≤b2 (50)

[0173] P i g i,t +Q i y i,t +R i w i,t +S i z i,t ≤b3 (51)

[0174]

[0175]

[0176] In the formula, F i,t,k This represents z′ after the k-th iteration. i,t,j The set of values ​​for z′ is the set of values ​​for z′ in the first k iterations. i,t,j All provisional solutions; n i,t,k For F i,t,k The number of elements in; This is a provisional solution to the higher-level problem;

[0177] To continuously approach the optimal solution, the main problem needs to be continuously updated. The specific iterative steps are as follows:

[0178] ① Initialization: Iteration count k = 1, set of integer variable values The number of elements n in the set i,t,k =0, penalty coefficient is M i The allowable error is ε;

[0179] ②for t=1 to T do:

[0180] ③ Input real-time external information for time period t;

[0181] ④ Solve the main problem (48)-(54) using the KKT method to obtain a provisional solution. The current optimal value V of the lower-level problem 1 ;

[0182] ⑤ The subproblems (55)-(56) corresponding to microgrid i are passed on.

[0183] ⑥ Solve each microgrid subproblem (55)-(56) in parallel to obtain provisional solutions. The current optimal value V of the subproblem 2 ;

[0184] ⑦if|V 2 -V 1 |≤εthen

[0185] ⑧ Stop the iteration, let t = t + 1, and return to step ③;

[0186] ⑨Else:

[0187] ⑩ Let n i,t,k =n i,t,k +1, and with The corresponding (52)-(53) are added to the main problem;

[0188] Let k = k + 1, and return to step ④;

[0189] end for;

[0190] Meanwhile, to reduce the complexity of solving the main problem, a minimum constraint set identification technique is employed: First, all inequality constraints in the main problem are extracted, and a constraint index set is constructed. For each constraint, a redundancy check is performed by solving an optimization problem containing relaxation constraints. If the optimal objective function value of the optimization problem is greater than the original right-hand side of the relaxation constraint, it indicates that the constraint has a real restriction on the feasible region (non-redundant at the current stage); if the optimal objective function value is less than or equal to the original right-hand side of the relaxation constraint, it indicates that the constraint can be deleted (redundant). Then, an auxiliary ray algorithm is used to obtain the constraint that restricts the feasible region most in a specific direction: a ray is emitted from a point in the feasible region along a specific direction, and the first hyperplane crossed by the ray is identified to obtain the constraint that restricts the feasible region most. Through redundancy checks and the auxiliary ray algorithm, a minimum constraint set is finally formed, significantly reducing the size of the main problem.

[0191] More specifically, the minimum constraint set identification technique is as follows:

[0192] In the above reconstruction and decomposition algorithm based on feasible region recovery, the number of lower-level problems contained in the main problem will also increase with the increase of the number of iterations. At this time, if the KKT method is directly used to solve the main problem, a large number of binary integer variables will be introduced, resulting in a huge solution burden. Since the number of introduced integer variables is equal to the number of inequality constraints in (53), if the minimum set of inequality constraints can be accurately identified and redundant constraints can be deleted before solving the main problem, the size of the main problem can be effectively reduced, thereby reducing the number of introduced integer variables and thus accelerating the solution. The specific details are as follows:

[0193] (a) Minimum constraint set identification technique:

[0194] For the lower-level constraints (53) in the main problem, the feasible region constructed from them is denoted as... Then it must exist. Makes (57)-(58) true. That is, the minimum set of constraints corresponding to microgrid i;

[0195]

[0196] In the formula, M represents the set of indices of all inequality constraints corresponding to microgrid i in (53). i represents the number of inequality constraints; l represents the index of the inequality constraints.

[0197] To identify the minimum constraint set The present invention constructs the test model shown in (59)-(61) to determine index l in Is there redundancy in the data?

[0198]

[0199] P i,l g i,t +Q i,l y′ i,t,j +R i,l w′ i,t,j +S i,l z′ i,t,j -T i,l ξ i,t,j ≤b 3,l +1 (61)

[0200] Note that due to the feasible region The above optimization model will always find the optimal solution. Since (61) relaxes the constant term on the right-hand side of the original l-th inequality constraint, the optimal objective function p of the above model will always find the optimal solution if and only if the optimal objective function p of the above model is found to be the optimal solution. * Strictly greater than b 3,l hour, It is non-redundant. Using the redundancy check model given in (59)-(61), by checking all... Perform a traversal to identify the minimum set of constraints.

[0201] (b) Improved minimum constraint set identification technique:

[0202] The above techniques provide a basic framework for identifying the minimum constraint set; however, for each microgrid i, M needs to be solved during the traversal process. i Next M i The LP problem of dimensionality leads to high computational complexity. To address this issue, this invention improves the aforementioned minimum constraint set identification technique by employing an auxiliary ray algorithm and iterative filling, which significantly enhances the identification efficiency of the proposed technique. Specific details are as follows:

[0203] First, given any point s within the feasible region, from Randomly select an untested index l and solve the test model. (in, (To store the set of non-redundant constraints), we obtain the point s corresponding to the optimal solution in the high-dimensional space. * With the optimal value p * If p * >b 3,l ,illustrate It is not yet included in ss * Directional constraints on feasible regions The constraints. However, constraint l may still be redundant because ss * There may be more restrictive constraints l′ in the direction. To address this issue, this application develops an auxiliary ray algorithm, which starts from point s and proceeds along ss... * Radiate a ray in a specific direction and identify the first hyperplane it passes through. Thus, the most restrictive constraint in that direction is obtained. * ,but Note that if l * If ≠ l, then l still exists. In the middle, it needs to be tested again; if the test model Determine the pair of l If it is redundant, then Repeat this process until... at this time That is, the minimum constraint set of microgrid i. The specific process of the improved minimum constraint set identification technique and auxiliary ray algorithm is as follows:

[0204] Improved minimum constraint set identification technique:

[0205] ① Initialization: Non-redundant constraint index set Constraint Index Complete Set Let s be a point within the feasible region, and let Boolean variable be Bool.

[0206] ②while do:

[0207] ③ In the set Randomly select index l;

[0208] ④ Solve the redundancy check model Get s * With p * ;

[0209] ⑤if p * >b 3,l then:

[0210] ⑥Bool = true;

[0211] ⑦ Obtain ss using the auxiliary ray algorithm * The index l with the most restrictive direction * ;

[0212] ⑧Else:

[0213] ⑨Bool = false;

[0214] ⑩End;

[0215] If Bool then:

[0216] index l * Add to the set of storage non-redundant constraints, i.e.

[0217] index l * Remove from the index set of all inequalities, i.e.

[0218] Else:

[0219] Remove index l from the index set of all inequalities, i.e.

[0220] End;

[0221] End;

[0222] The auxiliary ray algorithm is specifically as follows:

[0223] Let s be the point in high-dimensional space corresponding to the optimal solution of the redundancy check model. * Then, take any point s within the feasible region, and start from point s, along ss * A line segment is emitted from the direction. If this line segment does not cross any hyperplane (each constraint corresponds to one hyperplane), the length of the line segment is increased until the number of hyperplanes crossed by the line segment equals 1. At this point, the constraint corresponding to the hyperplane crossed by the line segment is ss. * The most restrictive constraint in direction. If increasing the length of a line segment results in the segment crossing multiple hyperplanes, the segment's increment needs to be rolled back, and the length increased by smaller increments until the segment crosses only one hyperplane. At this point, the constraint corresponding to the hyperplane crossed by the segment is ss. * The most restrictive constraint in direction;

[0224] ① Initialization: Let s be the point in the feasible region, and let s be the point corresponding to the optimal solution in the high-dimensional space. * The set of hyperplanes through which rays pass The ray increment is δ, and the ray direction is...

[0225] ②while do:

[0226] ③ Add an increment to the ray, s = s + δr;

[0227] ④ Add the index of the hyperplane through which the ray passes to the set. Right now:

[0228]

[0229] ⑤if then:

[0230] ⑥ Regression ray increment, s = s - δr;

[0231] ⑦ Shorten the increment step size, δ = δ / 10;

[0232] ⑧ Obtain ss using the auxiliary ray algorithm * The index l with the most restrictive direction * ;

[0233] ⑨End;

[0234] ⑩End;

[0235] Finally, the simplified main problem and sub-problems are solved iteratively based on the decision variables: the main problem outputs a provisional upper-level electricity carbon trading strategy, and the sub-problems return the optimal response for each microgrid accordingly; if the difference between the objective functions of the two satisfies the convergence condition, the final electricity carbon resource collaborative optimization scheduling instruction is obtained, covering the electricity trading plan, carbon quota allocation scheme, and equipment operating parameter adjustment values ​​of the distribution network and microgrids. The entire process, through the synergy of two technologies, ensures the accuracy of the optimization results while significantly improving the solution efficiency, meeting the requirements of real-time scheduling.

[0236] like Figure 2 As shown, the optimization logic of this application is as follows: Utilizing the aforementioned reconstruction and decomposition techniques based on feasible region recovery and the improved minimum constraint set identification technique, a solution algorithm for the collaborative optimization of electric carbon resources in multi-microgrid systems can be constructed. The complete iterative steps are as follows:

[0237] ① Initialization: Iteration count k = 1, set of integer variable values The number of elements n in the set i,t,k =0, penalty coefficient is M i The allowable error is ε;

[0238] ②for t=1 to T do:

[0239] ③ Input real-time external information for time period t;

[0240] ④ Solve the main problem (48)-(54) using the KKT method to obtain a provisional solution. The current optimal value V of the lower-level problem 1 ;

[0241] ⑤ The subproblems (55)-(56) corresponding to microgrid i are passed on.

[0242] ⑥ Solve each microgrid subproblem (55)-(56) in parallel to obtain provisional solutions. The current optimal value V of the subproblem 2 ;

[0243] ⑦if|V 2 -V 1 |≤ε then:

[0244] ⑧ Stop the iteration, let t = t + 1, and return to step ③;

[0245] ⑨else:

[0246] ⑩ Let n i,t,k =n i,t,k +1, Building and The corresponding lower-level model;

[0247] Identify using an improved minimal constraint set identification technique

[0248] Delete and The corresponding redundant constraints in the lower-level model, the index set of redundant constraints is

[0249]

[0250] Add the lower-level model, after removing redundant constraints, to the main problem;

[0251] Let k = k + 1, and return to step ④;

[0252] nd for.

[0253] S03: Based on the aforementioned coordinated optimization scheduling instruction for electricity carbon resources, adjust the electricity energy trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system in real time to optimize the electricity carbon resources of the multi-microgrid system.

[0254] In a preferred embodiment of this invention, the step of adjusting the electricity trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system in real time according to the electricity carbon resource collaborative optimization scheduling instruction to optimize the electricity carbon resources of the multi-microgrid system specifically involves:

[0255] Real-time adjustments to the multi-microgrid system based on the coordinated optimization scheduling instructions for electricity carbon resources are the final execution step in achieving coordinated optimization of electricity carbon resources. This is accomplished through actions in the following three dimensions:

[0256] Regarding power trading adjustments, the dispatch instructions clearly define the power exchange rules between the distribution network and the external grid, and between the distribution network and various microgrids. For example, when the instructions indicate that a certain period is peak grid time (when the purchase price is higher), the distribution network will reduce its purchases from the external grid and increase its purchases from various microgrids. If an industrial park microgrid has a 300kW surplus of renewable energy output at this time, the distribution network will purchase 200kW of power from that microgrid according to the instructions, while temporarily adjusting 100kW of transferable load (such as electric vehicle charging) from residential microgrids to this period to absorb the surplus power. During off-peak hours (when the purchase price is lower), the distribution network will purchase 500kW of power from the external grid according to the instructions and discharge 300kW to commercial microgrids with sufficient energy storage capacity, achieving low-cost power storage. Through this time-sharing and entity-specific power exchange adjustment, both the cost of electricity purchase can be reduced and the renewable energy absorption rate can be improved.

[0257] Regarding carbon quota allocation adjustments, based on the carbon quota allocation instructions in the dispatch directives, the distribution network first allocates initial carbon quotas to each microgrid (e.g., 800 tons / year to high-energy-consuming industrial park microgrids and 500 tons / year to residential microgrids primarily reliant on renewable energy), and then dynamically adjusts them based on real-time carbon emission data. For example, if an industrial park microgrid exceeds its carbon emission limit by 10 tons due to increased output from traditional generating units during a certain period, the directive will trigger the distribution network to transfer 10 tons of quota to it (adjusted from the surplus quotas of residential microgrids), settled at an internal carbon price (e.g., 50 yuan / ton); if a commercial microgrid reduces the output of traditional generating units through energy storage charging and discharging, generating a 20-ton quota surplus, the directive allows it to sell the surplus quotas to the external carbon market to obtain additional revenue. This dynamic adjustment ensures that the total carbon emissions of the system do not exceed the limit, and also incentivizes each microgrid to actively reduce emissions through quota trading.

[0258] Regarding the adjustment of equipment operating parameters, the instructions will directly affect various types of equipment. For traditional generating units, the instructions will limit their output to no more than the ramp rate (e.g., the maximum increase in output per hour for a certain unit should not exceed 100kW) to avoid a surge in carbon emissions. For energy storage devices, the instructions will specify the charging and discharging periods and power (e.g., charging to 80% state of charge at 200kW during off-peak hours and discharging at 150kW during peak hours) to achieve "peak shaving and valley filling". For transferable loads, the instructions will transfer the air conditioning load of commercial buildings from 14:00 (peak hours) to 16:00 (normal hours), with the transfer amount controlled within 50kW, and ensuring that the room temperature fluctuation does not exceed ±2℃. For reactive power compensation devices, the instructions will adjust their output capacity in real time (e.g., increasing the compensation by 20kvar when the voltage at a certain node is low) to ensure the stable operation of the power grid.

[0259] Through the coordinated adjustments in the above three aspects, the multi-microgrid system can achieve both economical and efficient flow of electrical energy and control carbon emissions within the target range, ultimately achieving the goal of "coordinated optimization of electricity and carbon resources". For example, through this adjustment mechanism, a pilot system reduced the average monthly electricity purchase cost by 12% and carbon emissions by 8%, verifying the effectiveness of dispatch instructions.

[0260] In summary, the proposed method for coordinated optimization of electricity and carbon resources provides accurate input for the optimization process by acquiring real-time operational data from multi-microgrid systems, including key information such as electricity and carbon trading prices, renewable energy output, and load demand. Utilizing a pre-defined two-level game model, this method simulates the complex interactions between distribution network operators and microgrid operators, ensuring the model reflects the real-world operating environment. By combining feasible region recovery-based reconfiguration decomposition technology and minimum constraint set identification technology, this method effectively processes real-time data, resolves the complexity caused by integer variables, and accelerates the solution process, resulting in a set of coordinated optimization scheduling instructions for electricity and carbon resources. These instructions are used to adjust electricity trading, carbon quota allocation, and equipment operating parameters in the multi-microgrid system in real time, achieving optimal allocation of electricity and carbon resources. This method not only improves the economic operating efficiency of multi-microgrid systems and reduces unnecessary electricity waste and carbon emissions, but also enhances the system's flexibility and reliability, thereby supporting the achievement of sustainable energy strategies and environmental goals. The implementation of this method may lead to a reduction in energy costs, a reduction in carbon emissions, and an optimization of the overall energy structure. It provides multi-microgrid system operators with an innovative and efficient means of co-optimizing electric carbon resources, effectively solving the problem that existing technologies cannot accurately and efficiently optimize the electric carbon resources of multi-microgrids.

[0261] Example 2

[0262] Please refer to Figure 3 This is a collaborative optimization device for carbon resources provided in the embodiments of this application.

[0263] In this embodiment, the collaborative optimization device for carbon resources includes an acquisition module 10, a processing module 20, and an optimization module 30.

[0264] The acquisition module 10 is used to acquire real-time operating data of the multi-microgrid system, including price data of electricity carbon trading, renewable energy output data and load data.

[0265] In a preferred embodiment of this invention, the acquisition of real-time operating data of the multi-microgrid system, including price data for electricity carbon trading, renewable energy output data, and load data, specifically involves:

[0266] In the collaborative optimization of carbon resources in multi-microgrid systems, the acquisition of real-time operational data is a fundamental step, and its accuracy and comprehensiveness directly affect the effectiveness of subsequent optimization decisions. Specifically, the acquisition methods and content of three core data types are as follows:

[0267] Price data for electricity carbon trading is primarily obtained by connecting to a pre-defined power grid dispatch system and regional carbon trading platform. Electricity trading prices include peak-hour, normal-hour, and off-peak electricity purchase and sale prices between the distribution network and the external power grid (e.g., peak-hour purchase price of RMB 1.2 / kWh and off-peak sales price of RMB 0.5 / kWh), time-of-use settlement prices for electricity purchases or sales from the distribution network to microgrids, and prices for load shedding due to insufficient load capacity (e.g., RMB 0.8 / kWh). Carbon trading prices encompass the purchase and sale prices of carbon allowances by the distribution network participating in external carbon market transactions (e.g., RMB 60 / ton CO2), as well as the carbon allowance transfer prices between the distribution network and microgrids based on internal negotiations.

[0268] Renewable energy output data is collected in real time through monitoring systems deployed in each microgrid, including the current active power output of photovoltaic (PV) and wind turbine units (e.g., 200 kW PV output and 150 kW wind power output in a microgrid). Simultaneously, short-term output forecasts for the next hour are also obtained (for forward-looking adjustments to support optimization decisions). This data directly reflects the power supply capacity of distributed energy resources and is a key basis for balancing carbon resources.

[0269] Load data is collected by the load monitoring systems of each microgrid and is subdivided into two categories: non-transferable loads and transferable loads. Non-transferable loads refer to rigid loads whose power supply periods cannot be adjusted (such as the power consumption of medical equipment and critical production equipment; the real-time value of non-transferable load in a residential microgrid is 300 kW). Transferable loads refer to flexible loads whose power consumption periods can be adjusted within a certain time range (such as electric vehicle charging loads and commercial building air conditioning loads; the total transferable load of a commercial microgrid is 500 kW, and it is allowed to be redistributed between 10:00 and 15:00). By distinguishing between these two types of loads, data support can be provided for load transfer strategies in subsequent optimized scheduling.

[0270] The above three types of data are aggregated into the multi-microgrid collaborative optimization system through a standardized interface to form a real-time updated dataset, providing initial input for the collaborative optimization of electric carbon resources based on a two-level game model.

[0271] The processing module 20 is used to process the preset electric carbon resource collaborative optimization model by using the reconstruction decomposition technology based on feasible domain recovery and combining the real-time running data to obtain the first optimization problem; by using the improved minimum constraint set identification technology, redundant constraints in the first optimization problem are identified and deleted, and the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system is obtained.

[0272] In a preferred embodiment of this invention, the pre-defined collaborative optimization model for electric carbon resources is processed using a reconstruction decomposition technique based on feasible region recovery, combined with the real-time operating data, to obtain a first optimization problem. Then, through an improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and removed, and the collaborative optimization scheduling instruction for electric carbon resources of the multi-microgrid system is obtained. Specifically:

[0273] In the collaborative optimization of carbon resources in a multi-microgrid system, the core step in achieving precise scheduling is to process real-time operational data based on a pre-defined two-layer game-theoretic collaborative optimization model for carbon resources, combined with feasible region recovery reconstruction decomposition technology and minimum constraint set identification technology. The specific process is as follows:

[0274] First, the collected real-time operational data (including carbon trading prices, renewable energy output, load data, etc.) is input into a two-layer game model. In this model, the upper-layer distribution network optimization model acts as the "leader," aiming to minimize its own operating costs. Based on real-time price data and load demand, it initially determines decision variables such as the power purchased and sold with the external grid, the carbon trading prices to each microgrid, and the initial carbon quota allocation. The lower-layer microgrid response model acts as the "follower," aiming to minimize its own costs based on the price signals set by the upper layer. It makes decisions regarding the amount of carbon trading with the distribution network, the output of internal traditional units, and the energy storage charging and discharging plan. However, this includes integer variables such as the energy storage charging and discharging status, making direct solution difficult.

[0275] To handle integer variables in the lower-level model, a reconstructed decomposition technique for recovering the feasible region is introduced: First, integer variables in the lower-level model are identified (such as identifiers for energy storage charging state 1 and discharging state 0). By replicating these variables and their corresponding constraints, non-negative relaxation variables and penalty terms are introduced to reconstruct the lower-level model. Relaxation variables are used to recover the feasible region that is reduced due to the discreteness of integer variables, while penalty terms force the relaxation variables to take values ​​close to zero, ensuring that the reconstructed model is equivalent to the original model. Subsequently, the reconstructed overall model is decomposed into a main problem and sub-problems. The main problem includes upper-level decision variables and reconstructed lower-level constraints, while the sub-problems are the response models of each microgrid based on the upper-level provisional strategy.

[0276] Meanwhile, to reduce the complexity of solving the main problem, a minimum constraint set identification technique is employed: First, all inequality constraints in the main problem are extracted, and a constraint index set is constructed. For each constraint, a redundancy check is performed by solving an optimization problem containing relaxation constraints. If the optimal objective function value of the optimization problem is greater than the original right-hand side of the relaxation constraint, it indicates that the constraint has a real restriction on the feasible region (non-redundant at the current stage); if the optimal objective function value is less than or equal to the original right-hand side of the relaxation constraint, it indicates that the constraint can be deleted (redundant). Then, an auxiliary ray algorithm is used to obtain the constraint that restricts the feasible region most in a specific direction: a ray is emitted from a point in the feasible region along a specific direction, and the first hyperplane crossed by the ray is identified to obtain the constraint that restricts the feasible region most. Through redundancy checks and the auxiliary ray algorithm, a minimum constraint set is finally formed, significantly reducing the size of the main problem.

[0277] Finally, the simplified main problem and sub-problems are solved iteratively based on the decision variables: the main problem outputs a provisional upper-level electricity carbon trading strategy, and the sub-problems return the optimal response for each microgrid accordingly; if the difference between the objective functions of the two satisfies the convergence condition, the final electricity carbon resource collaborative optimization scheduling instruction is obtained, covering the electricity trading plan, carbon quota allocation scheme, and equipment operating parameter adjustment values ​​of the distribution network and microgrids. The entire process, through the synergy of two technologies, ensures the accuracy of the optimization results while significantly improving the solution efficiency, meeting the requirements of real-time scheduling.

[0278] like Figure 2 As shown, the optimization logic of this application is as follows: Utilizing the aforementioned reconstruction and decomposition techniques based on feasible region recovery and the improved minimum constraint set identification technique, a solution algorithm for the collaborative optimization of electric carbon resources in multi-microgrid systems can be constructed. The complete iterative steps are as follows:

[0279] ① Initialization: Iteration count k = 1, set of integer variable values The number of elements n in the set i,t,k =0, penalty coefficient is M i The allowable error is ε;

[0280] ②for t=1 to T do;

[0281] ③ Input real-time external information for time period t;

[0282] ④ Solve the main problem (48)-(54) using the KKT method to obtain a provisional solution. The current optimal value V of the lower-level problem 1 ;

[0283] ⑤ The subproblems (55)-(56) corresponding to microgrid i are passed on.

[0284] ⑥ Solve each microgrid subproblem (55)-(56) in parallel to obtain provisional solutions. The current optimal value V of the subproblem 2 ;

[0285] ⑦if|V 2 -V 1 |≤ε then;

[0286] ⑧ Stop the iteration, let t = t + 1, and return to step ③;

[0287] ⑨else:

[0288] ⑩ Let n i,t,k =n i,t,k +1, Building and The corresponding lower-level model;

[0289] Identify using an improved minimal constraint set identification technique

[0290] Delete and The corresponding redundant constraints in the lower-level model, the index set of redundant constraints is

[0291]

[0292] Add the lower-level model, after removing redundant constraints, to the main problem;

[0293] Let k = k + 1, and return to step ④;

[0294] nd for.

[0295] The optimization module 30 is used to adjust the electricity trading, carbon quota allocation and equipment operation parameters of the multi-microgrid system in real time according to the electricity carbon resource collaborative optimization scheduling instruction, so as to optimize the electricity carbon resources of the multi-microgrid system.

[0296] In a preferred embodiment of this invention, the step of adjusting the electricity trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system in real time according to the electricity carbon resource collaborative optimization scheduling instruction to optimize the electricity carbon resources of the multi-microgrid system specifically involves:

[0297] Real-time adjustments to the multi-microgrid system based on the coordinated optimization scheduling instructions for electricity carbon resources are the final execution step in achieving coordinated optimization of electricity carbon resources. This is accomplished through actions in the following three dimensions:

[0298] Regarding power trading adjustments, the dispatch instructions clearly define the power exchange rules between the distribution network and the external grid, and between the distribution network and various microgrids. For example, when the instructions indicate that a certain period is peak grid time (when the purchase price is higher), the distribution network will reduce its purchases from the external grid and increase its purchases from various microgrids. If an industrial park microgrid has a 300kW surplus of renewable energy output at this time, the distribution network will purchase 200kW of power from that microgrid according to the instructions, while temporarily adjusting 100kW of transferable load (such as electric vehicle charging) from residential microgrids to this period to absorb the surplus power. During off-peak hours (when the purchase price is lower), the distribution network will purchase 500kW of power from the external grid according to the instructions and discharge 300kW to commercial microgrids with sufficient energy storage capacity, achieving low-cost power storage. Through this time-sharing and entity-specific power exchange adjustment, both the cost of electricity purchase can be reduced and the renewable energy absorption rate can be improved.

[0299] Regarding carbon quota allocation adjustments, based on the carbon quota allocation instructions in the dispatch directives, the distribution network first allocates initial carbon quotas to each microgrid (e.g., 800 tons / year to high-energy-consuming industrial park microgrids and 500 tons / year to residential microgrids primarily reliant on renewable energy), and then dynamically adjusts them based on real-time carbon emission data. For example, if an industrial park microgrid exceeds its carbon emission limit by 10 tons due to increased output from traditional generating units during a certain period, the directive will trigger the distribution network to transfer 10 tons of quota to it (adjusted from the surplus quotas of residential microgrids), settled at an internal carbon price (e.g., 50 yuan / ton); if a commercial microgrid reduces the output of traditional generating units through energy storage charging and discharging, generating a 20-ton quota surplus, the directive allows it to sell the surplus quotas to the external carbon market to obtain additional revenue. This dynamic adjustment ensures that the total carbon emissions of the system do not exceed the limit, and also incentivizes each microgrid to actively reduce emissions through quota trading.

[0300] Regarding the adjustment of equipment operating parameters, the instructions will directly affect various types of equipment. For traditional generating units, the instructions will limit their output to no more than the ramp rate (e.g., the maximum increase in output per hour for a certain unit should not exceed 100kW) to avoid a surge in carbon emissions. For energy storage devices, the instructions will specify the charging and discharging periods and power (e.g., charging to 80% state of charge at 200kW during off-peak hours and discharging at 150kW during peak hours) to achieve "peak shaving and valley filling". For transferable loads, the instructions will transfer the air conditioning load of commercial buildings from 14:00 (peak hours) to 16:00 (normal hours), with the transfer amount controlled within 50kW, and ensuring that the room temperature fluctuation does not exceed ±2℃. For reactive power compensation devices, the instructions will adjust their output capacity in real time (e.g., increasing the compensation by 20kvar when the voltage at a certain node is low) to ensure the stable operation of the power grid.

[0301] Through the coordinated adjustments in the above three aspects, the multi-microgrid system can achieve both economical and efficient flow of electrical energy and control carbon emissions within the target range, ultimately achieving the goal of "coordinated optimization of electricity and carbon resources". For example, through this adjustment mechanism, a pilot system reduced the average monthly electricity purchase cost by 12% and carbon emissions by 8%, verifying the effectiveness of dispatch instructions.

[0302] In summary, the device of this application consists of an acquisition module, a processing module, and an optimization module, which together realize real-time optimization management of a multi-microgrid system. First, the acquisition module is responsible for collecting key real-time operational data, including electricity carbon trading prices, renewable energy output, and load demand, providing data support for system decision-making. Next, the processing module utilizes this data, employing a pre-defined two-layer game-theoretic electricity carbon resource collaborative optimization model, combined with feasible region recovery reconstruction decomposition technology and improved minimum constraint set identification technology, to perform in-depth analysis and processing of the data, generating electricity carbon resource collaborative optimization scheduling instructions. Finally, the optimization module adjusts the electricity trading, carbon quota allocation, and equipment operating parameters in the multi-microgrid system in real time according to these instructions, ensuring the system operates in the most economical and environmentally friendly manner. The reasoning process shows that through this modular and automated processing flow, this application not only improves the operating efficiency and economy of the multi-microgrid system but also enhances the system's flexibility and capacity to absorb renewable energy, thereby promoting the green transformation of the energy structure and sustainable environmental development.

[0303] Example 3

[0304] This application provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the aforementioned collaborative optimization method for electric carbon resources.

[0305] The method for synergistic optimization of carbon resources, if implemented as a software functional unit and used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0306] Example 4

[0307] This embodiment provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements any one of the collaborative optimization methods for electric carbon resources as described in Embodiment 1.

[0308] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for synergistic optimization of electrical carbon resources, characterized in that, include: Acquire real-time operational data of the multi-microgrid system, including price data for electricity carbon trading, renewable energy output data, and load data; A reconstruction decomposition technique based on feasible region recovery is employed, combined with the real-time operational data, to process the preset collaborative optimization model for electric carbon resources, resulting in a first optimization problem. Through an improved minimum constraint set identification technique, redundant constraints in the first optimization problem are identified and removed, and the collaborative optimization scheduling instructions for electric carbon resources of the multi-microgrid system are obtained by solving the problem. Specifically: The real-time operating data is input into a preset two-layer game-based collaborative optimization model for electric carbon resources. This model determines the decision variables in the upper-level distribution network optimization model based on the real-time operating data, and processes the integer variables in the lower-level microgrid response model using a feasible region recovery-based reconstruction decomposition technique. The electric carbon resource collaborative optimization model is then reconstructed and decomposed into a main problem and sub-problems. The electric carbon resource collaborative optimization model includes an upper-level distribution network optimization model and a lower-level microgrid response model. The first optimization problem is the main problem and sub-problems obtained after applying the feasible region recovery-based reconstruction decomposition technique. Based on the improved minimum constraint set identification technology, redundant constraints in the decomposed main problem are identified and deleted to obtain a simplified main problem. The simplified main problem and the decomposed subproblems are solved iteratively to update the decision variables in the upper-level distribution network optimization model and the decision variables in the lower-level microgrid response model until the preset convergence condition is met, and the collaborative optimization scheduling instruction of the electric carbon resources of the multi-microgrid system is obtained. The reconstruction decomposition technique based on feasible domain recovery processes the integer variables in the lower-level microgrid response model, reconstructing and decomposing the co-optimization model of electricity carbon resources into a main problem and sub-problems, specifically as follows: Identify all integer variables in the lower-level microgrid response model; After replacing all the identified integer variables with a preset set of discrete values, non-negative relaxation variables are introduced into the lower-level microgrid response model, and a penalty term corresponding to the non-negative relaxation variables is added to the objective function of the lower-level microgrid response model to reconstruct the lower-level microgrid response model. The reconstructed response model of the lower-level microgrid is as follows: In the formula, include , and ; include 0 elements and ; microgrid i exist t A vector of continuous variables over a period of time, including , , , , , and ; and For the first j The replication of continuous variables in the lower-level problem within the alternative constraints; A constant vector representing a microgrid i The first energy storage unit j Group feasible charge / discharge states; , , , , , , and For a constant parameter vector or matrix of appropriate dimension, For the first j Non-negative relaxation variables in alternative constraints; This is the feasible state space; for The number of states in the system; After combining the reconstructed lower-level microgrid response model with the upper-level distribution network optimization model, the main problem and sub-problems are obtained by further decomposition. The main problem and sub-problems obtained from the decomposition are: In the formula, To optimize the set of time periods; For microgrid collection; Indicates the distribution network in t A vector of continuous variables over a period of time, including , , , , , , , , and ; followers i exist t A vector of integer variables for a given time period, including and ; for The number of elements in the middle; , , , A, B, C, and D are constant parameter vectors or matrices with a preset dimension; Indicates the first k After the second iteration The set of possible values, element is before k In the next iteration All provisional solutions; for The number of elements in; This is a provisional solution to the higher-level problem; The improved minimum constraint set identification technique identifies and removes redundant constraints in the decomposed main problem, resulting in a simplified main problem, specifically: Identify all constraints of the lower-level model in the decomposed main problem and obtain the index set of all constraints; Based on the improved minimum constraint set identification technology, an index is randomly selected from the index set, and a redundancy check is performed on the constraints corresponding to the index. If the check fails, the index is deleted from the index set. If the test is passed, the first hyperplane through which the ray passes is identified using a preset auxiliary ray algorithm. The constraint that restricts the feasible region the most in the current direction is obtained, and the index of the constraint that restricts the feasible region the most is deleted from the index set. The index of the constraint that restricts the feasible region the most is a non-redundant constraint index. The redundancy check is completed and the minimum constraint set is obtained when the index set becomes an empty set. The first hyperplane is: In the formula, The index that imposes the greatest constraint on the feasible region; The redundancy check is determined by solving an optimization problem with relaxed constraints, which is: In the formula, For the set of indexes of all constraints, For the number of constraints, l The index for the constraint being tested; The main problem is updated based on the minimum constraint set to obtain the simplified main problem; Based on the aforementioned collaborative optimization scheduling instructions for electricity carbon resources, the electricity energy trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system are adjusted in real time to optimize the electricity carbon resources of the multi-microgrid system.

2. The method for synergistic optimization of electrical carbon resources according to claim 1, characterized in that, The acquisition of real-time operational data of the multi-microgrid system specifically includes: Price data for electricity carbon trading is obtained from a pre-set power grid dispatch system. The price data includes the price of abandoned load, the purchase and sale price of electricity between the distribution network and the external power grid, and the carbon price of the distribution network participating in external carbon market transactions. Renewable energy output data is obtained from the monitoring system of each microgrid, including the maximum active power output value that renewable energy units in each microgrid can obtain; Load data is obtained from the load monitoring system of each microgrid, and the load data includes the non-transferable load and the transferable load of each microgrid.

3. The method for synergistic optimization of carbon resources according to claim 1, characterized in that, The step of adjusting the electricity trading, carbon quota allocation, and equipment operating parameters of the multi-microgrid system in real time according to the electricity carbon resource collaborative optimization scheduling instruction is as follows: The electricity carbon resource collaborative optimization scheduling instructions include electricity trading strategies, carbon quota allocation instructions, and equipment operating parameter adjustment instructions. Adjust the active power exchange between the distribution network and the external power grid, and between the distribution network and each microgrid, according to the electricity trading strategy; Adjust the carbon quota trading volume between the distribution network and the external carbon market, and between the distribution network and each microgrid, in accordance with the carbon quota allocation directive; Based on the equipment operating parameter adjustment instructions, the output of traditional units, the charging and discharging power of energy storage devices, the time period allocation of transferable loads, and the output of reactive power compensation devices within the multi-microgrid system are controlled in real time.

4. A synergistic optimization device for electrical carbon resources, characterized in that, include: The acquisition module is used to acquire real-time operating data of the multi-microgrid system, including price data of electricity carbon trading, renewable energy output data, and load data. The processing module is used to process the preset electric carbon resource collaborative optimization model using a reconstruction decomposition technique based on feasible region recovery, combined with the real-time operating data, to obtain a first optimization problem; and to identify and delete redundant constraints in the first optimization problem using an improved minimum constraint set identification technique, thereby solving for the electric carbon resource collaborative optimization scheduling instruction of the multi-microgrid system, specifically: The processing module inputs the real-time operating data into a preset two-layer game-based collaborative optimization model for electric carbon resources. This model determines the decision variables in the upper-level distribution network optimization model based on the real-time operating data, and processes the integer variables in the lower-level microgrid response model using a feasible region recovery-based reconstruction decomposition technique. The electric carbon resource collaborative optimization model is then reconstructed and decomposed into a main problem and sub-problems. The electric carbon resource collaborative optimization model includes an upper-level distribution network optimization model and a lower-level microgrid response model. The first optimization problem is the main problem and sub-problems obtained after applying the feasible region recovery-based reconstruction decomposition technique. Based on the improved minimum constraint set identification technology, redundant constraints in the decomposed main problem are identified and deleted to obtain a simplified main problem. The simplified main problem and the decomposed subproblems are solved iteratively to update the decision variables in the upper-level distribution network optimization model and the decision variables in the lower-level microgrid response model until the preset convergence condition is met, and the collaborative optimization scheduling instruction of the electric carbon resources of the multi-microgrid system is obtained. The reconstruction decomposition technique based on feasible domain recovery processes the integer variables in the lower-level microgrid response model, reconstructing and decomposing the co-optimization model of electricity carbon resources into a main problem and sub-problems, specifically as follows: Identify all integer variables in the lower-level microgrid response model; After replacing all the identified integer variables with a preset set of discrete values, non-negative relaxation variables are introduced into the lower-level microgrid response model, and a penalty term corresponding to the non-negative relaxation variables is added to the objective function of the lower-level microgrid response model to reconstruct the lower-level microgrid response model. The reconstructed response model of the lower-level microgrid is as follows: In the formula, include , and ; include 0 elements and ; microgrid i exist t A vector of continuous variables over a period of time, including , , , , , and ; and For the first j The replication of continuous variables in the lower-level problem within the alternative constraints; A constant vector representing a microgrid i The first energy storage unit j Group feasible charge / discharge states; , , , , , , and For a constant parameter vector or matrix of appropriate dimension, For the first j Non-negative relaxation variables in alternative constraints; This is the feasible state space; for The number of states in the system; After combining the reconstructed lower-level microgrid response model with the upper-level distribution network optimization model, the main problem and sub-problems are obtained by further decomposition. The main problem and sub-problems obtained from the decomposition are: In the formula, To optimize the set of time periods; For microgrid collection; Indicates the distribution network in t A vector of continuous variables over a period of time, including , , , , , , , , and ; followers i exist t A vector of integer variables for a given time period, including and ; for The number of elements in the middle; , , , A, B, C, and D are constant parameter vectors or matrices with a preset dimension; Indicates the first k After the second iteration The set of possible values, element is before k In the next iteration All provisional solutions; for The number of elements in; This is a provisional solution to the higher-level problem; The improved minimum constraint set identification technique identifies and removes redundant constraints in the decomposed main problem, resulting in a simplified main problem, specifically: Identify all constraints of the lower-level model in the decomposed main problem and obtain the index set of all constraints; Based on the improved minimum constraint set identification technology, an index is randomly selected from the index set, and a redundancy check is performed on the constraints corresponding to the index. If the check fails, the index is deleted from the index set. If the test is passed, the first hyperplane through which the ray passes is identified using a preset auxiliary ray algorithm. The constraint that restricts the feasible region the most in the current direction is obtained, and the index of the constraint that restricts the feasible region the most is deleted from the index set. The index of the constraint that restricts the feasible region the most is a non-redundant constraint index. The redundancy check is completed and the minimum constraint set is obtained when the index set becomes an empty set. The first hyperplane is: In the formula, The index that imposes the greatest constraint on the feasible region; The redundancy check is determined by solving an optimization problem with relaxed constraints, which is: In the formula, For the set of indexes of all constraints, For the number of constraints, l The index for the constraint being tested; The main problem is updated based on the minimum constraint set to obtain the simplified main problem; The optimization module is used to adjust the electricity trading, carbon quota allocation and equipment operating parameters of the multi-microgrid system in real time according to the collaborative optimization scheduling instruction for electricity carbon resources, thereby optimizing the electricity carbon resources of the multi-microgrid system.

5. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the collaborative optimization method for electric carbon resources as described in any one of claims 1 to 3.

6. A terminal device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the collaborative optimization method for electric carbon resources as described in any one of claims 1 to 3.

Images