Multi-microgrid cooperative scheduling method based on master-slave cooperative game and target cascade analysis

Abstract

The invention belongs to the field of electric power system scheduling, and relates to a multi-microgrid cooperative scheduling method based on a master-slave cooperative game and target cascade analysis, which utilizes the cooperative scheduling problem of electric power resources and carbon emission of a multi-microgrid system to construct a double-layer master-slave cooperative game framework. Power scheduling and operation strategies of the micro-grid alliance are planned as a whole; the micro-grids of the lower-layer micro-grid alliance realize power transaction and resource sharing through a cooperative game; based on a target cascade analysis algorithm, introducing an Anderson acceleration method to dynamically adjust tie line power, and combining an augmented Lagrangian penalty function to process constraints, optimizing an upper and lower layer decoupling iteration process and accelerating convergence; the problem of low efficiency of multi-agent collaborative optimization is effectively solved, stable system power and reasonable distribution of electric energy resources are ensured, and technical support is provided for operation of a multi-microgrid system.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system dispatching, and in particular relates to a multi-microgrid coordinated dispatching method based on master-slave cooperative game and target cascade analysis. Background Art

[0002] The global energy structure is accelerating its transition to low-carbonization, and carbon reduction policies in various countries are driving the power system to shift to renewable energy. Microgrid technology can integrate clean energy such as photovoltaics and wind power with its flexible and controllable characteristics, but it is difficult to achieve global resource optimization due to the limited equipment capacity of a single microgrid. Multi-microgrid systems have become a key path to improve energy efficiency and reduce carbon emissions through coordinated scheduling and resource sharing.

[0003] Driven by low-carbon policies, multiple microgrids interact deeply with grid operators to form a regional energy coordination network. Through power complementarity and peer-to-peer (P2P) transactions between microgrids, electricity and carbon quotas are flexibly allocated, which not only ensures stable energy supply but also improves environmental benefits. For example, electricity purchase support is provided during peak hours, and electricity is fed back to the grid or neighboring microgrids when there is excess renewable energy, reducing dependence on high-carbon units. This model significantly improves the clean energy consumption rate and provides new possibilities for building a low-carbon and efficient power system.

[0004] In the collaborative optimization of multiple microgrids and grid operators, reasonable allocation of resources and optimization efficiency are the key to achieving economic and environmental benefits. However, there are differences in supply and demand targets and cost allocation between microgrids and grid operators, which leads to challenges in resource sharing. In addition, the complexity of multi-agent game relationships and multi-dimensional constraints makes it difficult for traditional centralized optimization methods (such as those relying on KKT conditions or heuristic algorithms) to cope with large-scale dynamic scenarios. Existing methods convert the two-layer model into a single-layer planning or combine it with the state transfer algorithm for solution. Although it is effective in specific scenarios, it still has shortcomings such as low computational efficiency and slow convergence speed when facing real-time scheduling needs, and cannot meet the system's rapid response requirements. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a multi-microgrid collaborative scheduling method based on master-slave cooperative game and target cascade analysis.

[0006] In order to achieve the purpose of the present invention, the present invention is implemented by adopting the following technical solutions.

[0007] A multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis includes the following steps: S1. According to the power flow operation parameters of the multi-microgrid system, by constructing the microgrid alliance model, the grid operator model and the carbon emission management model, combined with the power balance constraints, exchange power and voltage constraints of the grid operator, the power balance constraints, power exchange constraints, energy storage equipment charging and discharging constraints, and load response constraints in the microgrid alliance, a two-layer master-slave cooperation game framework between the grid operator and the microgrid alliance is established; The grid operator, as the leader of the master-slave cooperative game framework, coordinates the power resources and interactive power between microgrids by adjusting the tie line power between the grid operator and each microgrid of the microgrid alliance under the condition of satisfying power balance constraints, exchange power and voltage constraints; As a collaborator in the master-slave cooperative game framework, the microgrid alliance allocates power resources between microgrids by adjusting the interactive power between microgrids while satisfying power balance constraints, power exchange constraints, energy storage device charging and discharging constraints, and load response constraints. At the same time, the Nash bargaining game theory is used to allocate the benefits of each microgrid to encourage each microgrid to participate in collaborative scheduling; Carbon emission management evaluates and controls the overall carbon emission level of the microgrid alliance during the coordinated dispatch process. By dynamically adjusting the carbon quota allocation of each microgrid, optimizing the unit operation mode and power trading strategy, the total carbon emission of the system is reduced, the low-carbon operation of the system is promoted, and ultimately the coordinated optimization of power resources and carbon emission targets is achieved; S2. Based on the idea of ​​target cascade analysis algorithm, the two-layer master-slave cooperative game problem between power grid operators and microgrid alliances is decoupled into an upper-layer power grid operator optimization problem and a lower-layer microgrid alliance problem that can be coupled through the tie line power variable. The Anderson acceleration method is introduced to dynamically adjust the tie line power variable, and the augmented Lagrangian penalty function is combined to handle constraints to achieve upper and lower decoupled iterative optimization. The convergence condition for iterative termination is set through the convergence criterion to accelerate convergence. S3. Based on the master-slave cooperative game framework built in step S1 and the target cascade analysis algorithm designed in step S2, the coordinated optimization of the power resources and carbon emissions of the microgrid alliance is achieved while satisfying the voltage stability, power flow constraints and power balance constraints. A scheduling plan that takes into account the fairness of all entities is output to adjust the interactive power within the microgrid alliance to ensure the low-carbon and stable operation of the microgrid alliance.

[0008] As a preferred solution of the present invention, the objective function of the power grid operator is: , is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the output power of the ith generator, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively.

[0009] As a preferred solution of the present invention, the constraints of the power grid operator include: (1) Power balance constraint: In each period, the power exchange between the grid operator, generators, loads and microgrids is balanced: , in, Represents the power exchange between the grid operator and the upper distribution network, is the output power of the ith generator, , are the number of units and microgrids, , are the number of fixed loads and adjustable loads, is the kth fixed load, is the kth adjustable load, is the exchange power between microgrid i and the grid operator; (2) Exchange power constraint: The power exchange between the microgrid alliance and the grid operator cannot exceed the upper limit of the interconnection line power and transmission line capacity to avoid overload risks: , in, is the exchange power between the grid operator and the upper distribution network, and are its upper and lower limits, is the tie line power between the grid operator and microgrid i, and are its minimum and maximum power limits; (3) Voltage constraint:

[0010] , in, , , represents the voltage at the grid operator at time t, and They represent the squares of the minimum and maximum voltages of the grid operator at time t respectively.

[0011] As a preferred solution of the present invention, the objective function of the microgrid alliance is: ,

[0012] In the formula, , are the total dispatch duration and the number of microgrids, respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are the electricity price and the interaction power between microgrid i and other microgrids respectively.

[0013] As a preferred solution of the present invention, the constraints of the microgrid alliance include: (1) Power balance constraints: , In the formula, is the power of the tie line between microgrid i and the grid operator, is the power of the CHP unit, is the output of photovoltaic power in microgrid i during period t, , are the charging and discharging powers of the energy storage device of microgrid i in period t, is the interaction power between microgrid i and other microgrids, , are the powers regulated by the fixed load and adjustable load of microgrid i, , are the total dispatch duration and the number of microgrids, respectively; (2) Power exchange constraints: , , In the formula, is the interaction power between microgrid i and other microgrids, , are the upper and lower limits of the interaction power.

[0014] (3) Constraints on charging and discharging of energy storage equipment: , , in, , are the charging and discharging power of the energy storage in period t, and are the maximum values ​​of energy storage charging and discharging power respectively.

[0015] (4) Load response constraints: The upper and lower limit constraints and total amount constraints of the transferable electric load are as follows: , , in, is the proportion of transferable load to total load, is the total time, For the pre-demand response electric load, For transferable electrical loads.

[0016] The load reduction constraints are as follows: , In the formula, To reduce the proportion of electrical load to the total electrical load, To reduce the electrical load, For demand response front electric load.

[0017] As a preferred solution of the present invention, Nash bargaining game theory is used to describe the transaction interaction within the microgrid alliance, and the constructed model is as follows: , , , In the formula, , are the total dispatch duration and the number of microgrids, respectively; represents the total operating cost of microgrid i; is the cost of the combined heat and power unit in microgrid i, , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation cost of energy storage equipment, is the degradation coefficient, , are the charging and discharging power of the energy storage in period t respectively; is the communication cost between microgrid i and the grid operator, , are the contact electricity price and power respectively; is the carbon emission cost of the microgrid, is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; is the demand response cost, , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; is the cost of power exchange between microgrids, , are electricity price and interaction power between microgrid i and other microgrids respectively; is the cost of microgrid i operating independently without participating in the alliance, that is, the breaking point of Nash negotiation. ; is the overall operating cost of the microgrid i participating in the cooperation; The payment income obtained from the microgrid’s participation in cooperation.

[0018] As a preferred solution of the present invention, the objective functions of the upper and lower layers after decoupling are respectively: , , In the formula, is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are electricity price and interaction power between microgrid i and other microgrids respectively; , denote the decoupling objective functions of the grid operator and the microgrid, respectively; and Represents the grid operator and i A vector of microgrid decision variables; , represents the multiplier of the primary and quadratic terms in the kth iteration; The power traded between the grid operator and microgrid i; Indicates the optimization result of the lower-level problem’s feedback to the upper-level problem; For Microgrid i The power of the tie line with the grid operator; It represents the optimization result of the upper-level problem’s feedback to the lower-level problem. and They represent the penalty factors of the primary and quadratic terms of the upper penalty function respectively. The penalty terms are used to constrain unreasonable transactions between microgrids and ensure that transactions are fair and reasonable.

[0019] As a preferred solution of the present invention, the update multipliers of the primary and secondary terms of the penalty function are as follows: , Where: , represents the multiplier of the primary and quadratic terms in the kth iteration; , represents the multiplier of the primary and quadratic terms in the k+1th iteration; represents the penalty parameter magnification factor; is the penalty parameter Updated judgment weights; , They represent the microgrid power and tie-line power solved in the kth iteration respectively.

[0020] As a preferred solution of the present invention, the process of updating the tie line power and the interaction power by the Anderson acceleration method comprises the following steps: S91, Microgrid power acceleration update: , in, is the power traded between the grid operator and the microgrid in the kth iteration, is the number of microgrids, is the power traded between the grid operator and the microgrid in the k+1th iteration, is the Anderson coefficient of the power traded between the grid operator and the microgrid; S92, tie line power accelerated update: , in, is the tie line power between the microgrid and the grid operator in the kth iteration, is the number of microgrids, is the tie line power between the microgrid and the grid operator in the k+1th iteration, is the Anderson coefficient of the tie line power between the microgrid and the grid operator; S93, Interaction power between microgrids is accelerated and updated: , in, is the interaction power between microgrids in the kth iteration, is the number of microgrids, is the interaction power between microgrids in the k+1th iteration, is the Anderson coefficient of the interaction power between microgrids; S94. Determine the Anderson acceleration factor: , in, It represents the residual difference between the power traded between the grid operator and the microgrid and the result of the previous iteration in the k+1th iteration, represents the residual difference between the tie line power between the microgrid and the grid operator in the k+1th iteration and the result of the previous iteration, represents the residual between the microgrid interaction power in the k+1th iteration and the result of the previous iteration; S95, solve the coefficient: , in, It represents the residual difference between the power traded between the grid operator and the microgrid in the kth iteration and the result of the previous i iterations. It represents the residual difference between the power of the tie line between the microgrid and the grid operator in the kth iteration and the result of the previous i iterations. It represents the residual of the interaction power between microgrids in the kth iteration and the result of the previous i iterations.

[0021] As a preferred embodiment of the present invention, the convergence criterion is: , in, is the power traded between the grid operator and the microgrid in the kth iteration, is the power traded between the grid operator and the microgrid in the k+1th iteration; is the tie line power between the microgrid and the grid operator in the kth iteration, is the tie line power between the microgrid and the grid operator in the k+1th iteration; The interaction power between microgrids in the kth iteration, is the interaction power between microgrids in the k+1th iteration; represents the convergence accuracy of the power traded between the grid operator and the microgrid, represents the convergence accuracy of the tie line power between the microgrid and the grid operator, Convergence accuracy of interactive power between microgrids; The convergence condition is: , In the formula, , represents the convergence criterion. The former indicates the accuracy that needs to be guaranteed when the connection variables are optimized independently, and the latter represents the convergence criterion of the overall benefit of the scheduling model.

[0022] Beneficial effects: The present invention proposes a multi-microgrid collaborative scheduling method based on master-slave cooperative game and target cascade analysis, constructs a two-level master-slave cooperative game between power grid operators and microgrid alliances, and realizes efficient collaboration among multiple subjects; the upper-level power grid operator coordinates the overall situation, dynamically formulates electricity prices and carbon quota strategies, and optimizes system economy and carbon emission control; the lower-level microgrid alliance realizes internal power trading and resource sharing through cooperative games, and reduces dependence on high-carbon units; in response to the efficiency bottleneck of large-scale optimization, an improved target cascade analysis algorithm is designed, and the Anderson acceleration method is introduced to dynamically update the interconnection line power variables and combined with the augmented Lagrangian penalty function method to realize decoupled iterative optimization of the upper and lower layers; experiments show that the improved target cascade analysis algorithm improves the convergence speed, enhances the accuracy and efficiency of the results, and provides technical support for the low-carbon and efficient collaborative scheduling of multi-microgrid systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a framework diagram of a multi-microgrid system; Figure 2 To improve the ATC algorithm flow chart; Figure 3 Improve the iterative convergence process of the ATC algorithm for different scenarios; Figure 4 It is the iterative process of the power of the interconnection line between the microgrid and the grid operator in different scenarios; Figure 5 Improve the convergence characteristics of the ATC algorithm penalty term in different scenarios; Figure 6 It is the power balance diagram of each microgrid; Figure 7 This is the power interaction diagram between microgrids. DETAILED DESCRIPTION

[0024] The present invention will be further described with reference to the embodiments and the accompanying drawings.

[0025] As an embodiment of the present invention, Figure 1 to Figure 2 As shown, a multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis includes the following steps: S1. According to the power flow operation parameters of the multi-microgrid system, by constructing the microgrid alliance model, the grid operator model and the carbon emission management model, combined with the power balance constraints, exchange power and voltage constraints of the grid operator, the power balance constraints, power exchange constraints, energy storage equipment charging and discharging constraints, and load response constraints in the microgrid alliance, a two-layer master-slave cooperation game framework between the grid operator and the microgrid alliance is established; The grid operator, as the leader of the master-slave cooperative game framework, coordinates the power resources and interactive power between microgrids by adjusting the tie line power between the grid operator and each microgrid of the microgrid alliance under the condition of satisfying power balance constraints, exchange power and voltage constraints; As a collaborator in the master-slave cooperative game framework, the microgrid alliance allocates power resources between microgrids by adjusting the interactive power between microgrids while satisfying power balance constraints, power exchange constraints, energy storage device charging and discharging constraints, and load response constraints. At the same time, the Nash bargaining game theory is used to allocate the benefits of each microgrid to encourage each microgrid to participate in collaborative scheduling; Carbon emission management evaluates and controls the overall carbon emission level of the microgrid alliance during the coordinated dispatch process. By dynamically adjusting the carbon quota allocation of each microgrid, optimizing the unit operation mode and power trading strategy, the total carbon emission of the system is reduced, the low-carbon operation of the system is promoted, and ultimately the coordinated optimization of power resources and carbon emission targets is achieved; S2. Based on the idea of ​​target cascade analysis algorithm, the two-layer master-slave cooperative game problem between power grid operators and microgrid alliances is decoupled into an upper-layer power grid operator optimization problem and a lower-layer microgrid alliance problem that can be coupled through the tie line power variable. The Anderson acceleration method is introduced to dynamically adjust the tie line power variable, and the augmented Lagrangian penalty function is combined to handle constraints to achieve upper and lower decoupled iterative optimization. The convergence condition for iterative termination is set through the convergence criterion to accelerate convergence. S3. Based on the master-slave cooperative game framework built in step S1 and the target cascade analysis algorithm designed in step S2, the coordinated optimization of the power resources and carbon emissions of the microgrid alliance is achieved while satisfying the voltage stability, power flow constraints and power balance constraints. A scheduling plan that takes into account the fairness of all entities is output to adjust the interactive power within the microgrid alliance to ensure the low-carbon and stable operation of the microgrid alliance.

[0026] As an embodiment of the present invention, the objective function of the power grid operator is: ,

[0027] is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the output power of the ith generator, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively.

[0028] As an embodiment of the present invention, the constraints of the power grid operator include: (1) Power balance constraint: In each period, the power exchange between the grid operator, generators, loads and microgrids is balanced: , in, Represents the power exchange between the grid operator and the upper distribution network, is the output power of the ith generator, , are the number of units and microgrids, , are the number of fixed loads and adjustable loads, is the kth fixed load, is the kth adjustable load, is the exchange power between microgrid i and the grid operator; (2) Exchange power constraint: The power exchange between the microgrid alliance and the grid operator cannot exceed the upper limit of the interconnection line power and transmission line capacity to avoid overload risks: , in, is the exchange power between the grid operator and the upper distribution network, and are its upper and lower limits, is the tie line power between the grid operator and microgrid i, and are its minimum and maximum power limits; (3) Voltage constraint: , in, , , represents the voltage at the grid operator at time t, and They represent the squares of the minimum and maximum voltages of the grid operator at time t respectively.

[0029] As a preferred solution of the present invention, the objective function of the microgrid alliance is: , In the formula, , are the total dispatch duration and the number of microgrids, respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are the electricity price and the interaction power between microgrid i and other microgrids respectively.

[0030] As an embodiment of the present invention, the constraints of the microgrid alliance include: (1) Power balance constraints: , In the formula, is the power of the tie line between microgrid i and the grid operator, is the power of the CHP unit, is the output of photovoltaic power in microgrid i during period t, , are the charging and discharging powers of the energy storage device of microgrid i in period t, is the interaction power between microgrid i and other microgrids, , are the powers regulated by the fixed load and adjustable load of microgrid i, , are the total dispatch duration and the number of microgrids, respectively; (2) Power exchange constraints: , , In the formula, is the interaction power between microgrid i and other microgrids, , are the upper and lower limits of the interaction power.

[0031] (3) Constraints on charging and discharging of energy storage equipment: , , in, , are the charging and discharging power of the energy storage in period t, and are the maximum values ​​of energy storage charging and discharging power respectively.

[0032] (4) Load response constraints: The upper and lower limit constraints and total amount constraints of the transferable electric load are as follows: , , in, is the proportion of transferable load to total load, is the total time, For the pre-demand response electric load, For transferable electrical loads.

[0033] The load reduction constraints are as follows: , In the formula, To reduce the proportion of electrical load to the total electrical load, To reduce the electrical load, For demand response front electric load.

[0034] As a preferred solution of the present invention, the Nash bargaining game theory is used to describe the transaction interaction within the microgrid alliance, and the constructed model is as follows: , , , In the formula, , are the total dispatch duration and the number of microgrids, respectively; represents the total operating cost of microgrid i; is the cost of the combined heat and power unit in microgrid i, , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation cost of energy storage equipment, is the degradation coefficient, , are the charging and discharging power of the energy storage in period t respectively; is the communication cost between microgrid i and the grid operator, , are the contact electricity price and power respectively; is the carbon emission cost of the microgrid, is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; is the demand response cost, , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; is the cost of power exchange between microgrids, , are electricity price and interaction power between microgrid i and other microgrids respectively; is the cost of microgrid i operating independently without participating in the alliance, that is, the breaking point of Nash negotiation. ; is the overall operating cost of the microgrid i participating in the cooperation; The payment income obtained from the microgrid’s participation in cooperation.

[0035] As an embodiment of the present invention, the objective functions of the upper and lower layers after decoupling are respectively: , , In the formula, is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are electricity price and interaction power between microgrid i and other microgrids respectively; , denote the decoupling objective functions of the grid operator and the microgrid, respectively; and Represents the grid operator and i A vector of microgrid decision variables; , represents the multiplier of the primary and quadratic terms in the kth iteration; The power traded between the grid operator and microgrid i; Indicates the optimization result of the lower-level problem’s feedback to the upper-level problem; For Microgrid i The power of the tie line with the grid operator; It represents the optimization result of the upper-level problem’s feedback to the lower-level problem. and They represent the penalty factors of the primary and quadratic terms of the upper penalty function respectively. The penalty terms are used to constrain unreasonable transactions between microgrids and ensure that transactions are fair and reasonable.

[0036] As an embodiment of the present invention, the update multipliers of the primary and secondary terms of the penalty function are as follows: , Where: , represents the multiplier of the primary and quadratic terms in the kth iteration; , represents the multiplier of the primary and quadratic terms in the k+1th iteration; represents the penalty parameter magnification factor; is the penalty parameter Updated judgment weights; , They represent the microgrid power and tie-line power solved in the kth iteration respectively.

[0037] As an embodiment of the present invention, the process of updating the tie line power and the interaction power by the Anderson acceleration method includes the following steps: S91, Microgrid power acceleration update: , in, is the power traded between the grid operator and the microgrid in the kth iteration, is the number of microgrids, is the power traded between the grid operator and the microgrid in the k+1th iteration, is the Anderson coefficient of the power traded between the grid operator and the microgrid; S92, tie line power accelerated update: , in, is the tie line power between the microgrid and the grid operator in the kth iteration, is the number of microgrids, is the tie line power between the microgrid and the grid operator in the k+1th iteration, is the Anderson coefficient of the tie line power between the microgrid and the grid operator; S93, Interaction power between microgrids is accelerated and updated: , in, is the interaction power between microgrids in the kth iteration, is the number of microgrids, is the interaction power between microgrids in the k+1th iteration, is the Anderson coefficient of the interaction power between microgrids; S94. Determine the Anderson acceleration factor: , in, It represents the residual difference between the power traded between the grid operator and the microgrid and the result of the previous iteration in the k+1th iteration, represents the residual difference between the tie line power between the microgrid and the grid operator in the k+1th iteration and the result of the previous iteration, represents the residual between the microgrid interaction power in the k+1th iteration and the result of the previous iteration; S95, solve the coefficient: , in, It represents the residual difference between the power traded between the grid operator and the microgrid in the kth iteration and the result of the previous i iterations. It represents the residual difference between the power of the tie line between the microgrid and the grid operator in the kth iteration and the result of the previous i iterations. It represents the residual of the interaction power between microgrids in the kth iteration and the result of the previous i iterations.

[0038] As an embodiment of the present invention, the convergence criterion is: , in, is the power traded between the grid operator and the microgrid in the kth iteration, is the power traded between the grid operator and the microgrid in the k+1th iteration; is the tie line power between the microgrid and the grid operator in the kth iteration, is the tie line power between the microgrid and the grid operator in the k+1th iteration; The interaction power between microgrids in the kth iteration, is the interaction power between microgrids in the k+1th iteration; represents the convergence accuracy of the power traded between the grid operator and the microgrid, represents the convergence accuracy of the tie line power between the microgrid and the grid operator, Convergence accuracy of interactive power between microgrids; The convergence condition is: , In the formula, , represents the convergence criterion. The former indicates the accuracy that needs to be guaranteed when the connection variables are optimized independently, and the latter represents the convergence criterion of the overall benefit of the scheduling model.

[0039] As an embodiment of the present invention, Figures 3 to 5 As shown in Figure 1, the model parameters of the test grid operator and each microgrid are initialized first. The grid operator includes three traditional diesel generators with powers of 350kw, 300kw and 250kw respectively. Microgrid 1 has a 200kw CHP unit, a 250kw photovoltaic unit, and a 200kw energy storage device; Microgrid 2 has a 150kw photovoltaic unit and a 200kw energy storage device; Microgrid 3 has a 400kw CHP unit, a 100kw photovoltaic unit, and a 100kw energy storage device.

[0040] The following two scenarios are set for the test case to verify the practicality of the dispatch strategy: 1) The transactions between sub-microgrids are not considered, and the sub-microgrids only trade electricity with the grid operator; 2) The transactions between sub-microgrids are considered, and the form of microgrid alliance cooperation game is adopted; at the same time, the microgrid alliance can also trade electricity with the grid operator.

[0041] In the first scenario, that is, without considering the transaction between microgrids, the coordinated dispatch process of the entire multi-microgrid system is as follows: First, the main grid issues an initial dispatch plan to the grid operator based on the system operation status, and the grid operator then determines the initial tie-line power plan with each microgrid based on the superior instructions and its own model. In this case, there is no direct power interaction between the microgrids, and each microgrid only interacts with the grid operator. Subsequently, the grid operator (upper-level problem) and each microgrid (lower-level problem) decouple the problem based on the ATC algorithm, and each independently optimizes the model parameters. Next, the grid operator and each microgrid use the tie-line power variable to process the constraints through the augmented Lagrangian penalty function method, and dynamically update the tie-line power variable with the help of the Anderson acceleration method to improve the iteration efficiency. In the iterative solution stage, the grid operator and each microgrid repeatedly exchange optimization results and update variables until the tie-line power variable meets the convergence criterion of the algorithm and reaches the preset convergence accuracy, and then outputs the final tie-line power plan between the grid operator and the microgrid. At this point, the system forms a stable collaborative dispatching plan, and each microgrid executes power interaction with the grid operator according to the obtained plan to achieve stable and efficient operation of the entire system.

[0042] In the second scenario, that is, when power trading between microgrids is allowed, the system scheduling process is different. At this time, the main grid also issues an initial overall scheduling plan to the grid operator, and the grid operator formulates the initial power strategy of the interconnection line based on this, and issues the scheduling plan to the microgrid alliance. In this mode, there is also power interaction between the sub-microgrids, that is, the transaction interaction within the microgrid alliance is carried out in the form of Nash bargaining game to achieve resource sharing and maximize overall economic benefits. Therefore, the optimization problem of the lower-level microgrid alliance is more complicated than the first scenario, and it is necessary to consider the interconnection line power trading with the grid operator and the interactive power trading problem between the microgrids at the same time. In this case, the upper-level problem of the grid operator and the lower-level problem of the microgrid alliance are decoupled through the target cascade analysis algorithm, and further subdivided into multiple sub-problems. Each sub-microgrid optimizes the local power generation and load scheduling at the same time, and determines the power trading strategy with other microgrids. Each microgrid feeds back the local optimization results to the grid operator, and the grid operator dynamically adjusts the interconnection line power strategy with the microgrid alliance as a whole. This process also uses the Anderson acceleration method to iteratively accelerate the update of the interconnection line power between the grid operator and the microgrid and the interaction power between the microgrids until the convergence criterion of the algorithm is reached and converges to the set accuracy. Finally, the system outputs the transaction plan between the microgrids and the interconnection line power allocation plan between the microgrid alliance as a whole and the grid operator. Each microgrid and grid operator performs corresponding power interaction and resource allocation according to this plan, achieving low-carbon, economical, efficient and stable operation under the conditions of multi-subject participation.

[0043] Based on the target cascade analysis (ATC) algorithm, the multi-microgrid coordinated dispatch model is solved in a distributed manner. The iterative convergence process of scenario 1 (without microgrid transactions) and scenario 2 (with microgrid transactions) is shown in the figure. Figure 3 As shown in the figure: Scenario 1 converges overall after 7 iterations, and Scenario 2 converges after 9 iterations due to the increase in the complexity of microgrid transactions, and the residuals are all controlled within 10^-3. The iterative process of the power of the microgrid and the grid operator's tie line in the two scenarios: In Scenario 1, when only the microgrid interacts with the grid, the tie line power tends to be consistent after 7 iterations; After the introduction of microgrid transactions in Scenario 2, the tie line power needs 9 iterations to converge, indicating that the increase in transaction complexity has no essential impact on the system equilibrium; Figure 4 shown.

[0044] like Figure 5 As shown in the figure, the convergence characteristics of the penalty term in the two scenarios are further compared. In scenario 1, due to the lack of interaction between microgrids, the initial deviation of the penalty term is small and changes slowly; in scenario 2, due to the need to meet the power balance constraint between microgrids (P1+P2+P3=0) and the differences in the structure of each microgrid unit, the initial deviation of the penalty term is significant, but it gradually converges to the same order of magnitude through 9 iterations, verifying the adaptability of the algorithm to complex constraints.

[0045] After the microgrid alliance cooperation, the revenue of each microgrid increased significantly: Microgrid 1, 2, and 3 increased by 4.5%, 8.79%, and 1.85% respectively, indicating that electricity sharing can optimize resource allocation and improve economic benefits; at the same time, the revenue of grid operators decreased due to the reduced dependence on microgrids, reflecting the characteristics of interest redistribution under multi-agent game.

[0046] From the perspective of energy structure, microgrids 1 and 2 are mainly photovoltaic, and can sell redundant electricity to microgrid 3 when the load is low; microgrid 3 is based on gas turbines, and can reverse power supply when the load is peak, forming a complementary relationship. Figure 6 As shown in the figure, power interaction between microgrids. Microgrid 3 needs additional power support during a specific period. At this time, microgrids 1 and 2 provide clean electricity through P2P transactions to reduce the output of their high-carbon units. Although the transaction volume between microgrids is limited, it can achieve local balance in the regional network by dynamically adjusting the purchase and sale strategy, thus reducing the pressure on grid dispatching. Figure 7 shown.

[0047] The above simulation results verify the effectiveness and practicality of the model and algorithm constructed by the present invention. Through the master-slave cooperative game framework and the improved ATC algorithm, efficient collaboration of multiple subjects is achieved, the dynamic optimization efficiency is improved, and the real-time scheduling requirements of the system are ensured. This method can provide a reference for low-carbon and efficient collaborative scheduling of multi-microgrid systems and has certain engineering practical value.

[0048] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis, characterized by: The steps include: S1. According to the power flow operation parameters of the multi-microgrid system, by constructing the microgrid alliance model, the grid operator model and the carbon emission management model, combined with the power balance constraints, exchange power and voltage constraints of the grid operator, the power balance constraints, power exchange constraints, energy storage equipment charging and discharging constraints, and load response constraints in the microgrid alliance, a two-layer master-slave cooperation game framework between the grid operator and the microgrid alliance is established; The grid operator, as the leader of the master-slave cooperative game framework, coordinates the power resources and interactive power between microgrids by adjusting the tie line power between the grid operator and each microgrid of the microgrid alliance under the condition of satisfying power balance constraints, exchange power and voltage constraints; As a collaborator in the master-slave cooperative game framework, the microgrid alliance allocates power resources between microgrids by adjusting the interactive power between microgrids while satisfying power balance constraints, power exchange constraints, energy storage device charging and discharging constraints, and load response constraints. At the same time, the Nash bargaining game theory is used to allocate the benefits of each microgrid to encourage each microgrid to participate in collaborative scheduling; Carbon emission management evaluates and controls the overall carbon emission level of the microgrid alliance during the coordinated dispatch process. By dynamically adjusting the carbon quota allocation of each microgrid, optimizing the unit operation mode and power trading strategy, the total carbon emission of the system is reduced, the low-carbon operation of the system is promoted, and ultimately the coordinated optimization of power resources and carbon emission targets is achieved; S2. Based on the idea of ​​target cascade analysis algorithm, the two-layer master-slave cooperative game problem between power grid operators and microgrid alliances is decoupled into an upper-layer power grid operator optimization problem and a lower-layer microgrid alliance problem that can be coupled through the tie line power variable. The Anderson acceleration method is introduced to dynamically adjust the tie line power variable, and the augmented Lagrangian penalty function is combined to handle constraints to achieve upper and lower decoupled iterative optimization. The convergence condition for iterative termination is set through the convergence criterion to accelerate convergence. S3. Based on the master-slave cooperative game framework built in step S1 and the target cascade analysis algorithm designed in step S2, the coordinated optimization of the power resources and carbon emissions of the microgrid alliance is achieved while satisfying the voltage stability, power flow constraints and power balance constraints. A scheduling plan that takes into account the fairness of all entities is output to adjust the interactive power within the microgrid alliance to ensure the low-carbon and stable operation of the microgrid alliance.

2. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The objective function of the grid operator is: ， is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the output power of the ith generator, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively.

3. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The grid operator's constraints include: (1) Power balance constraint: In each period, the power exchange between the grid operator, generators, loads and microgrids is balanced: ， in, Represents the power exchange between the grid operator and the upper distribution network, is the output power of the ith generator, , are the number of units and microgrids, , are the number of fixed loads and adjustable loads, is the kth fixed load, is the kth adjustable load, is the exchange power between microgrid i and the grid operator; (2) Exchange power constraint: The power exchange between the microgrid alliance and the grid operator cannot exceed the upper limit of the interconnection line power and transmission line capacity to avoid overload risks: ， in, is the exchange power between the grid operator and the upper distribution network, and are its upper and lower limits, is the tie line power between the grid operator and microgrid i, and are its minimum and maximum power limits; (3) Voltage constraint: ， in, , , represents the voltage at the grid operator at time t, and They represent the squares of the minimum and maximum voltages of the grid operator at time t respectively.

4. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The objective function of the microgrid alliance is: ， In the formula, , are the total dispatch duration and the number of microgrids, respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are the electricity price and the interaction power between microgrid i and other microgrids respectively.

5. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The constraints of the microgrid alliance include: (1) Power balance constraints: ， In the formula, is the power of the tie line between microgrid i and the grid operator, is the power of the CHP unit, is the output of photovoltaic power in microgrid i during period t, , are the charging and discharging powers of the energy storage device of microgrid i in period t, is the interaction power between microgrid i and other microgrids, , are the powers regulated by the fixed load and adjustable load of microgrid i, , are the total dispatch duration and the number of microgrids, respectively; (2) Power exchange constraints: ， ， In the formula, is the interaction power between microgrid i and other microgrids, , are the upper and lower limits of the interaction power; (3) Constraints on charging and discharging of energy storage equipment: ， ， in, , are the charging and discharging power of the energy storage in period t, and are the maximum values ​​of energy storage charging and discharging power respectively; (4) Load response constraints: The upper and lower limit constraints and total amount constraints of the transferable electric load are as follows: ， ， in, is the proportion of transferable load to total load, is the total time, For the pre-demand response electric load, For transferable electrical loads; The load reduction constraints are as follows: ， In the formula, To reduce the proportion of electrical load to the total electrical load, To reduce the electrical load, For demand response front electric load.

6. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The Nash bargaining game theory is used to describe the transaction interaction within the microgrid alliance. The constructed model is as follows: ， ， ， In the formula, , are the total dispatch duration and the number of microgrids, respectively; represents the total operating cost of microgrid i; is the cost of the combined heat and power unit in microgrid i, , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation cost of energy storage equipment, is the degradation coefficient, , are the charging and discharging power of the energy storage in period t respectively; is the communication cost between microgrid i and the grid operator, , are the contact electricity price and power respectively; is the carbon emission cost of the microgrid, is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; is the demand response cost, , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; is the cost of power exchange between microgrids, , are electricity price and interaction power between microgrid i and other microgrids respectively; is the cost of microgrid i operating independently without participating in the alliance, that is, the breaking point of Nash negotiation. ; is the overall operating cost of the microgrid i participating in the cooperation; The payment income obtained from the microgrid’s participation in cooperation.

7. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: After decoupling, the objective functions of the upper and lower layers are: ， ， In the formula, is the total dispatch time, , are the number of units and microgrids, is the operating coefficient of the diesel generator set, and are the model coefficients of the units, is the start and stop 0-1 variable of unit i in period t, and is the start-stop cost coefficient; , are the electricity price and power traded between the grid operator and the distribution network, respectively; , are the electricity price and power traded between the grid operator and microgrid i respectively; is the transaction price of carbon quotas, , are the CO2 emissions and carbon quota of unit i of the grid operator respectively; , , is the fuel coefficient of the CHP unit, is the power generated by the CHP unit; is the degradation coefficient of energy storage equipment, , are the charging and discharging power of the energy storage in period t respectively; , are the contact electricity price and power respectively; is the price of carbon quota trading between microgrids, , are the CO2 emissions and carbon quota of microgrid i respectively; , They are the compensation electricity price of controllable load and the adjustment power of adjustable load respectively; , are electricity price and interaction power between microgrid i and other microgrids respectively; , denote the decoupling objective functions of the grid operator and the microgrid, respectively; and Represents the grid operator and i A vector of microgrid decision variables; , represents the multiplier of the primary and quadratic terms in the kth iteration; The power traded between the grid operator and microgrid i; Indicates the optimization result of the lower-level problem’s feedback to the upper-level problem; For Microgrid i The power of the tie line with the grid operator; It represents the optimization result of the upper-level problem’s feedback to the lower-level problem. and They represent the penalty factors of the primary and quadratic terms of the upper penalty function respectively. The penalty terms are used to constrain unreasonable transactions between microgrids and ensure that transactions are fair and reasonable.

8. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 7 is characterized by: The update multipliers of the first and second terms of the penalty function are as follows: ， Where: , represents the multiplier of the primary and quadratic terms in the kth iteration; , represents the multiplier of the primary and quadratic terms in the k+1th iteration; represents the penalty parameter magnification factor; is the penalty parameter Updated judgment weights; , They represent the microgrid power and tie-line power solved in the kth iteration respectively.

9. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1 is characterized by: The process of updating the tie line power and the interaction power by the Anderson acceleration method comprises the following steps: S91, Microgrid power acceleration update: ， in, is the power traded between the grid operator and the microgrid in the kth iteration, is the number of microgrids, is the power traded between the grid operator and the microgrid in the k+1th iteration, is the Anderson coefficient of the power traded between the grid operator and the microgrid; S92, tie line power accelerated update: , in, is the tie line power between the microgrid and the grid operator in the kth iteration, is the number of microgrids, is the tie line power between the microgrid and the grid operator in the k+1th iteration, is the Anderson coefficient of the tie line power between the microgrid and the grid operator; S93, Interaction power between microgrids is accelerated and updated: , in, is the interaction power between microgrids in the kth iteration, is the number of microgrids, is the interaction power between microgrids in the k+1th iteration, is the Anderson coefficient of the interaction power between microgrids; S94. Determine the Anderson acceleration factor: , in, It represents the residual difference between the power traded between the grid operator and the microgrid and the result of the previous iteration in the k+1th iteration, represents the residual difference between the tie line power between the microgrid and the grid operator in the k+1th iteration and the result of the previous iteration, represents the residual between the microgrid interaction power in the k+1th iteration and the result of the previous iteration; S95, solve the coefficient: , in, It represents the residual difference between the power traded between the grid operator and the microgrid in the kth iteration and the power traded i times before. It represents the residual difference between the power of the tie line between the microgrid and the grid operator in the kth iteration and the result of the previous i iterations. It represents the residual of the interaction power between microgrids in the kth iteration and the result of the previous i iterations.

10. The multi-microgrid coordinated scheduling method based on master-slave cooperative game and target cascade analysis according to claim 1, characterized in that: The convergence criterion is: ， in, is the power traded between the grid operator and the microgrid in the kth iteration, is the power traded between the grid operator and the microgrid in the k+1th iteration; is the tie line power between the microgrid and the grid operator in the kth iteration, is the tie line power between the microgrid and the grid operator in the k+1th iteration; The interaction power between microgrids in the kth iteration, is the interaction power between microgrids in the k+1th iteration; represents the convergence accuracy of the power traded between the grid operator and the microgrid, represents the convergence accuracy of the tie line power between the microgrid and the grid operator, Convergence accuracy of interactive power between microgrids; The convergence condition is: ， , represents the convergence criterion. The former indicates the accuracy that needs to be guaranteed when the connection variables are optimized independently, and the latter represents the convergence criterion of the overall benefit of the scheduling model.

Images