Optimized scheduling method for micro-grid alliance containing shared energy storage and power distribution network based on mixed game

Abstract

The invention provides an optimal scheduling method for a micro-grid alliance containing shared energy storage and a power distribution network based on a mixed game. The method comprises the following steps of: constructing an integral system which takes an active power distribution network as a dominant and takes multiple accessed micro-grids containing energy storage as participants to coordinate a main body; a hybrid game system is constructed, and two-stage optimization is adopted in the game process: firstly, a multi-microgrid alliance is optimized, and charging and discharging strategies of all microgrids are jointly adjusted and optimized through shared energy storage through an asymmetric Nash equilibrium game; and then, the master-slave body is optimized, the active power distribution network adopts time-of-use electricity price to implement peak clipping and valley filling scheduling, and the energy storage-containing multi-microgrid alliance responds to an active power distribution network pricing strategy. The optimized scheduling model containing shared energy storage power distribution established by the invention can achieve the maximum overall benefit and a better peak regulation scheduling result.

Description

Technical Field

[0001] The present invention relates to the field of integrated energy system optimization, and in particular to a hybrid game-based optimization dispatching method for a microgrid alliance and a distribution network containing shared energy storage. Background Art

[0002] With the advancement of global industrialization, energy demand is increasing, and the consumption of traditional mineral energy is increasing. Against this background, my country has made remarkable achievements in the development of renewable energy in recent years. However, since renewable energy power generation is easily affected by external factors such as the natural environment, it has the characteristics of unstable output, uncontrollability and intermittency. Therefore, connecting renewable energy units to traditional distribution networks is a severe challenge for the operation and control of the power grid. In order to solve the global problem of new energy grid connection, microgrids have developed rapidly because they can effectively integrate and utilize decentralized renewable energy.

[0003] However, when multiple microgrids are connected to the active distribution network at the same time, the charging and discharging behavior of the overall system will become disordered, which will lead to ineffective utilization of energy storage resources. Moreover, from a macro perspective, each microgrid also causes ineffective utilization of energy resources due to its energy generation and consumption characteristics. Energy storage is an important means to smooth the fluctuations of new energy in microgrids. Through its efficient charging and discharging characteristics, it can offset the power fluctuations generated by new energy in grid-connected microgrids. However, energy storage usually has limited energy capacity in accommodating renewable energy. Therefore, exploring efficient integration strategies for new energy, shared energy storage and distribution networks in multiple microgrids, and promoting the friendly access of multiple microgrids to the grid, can provide new solutions for peak and valley regulation of ADN. Summary of the invention

[0004] The purpose of the present invention is to propose an optimized dispatching method for a microgrid alliance and distribution network containing shared energy storage based on a hybrid game. First, a distribution network is built as the leading factor, and multiple microgrids with energy storage connected are used as the coordinating entities to form an overall system. A hybrid game system is constructed, and the game process adopts a two-stage optimization: first, the microgrid alliance is optimized, and each microgrid uses an asymmetric Nash equilibrium game to jointly optimize its own charging and discharging strategies through shared energy storage regulation; then the master and slave are optimized, and the distribution network adopts time-of-use electricity prices to implement peak shaving and valley filling scheduling, and the microgrid alliance containing energy storage responds to the pricing strategy of the distribution network. The optimized dispatching model for power distribution containing shared energy storage established by the present invention can achieve the maximum overall benefit and better peak-shaving dispatching results.

[0005] To achieve the above purpose, the technical solution of the present invention is: a hybrid game-based optimization scheduling method for a microgrid alliance with shared energy storage and a distribution network, comprising the following steps:

[0006] Step S1: Collect historical output data and historical load data of new energy sources in the region, and establish an active distribution network framework including an active distribution network ADN and a multi-microgrid alliance MGA including leased shared energy storage RSES;

[0007] Step S2: Establish a hybrid game system mathematical model of the active distribution network framework and adopt a two-stage optimization strategy;

[0008] Phase 1: Multiple microgrids conduct asymmetric Nash equilibrium games by leasing shared energy storage, and build an asymmetric Nash equilibrium model that takes into account energy storage to smooth power fluctuations and reduce costs;

[0009] Phase 2: Multiple microgrid alliances operate together to form a master-slave game with the active distribution network, and a master-slave game model that takes into account both the master and the slave is constructed;

[0010] Step S3: Use the fast non-dominated sorting genetic algorithm with elite retention strategy to solve the asymmetric Nash equilibrium game problem, and build a solution model for the master-slave game. Use the CPLEX solver to solve the lower model and obtain a hybrid game optimization scheduling scheme for the active distribution network framework.

[0011] Step S4: Allocate the revenue of the multi-microgrid alliance based on the Shapley value.

[0012] Preferably, the active distribution network framework including the active distribution network ADN and the multi-microgrid alliance MGA including the leased shared energy storage RSES in step S1 is specifically described as follows:

[0013] ADN includes new energy generation systems and conventional loads, and multiple microgrids MG are connected to ADN at different locations;

[0014] The power interaction benefit of ADN and RSES is V RSES It is expressed as:

[0015]

[0016] In the formula: is the power sold and purchased by RSES and ADN during period t; is the ADN electricity purchase price; T is the number of time periods in the dispatching stage;

[0017] Power interaction benefits of ADN and multi-microgrid alliances V MGA It is expressed as:

[0018]

[0019] Where: , is the total power sold and purchased by MGA and ADN; n is the number of microgrids; is the power sold and purchased by microgrid i and ADN.

[0020] Preferably, in the hybrid game system mathematical model of step S2, the asymmetric Nash equilibrium game of multiple microgrids in phase 1 through leasing shared energy storage is specifically described as follows:

[0021] Objective function 1: Minimize the mean square error of net energy storage load;

[0022]

[0023] Where: I 1 Load mean square error after using leased shared energy storage for microgrids; is the load power, energy storage charging capacity, wind power generation power, photovoltaic power generation and energy storage release capacity of the i-th microgrid in period t; is the load average;

[0024] Objective function 2: Minimize the cost of energy storage application;

[0025] I 2 =J rent +J RSES

[0026]

[0027] Where: I 2 The application cost of leasing shared energy storage for microgrids; a, b, and c are the unit energy leasing cost, unit capacity leasing cost, and unit storage cost of energy storage, respectively; J rent , J RSES They represent the energy storage leasing cost and the energy storage and release power cost respectively; They represent the power demand and energy demand of microgrid i respectively;

[0028] The constraints of the microgrid include constant input and output power constraints and energy storage charging and discharging power and energy constraints.

[0029] Preferably, the input-output power constant constraint is expressed as:

[0030]

[0031] In the formula, are the charging and discharging efficiencies of microgrid i in one cycle respectively;

[0032] The energy storage charging and discharging power and energy constraints are expressed as:

[0033] SOC min E RSES ≤E′ RSES ≤SOC max E RSES

[0034] SOC' min P RSES ≤P′ RSES ≤SOC' max P RSES

[0035] Where: E RESE and P RESE To operate the energy capacity and power capacity of the shared energy storage power station; E' RSES , P' RSES Shared energy storage power station power and energy; SOC min , SOC max and SOC' min , SOC' max The minimum and maximum charge states of the energy and power of the shared energy storage station;

[0036]

[0037] Where: The maximum value of charging and discharging power of the shared energy storage power station; out , RSES It is the status bit of charge and discharge; and Δt are the energy capacity and time difference of the shared energy storage power station at time t, respectively.

[0038] Preferably, in the hybrid game system mathematical model of step S2, the multi-microgrid alliance in stage 2 jointly operates and forms a master-slave game with the active distribution network as specifically described as follows:

[0039] In the master-slave game model of ADN, MGA and RSES, the objective function of the main ADN is ADN It is expressed as:

[0040] I ADN =V MGA +V RSES -B grid

[0041] Where: V MGA 、V RSES , B grid are the power of active distribution network and multi-microgrid alliance, the interactive power benefit of shared energy storage, and the power purchase cost from the upper power grid, B grid It is expressed as:

[0042]

[0043] Where: t represents the unit time period; T is the number of time periods in the scheduling phase; The price of electricity from the upper power grid; The amount of electricity purchased by the active distribution network from the upper power grid;

[0044] The constraints of the subject ADN include the following:

[0045]

[0046] Where: i,t,l , i,t,h are the electricity purchase prices sold by the active distribution network to microgrid i during the peak and valley phases; ρ i,t,l,max , i,t,l,min are the upper and lower limits of the electricity price sold by the active distribution network to microgrid i; ρ i,t,h,max , i,t,h,min are the upper and lower limits of the electricity price purchased by the active distribution network from microgrid i;

[0047]

[0048] Where: is the maximum value of the average electricity price sold by ADN to microgrid i; i,t , i ' ,t is the average electricity price sold or purchased by ADN to microgrid i;

[0049] The objective function U of the MGA model containing RSES MGA It is expressed as follows:

[0050] U MGA =I ADNΣ -J MTΣ

[0051] Where: I ADNΣ is the total interactive power benefit between the multi-microgrid alliance and the active distribution network, J MTΣ The total power generation cost of the generator sets in the multi-microgrid alliance is:

[0052]

[0053] J MTΣ =J MTΣ,fuel +J MTΣ,o&m +J MTΣ,st +J MTΣ,em

[0054]

[0055] Where: J MTΣ,fuel , J MTΣ,o&m , J MTΣ,st , J MTΣ,em Respectively represent the fuel cost, operation and maintenance cost, startup and shutdown cost, and emission cost of the generator set; P MGi (t), EMMGi (t) is the output power and emission of the unit at time t; C fuel,p (t), C o&m,p (t), C em,p (t) is the unit price of fuel cost, operation and maintenance cost and emission cost; N sta 、N sh is the number of startup and shutdown; C sta,p , C sh,p The unit costs for startup and shutdown;

[0056] The constraint function of the MGA model containing RSES is expressed as follows:

[0057] Active power balance constraints of multi-microgrid alliance:

[0058]

[0059] Where: They are respectively represented as the interaction power between microgrid ij and microgrid ji during period t; represents the gas turbine power; and is the behavior variable of the multi-microgrid alliance purchasing and selling electricity from ADN. When it is 1, it means that the multi-microgrid alliance purchases electricity from ADN, and when it is 0, it means that the multi-microgrid alliance sells electricity to ADN;

[0060] Shared energy storage charging and discharging balance constraints:

[0061]

[0062] Where: The discharge power and charging power of the shared energy storage station;

[0063] Gas turbine output upper and lower limit constraints:

[0064]

[0065] Where: and are the upper and lower limits of gas turbine power;

[0066] Power exchange constraints between microgrids:

[0067]

[0068] In the formula, Respectively represent the conversion coefficients between different energy forms; They represent the remaining energy capacity between microgrids and the maximum power exchange after comprehensive consideration.

[0069] Preferably, the step S4 allocates the income of the multi-microgrid alliance based on the Shapley value, specifically as follows:

[0070] The multi-microgrid alliance adheres to the following principles: the overall operating cost shall not exceed the total cost of each microgrid when it operates independently; and the cost saved through cooperation, when allocated to each microgrid, must be lower than the cost of each microgrid operating independently;

[0071] The Shapley value distribution method is expressed as:

[0072]

[0073] Where: x i is the remaining cooperation amount allocated to alliance member i; e(i) is the benefit obtained by alliance member i; s i is the sub-alliance composed of member i; ω(|s|) is the weight coefficient; e(s) and e(s / i) are the remainder of sub-alliance s and the remainder after excluding member i; |s| is the number of members in the alliance.

[0074] Preferably, the step S3 uses a fast non-dominated sorting genetic algorithm with an elite retention strategy to solve the asymmetric Nash equilibrium game problem, specifically:

[0075] After obtaining the uniformly distributed Pareto optimal solution through the fast non-dominated sorting genetic algorithm with elite retention strategy, the grey system membership function is established, the compromise solution is screened by the grey system membership function, and the optimal charging and discharging strategy of each microgrid is recorded.

[0076] Preferably, the grey system membership function is expressed as:

[0077]

[0078] Where: μ(x) is the degree of membership; x is the independent variable; a 1 、a 2 is a constant.

[0079] Preferably, the master-slave game solution model of step S3 uses a CPLEX solver to solve the lower model, specifically:

[0080] The optimization model of the master-slave game is expressed as:

[0081] G={M;ρ j,t ; {ξ B ,ξ S ,θ B ,θ S};I ADN ; U MGA}

[0082] Participant set: ADN and MGA with RSES constitute the participant set M, where ADN is the main participant and MGA with RSES is the slave participant; Decision variables: ADN decision variables are in the form of time period electricity price ρ j,t The decision variables of the MGA with RSES are the power of each microgrid to purchase and sell electricity to the main grid. And the power of shared energy storage to purchase and sell electricity to the main grid

[0083] The master-slave equilibrium condition of the Stackelberg game is:

[0084]

[0085] In the formula, the superscript * indicates the equilibrium solution after solving; then The master-slave game is solved.

[0086] Preferably, the upper-layer ADN of the master-slave game optimization model is solved using the whale optimization algorithm, and the whale optimization algorithm is used to calculate and update the ADN time-of-use electricity price.

[0087] Compared with the prior art, the present invention has the following beneficial effects:

[0088] In the present invention, with the help of the time-of-use pricing mechanism of ADN, the microgrid containing leased shared energy storage can design and execute its own charging, discharging and energy management strategies. And these strategies cooperate to participate in the peak-valley load balancing task of the distribution network. At the same time, the optimization algorithm determines the appropriate energy storage capacity, so that when the load fluctuates in the microgrid, it can be smoothed by the energy storage system, ensuring the effective balance of the economic benefits of shared energy storage. And according to the current energy demand of the microgrid and the status of the energy storage equipment, it is determined in real time whether shared energy storage services are needed. And by implementing a shared energy storage mechanism, it is beneficial to reduce the cost of microgrid operation. And within the microgrid alliance system, by giving priority to the internal power coordination mechanism, the effective consumption of new energy in the local area is significantly improved. And by adopting the Shapley value analysis method, the problem of interest distribution in the microgrid alliance is successfully solved, thereby ensuring its fairness and operating efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 This is a distribution network framework diagram including leased shared energy storage and microgrids according to an embodiment of the present invention;

[0090] Figure 2 A schematic diagram of a two-stage hybrid game scheduling strategy for an active power distribution network system according to an embodiment of the present invention;

[0091] Figure 3 A flow chart of solving a hybrid game model according to an embodiment of the present invention;

[0092] Figure 4 This is a schematic diagram of the structure of a 33-node active power distribution network according to an embodiment of the present invention;

[0093] Figure 5 This is a comparison diagram of the net load before and after charging and discharging of the microgrid 1 according to an embodiment of the present invention;

[0094] Figure 6 This is a comparison diagram of the net load before and after charging and discharging of the microgrid 2 according to an embodiment of the present invention;

[0095] Figure 7 This is a comparison diagram of the net load before and after charging and discharging of the microgrid 3 according to an embodiment of the present invention. DETAILED DESCRIPTION

[0096] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings.

[0097] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0098] The present invention proposes a hybrid game-based optimization scheduling method for a microgrid alliance and a distribution network containing shared energy storage.

[0099] like Figure 1 The following is a framework diagram of the active distribution network. It mainly includes the active distribution network and a multi-microgrid alliance with leased shared energy storage; Figure 2 The hybrid game dispatching model is established in two stages: in the first stage, multiple power grids conduct asymmetric Nash equilibrium games by leasing shared energy storage; in the second stage, multiple microgrid alliances jointly operate and form a master-slave game with the active distribution network. Then, the corresponding mathematical model is established based on this game model, and the solution method is as follows: Figure 3 As shown in the flowchart, the fast non-dominated sorting genetic algorithm with elite retention strategy, the whale optimization algorithm and the CPLEX solver are used to solve the game model. Finally, substitute Figure 4 The IEEE distribution network was verified to prove that the maximum benefit of the overall distribution network and the best peak-shaving and valley-filling effect can be achieved. The present invention provides an optimization scheduling method for a microgrid alliance with shared energy storage and a distribution network based on a hybrid game, which has the best peak-shaving and valley-filling effect and maximum benefit compared with other models.

[0100] Next, the implementation steps of the present invention are described in detail in conjunction with the specific process.

[0101] 1. Establish an active distribution network system framework that includes an active distribution network (ADN) and a multi-microgrid alliance (MGA) with rental shared energy storage (RSES).

[0102] The operation goal of ADN is to achieve load peak shaving and valley filling, and maximize dispatching and operation benefits. ADN contains its own wind turbines, photovoltaic and other new energy power generation systems and conventional loads. Multiple micro-grids (MG) are connected to ADN at different locations; the operation goal of the multi-microgrid alliance with shared energy storage is to coordinate the smoothing of power fluctuations of multiple microgrids and the interconnection between microgrids, and promote the local consumption of new energy. The power interaction benefits of ADN and RSES are V RSES It is expressed as:

[0103]

[0104] Where: is the power sold and purchased by RSES and ADN during period t; is the ADN electricity purchase price; T is the number of time periods in the dispatching stage;

[0105] The power interaction benefit between ADN and multi-microgrid alliance is expressed as:

[0106]

[0107] Where: is the total power sold and purchased by MGA and ADN; n is the number of microgrids; is the power sold and purchased by microgrid i and ADN.

[0108] 2. A hybrid game scheduling mathematical model is established and carried out in two stages: In the first stage, multiple power grids conduct asymmetric Nash equilibrium games by leasing shared energy storage; in the second stage, multiple microgrid alliances operate together and form a master-slave game with the active distribution network.

[0109] In the asymmetric Nash equilibrium game of leasing energy storage in a multi-microgrid alliance:

[0110] Objective function 1: Minimize the mean square error of net energy storage load;

[0111]

[0112] Where: I 1 Load mean square error after using leased shared energy storage for microgrids; is the load power, energy storage charging capacity, wind power generation power, photovoltaic power generation and energy storage release capacity of the i-th microgrid in period t; is the load average;

[0113] Objective function 2: Minimize the cost of energy storage application;

[0114] I 2 =J rent +J RSES

[0115]

[0116] Where: I 2 The application cost of leasing shared energy storage for microgrids; a, b, and c are the unit energy leasing cost, unit capacity leasing cost, and unit storage cost of energy storage, respectively; J rent , J RSES They represent the energy storage leasing cost and the energy storage and release power cost respectively; They represent the power demand and energy demand of microgrid i respectively;

[0117] The input and output power constant constraint of the microgrid is expressed as:

[0118]

[0119] In the formula, are the charging and discharging efficiencies of microgrid i in one cycle respectively;

[0120] The energy storage charging and discharging power and energy constraints are expressed as:

[0121] SOC min E RSES ≤E′ RSES ≤SOC max E RSES

[0122] SOC' min P RSES ≤P′ RSES ≤SOC' max P RSES

[0123] Where: E RESE and P RESE To operate the energy capacity and power capacity of the shared energy storage power station; E' RSES , P' RSES Shared energy storage power station power and energy; SOC min , SOC max and SOC' min , SOC' max The minimum and maximum charge states of the energy and power of the shared energy storage power station;

[0124]

[0125] Where: The maximum value of charging and discharging power of the shared energy storage power station; out , RSES It is the status bit of charge and discharge; and Δt are the energy capacity and time difference of the shared energy storage power station at time t, respectively.

[0126] In the master-slave game model of ADN, MGA and RSES, the objective function of the main ADN is ADN It is expressed as:

[0127] I ADN =V MGA +V RSES -B grid

[0128] Where: V MGA 、V RSES , B grid are the power of active distribution network and multi-microgrid alliance, the interactive power benefit of shared energy storage, and the power purchase cost from the upper power grid, B grid It is expressed as:

[0129]

[0130] Where: t represents the unit time period; T is the number of time periods in the scheduling phase; The price of electricity from the upper power grid; The amount of electricity purchased by the active distribution network from the upper power grid;

[0131] The constraints of the subject ADN include the following:

[0132]

[0133] Where: i,t,l , i,t,h are the electricity purchase prices sold by the active distribution network to microgrid i during the peak and valley phases; ρ i,t,l,max , i,t,l,min are the upper and lower limits of the electricity price sold by the active distribution network to microgrid i; ρ i,t,h,max , i,t,h,min are the upper and lower limits of the electricity price purchased by the active distribution network from microgrid i;

[0134]

[0135] Where: is the maximum value of the average electricity price sold by ADN to microgrid i; i,t ,i ' ,t is the average electricity price sold or purchased by ADN to microgrid i;

[0136] The objective function U of the MGA model containing RSES MGA It is expressed as follows:

[0137] U MGA =I ADNΣ -J MTΣ

[0138] Where: I ADNΣ is the total interactive power benefit between the multi-microgrid alliance and the active distribution network, J MTΣ is the total power generation cost of the generator sets in the multi-microgrid alliance, as shown below:

[0139]

[0140] J MTΣ =J MTΣ,fuel +J MTΣ,o&m +J MTΣ,st +J MTΣ,em

[0141]

[0142] Where: J MTΣ,fuel , J MTΣ,o&m , J MTΣ,st , J MTΣ,em Respectively represent the fuel cost, operation and maintenance cost, startup and shutdown cost, and emission cost of the generator set; P MGi (t), EM MGi (t) is the output power and emission of the unit at time t; C fuel,p (t), C o&m,p (t), C em,p (t) is the unit price of fuel cost, operation and maintenance cost and emission cost; N sta 、N sh is the number of startup and shutdown; C sta,p , C sh,p The unit costs for startup and shutdown;

[0143] The constraint function of the MGA model containing RSES is expressed as follows:

[0144] Active power balance constraints of multi-microgrid alliance:

[0145]

[0146] Where: They are respectively represented as the interaction power between microgrid ij and microgrid ji during period t; represents the gas turbine power; and is the behavior variable of the multi-microgrid alliance purchasing and selling electricity from ADN. When it is 1, it means that the multi-microgrid alliance purchases electricity from ADN, and when it is 0, it means that the multi-microgrid alliance sells electricity to ADN;

[0147] Shared energy storage charging and discharging balance constraints:

[0148]

[0149] Where: The discharge power and charging power of the shared energy storage station;

[0150] Gas turbine output upper and lower limit constraints:

[0151]

[0152] Where: and are the upper and lower limits of gas turbine power;

[0153] Power exchange constraints between microgrids:

[0154]

[0155] In the formula, Respectively represent the conversion coefficients between different energy forms; They represent the remaining energy capacity between microgrids and the maximum power exchange after comprehensive consideration.

[0156] 3. The fast non-dominated sorting genetic algorithm with elite retention strategy, whale optimization algorithm and CPLEX solver are used to solve the game model.

[0157] After obtaining the uniformly distributed Pareto optimal solution through the fast non-dominated sorting genetic algorithm with elite retention strategy, a grey system membership function is established in this scenario, and the compromise solution is screened using the grey system membership function, and the optimal charging and discharging strategy of each microgrid is recorded. Such a grey system membership function can be used to support tasks such as fuzzy logic reasoning and fuzzy decision-making to help deal with decision-making problems with uncertainty or fuzzy information, and provide effective decision support by modeling and analyzing limited data. The grey system membership function is expressed as:

[0158]

[0159] Where: μ(x) is the degree of membership; x is the independent variable; a 1 、a 2 is a constant.

[0160] Furthermore, the master-slave game solution model of step S4 uses the CPLEX solver to solve the lower layer model. The specific description is:

[0161] The optimization model of the master-slave game is expressed as:

[0162] G={M;ρ j,t ; {ξ B ,ξ S ,θ B ,θ S};I ADN ; U MGA}

[0163] Participant set: ADN and MGA with RSES constitute the participant set M, where ADN is the main participant and MGA with RSES is the slave participant; Decision variables: ADN decision variables are in the form of time period electricity price ρ j,t The decision variables of the MGA with RSES are the power of each microgrid to purchase and sell electricity to the main grid. And the power of shared energy storage to purchase and sell electricity to the main grid

[0164] The master-slave equilibrium condition of the Stackelberg game is:

[0165]

[0166] In the formula, the superscript * indicates the equilibrium solution after solving; then The master-slave game is solved. The whale optimization algorithm and CPLEX solver are used to solve the game model. The upper layer ADN is solved by the whale optimization algorithm. The whale optimization algorithm is used to calculate and update the ADN time-of-use electricity price, and the lower layer model is solved by the CPLEX solver.

[0167] 4. Profit distribution of multi-microgrid alliance based on Shapley value.

[0168] The key to alliance collaboration lies in collective application, and its effectiveness depends on two core principles: first, the overall operating cost should not exceed the total cost of each microgrid independently allowed; second, the cost saved through cooperation must be lower than the cost of their independent operation when allocated to each microgrid. By following these two principles, it helps the microgrid alliance reduce its dependence on the main grid and improve the overall operating efficiency of the system. The Shapley value allocation method is expressed as:

[0169]

[0170] Where: x i is the remaining cooperation amount allocated to alliance member i; e(i) is the benefit obtained by alliance member i; s iis the sub-alliance composed of member i; ω(|s|) is the weight coefficient; e(s) and e(s / i) are the remainder of sub-alliance s and the remainder after excluding member i; |s| is the number of members in the alliance.

[0171] 5. Substitute the IEEE33 node distribution network system for verification and obtain the optimal scheduling results of the hybrid model.

[0172] The IEEE33-node distribution network system was selected for verification. The hybrid game mathematical model was established through the above; the whale optimization algorithm and CPLEX commercial solver were used to optimize the dispatch of the distribution network system. In order to conduct in-depth empirical research, three groups of cases were constructed and compared to analyze the effect of the distribution network in peak shaving and valley filling under various circumstances.

[0173] Solution 1: The distribution network and other subordinates are all fixed priced, and there is no interaction between the parts; Solution 2: Based on Solution 1, ADN implements a time-of-use electricity price system, but only the microgrid alliance responds to this mechanism, and the shared energy storage does not respond; Solution 3 (the strategy proposed in this article): Based on Solution 2, ADN introduces a dispatching mechanism based on game theory, allowing the microgrid alliance with shared energy storage to respond to changes in time-of-use electricity prices, thereby optimizing their respective energy management and promoting power complementarity between microgrids.

[0174] Table 1 analyzes the results of the three schemes in detail. Scheme 3 fully considers the application of RSES in the peak-shaving and valley-filling dispatch of the distribution network. The distribution network introduces a dispatch mechanism based on game theory. Through the master-slave game model, the standard deviation of the ADN net load is 382.12kW, while Scheme 1 adopts a fixed electricity price, which makes the standard deviation of the ADN net load 625.14kW. Compared with Scheme 3, the net load peak-valley difference rate is reduced by 9.93%. It shows that under the master-slave game model of Scheme 3, the distribution network load presents a better peak-shaving and valley-filling effect. Therefore, Scheme 3 has a better ADN peak-shaving performance.

[0175] Table 1: Peak load regulation results of distribution networks in three examples

[0176]

[0177] In order to determine the optimal charging and discharging power scheme of each microgrid and the required energy storage capacity demand, the NSGA-II algorithm is used to obtain the solution of the asymmetric Nash equilibrium game, and then the optimal solution is obtained through the grey system membership function. Figure 5-7 The results show how each microgrid uses energy storage to offset net load fluctuations. The net load curve can be significantly smoothed through the energy storage charging and discharging strategy, clearly demonstrating the effectiveness of each microgrid in managing fluctuations.

[0178] The Shapley value allocation principle ensures the fair distribution of the value generated by the multi-party cooperation of microgrids. Table 2 shows the operation benefit analysis before and after the microgrid alliance. It can be obtained that the total operation benefit after the MG alliance is 7930.2 yuan, which is 5338.8 yuan higher than the sum of the operation benefits of each MG operating independently, which is 2591.4 yuan. In addition, the operation benefit distribution of each microgrid is higher than that of the single operation, which confirms the feasibility of the cooperation.

[0179] Table 2: Profit distribution of each microgrid alliance before and after

[0180]

[0181] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions do not exceed the scope of the technical solution of the present invention, belong to the protection scope of the present invention.

Claims

1. A hybrid game-based optimization dispatching method for a microgrid alliance and a distribution network with shared energy storage, characterized in that: The following steps are involved: Step S1: Collect historical output data and historical load data of new energy sources in the region, and establish an active distribution network framework including an active distribution network ADN and a multi-microgrid alliance MGA including leased shared energy storage RSES; Step S2: Establish a hybrid game system mathematical model of the active distribution network framework and adopt a two-stage optimization strategy; Phase 1: Multiple microgrids conduct asymmetric Nash equilibrium games by leasing shared energy storage, and build an asymmetric Nash equilibrium model that takes into account energy storage to smooth power fluctuations and reduce costs; Phase 2: Multiple microgrid alliances operate together to form a master-slave game with the active distribution network, and a master-slave game model that takes into account both the master and the slave is constructed; Step S3: Use the fast non-dominated sorting genetic algorithm with elite retention strategy to solve the asymmetric Nash equilibrium game problem, and build a solution model for the master-slave game. Use the CPLEX solver to solve the lower model and obtain a hybrid game optimization scheduling scheme for the active distribution network framework. Step S4: Allocate the revenue of the multi-microgrid alliance based on the Shapley value.

2. The method for optimizing the scheduling of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: The active distribution network framework of step S1, which includes an active distribution network ADN and a multi-microgrid alliance MGA including leased shared energy storage RSES, is specifically described as follows: ADN includes new energy generation systems and conventional loads, and multiple microgrids MG are connected to ADN at different locations; The power interaction benefit of ADN and RSES is V RSES It is expressed as: Where: is the power sold and purchased by RSES and ADN during period t; is the ADN electricity purchase price; T is the number of time periods in the dispatching stage; Power interaction benefits of ADN and multi-microgrid alliances V MGA It is expressed as: Where: is the total power sold and purchased by MGA and ADN; n is the number of microgrids; is the power sold and purchased by microgrid i and ADN.

3. The method for optimizing the scheduling of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: In the hybrid game system mathematical model of step S2, the asymmetric Nash equilibrium game of multi-microgrids in phase 1 through leasing shared energy storage is specifically described as: Objective function 1: Minimize the mean square error of net energy storage load; Where: I1 is the mean square error of load after the microgrid uses leased shared energy storage; is the load power, energy storage charging capacity, wind power generation power, photovoltaic power generation and energy storage release capacity of the i-th microgrid in period t; is the load average; Objective function 2: Minimize the cost of energy storage application; I2=J rent +J RSES Where: I2 is the application cost of using leased shared energy storage in microgrids; a, b, c are the energy storage unit leasing cost, unit capacity leasing cost, and unit storage and release cost, respectively; J rent , J RSES They represent the energy storage leasing cost and the energy storage and release power cost respectively; They represent the power demand and energy demand of microgrid i respectively; The constraints of the microgrid include constant input and output power constraints and energy storage charging and discharging power and energy constraints.

4. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 3, characterized in that: The input-output power constant constraint is expressed as: In the formula, are the charging and discharging efficiencies of microgrid i in one cycle respectively; The energy storage charging and discharging power and energy constraints are expressed as: SOC min IN RSES ≤E′ RSES ≤SOC max IN RSES SOCIETY min P RSES ≤P′ RSES ≤SOC' max P RSES Where: E RESE and P RESE To operate a shared energy storage power station with energy and power capacity; E' RSES , P' RSES Shared energy storage power station power and energy; SOC min , SOC max and SOC' min , SOC' max The minimum and maximum charge states of the energy and power of the shared energy storage power station; Where: The maximum value of charging and discharging power of the shared energy storage power station; out , RSES It is the status bit of charge and discharge; and Δt are the energy capacity and time difference of the shared energy storage power station at time t, respectively.

5. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: In the hybrid game system mathematical model of step S2, the joint operation of the multi-microgrid alliance in stage 2 and the active distribution network to form a master-slave game are specifically described as follows: In the master-slave game model of ADN, MGA and RSES, the objective function of the main ADN is ADN It is expressed as: I ADN =V MGA +V RSES -B grid Where: V MGA 、V RSES , B grid are the power of the active distribution network and the multi-microgrid alliance, the interactive power benefit of shared energy storage, and the power purchase cost from the upper power grid, B grid It is expressed as: Where: t represents the unit time period; T is the number of time periods in the scheduling phase; The price of electricity from the upper power grid; The amount of electricity purchased by the active distribution network from the upper power grid; The constraints of the subject ADN include the following: Where: i,t,l , i,t,h They are the electricity purchase prices sold by the active distribution network to microgrid i during the peak and valley phases; ρ i,t,l,max , i,t,l,min are the upper and lower limits of the electricity price sold by the active distribution network to microgrid i; ρ i,t,h,max , i,t,h,min are the upper and lower limits of the electricity price purchased by the active distribution network from microgrid i; Where: is the maximum value of the average electricity price sold by ADN to microgrid i; i,t , i ' ,t is the average electricity price sold or purchased by ADN to microgrid i; The objective function U of the MGA model containing RSES MGA It is expressed as follows: IN MGA =I ADNΣ -J MTΣ Where: I ADNΣ is the total interactive power benefit between the multi-microgrid alliance and the active distribution network, J MTΣ The total power generation cost of the generator sets in the multi-microgrid alliance is: I MTΣ =J MTΣ,fuel +J MTΣ,o&m +J MTΣ,st +J MTΣ,em Where: J MTΣ,fuel , J MTΣ,o&m , J MTΣ,st , J MTΣ,em Respectively represent the fuel cost, operation and maintenance cost, startup and shutdown cost, and emission cost of the generator set; P MGi (t), EM MGi (t) is the output power and emission of the unit at time t; C fuel,p (t), C o&m,p (t), C em,p (t) is the unit price of fuel cost, operation and maintenance cost and emission cost; N sta 、N sh is the number of startup and shutdown; C sta,p , C sh,p The unit costs for startup and shutdown; The constraint function of the MGA model containing RSES is expressed as follows: Active power balance constraints of multi-microgrid alliance: Where: They are respectively represented as the interaction power between microgrid ij and microgrid ji during period t; represents the gas turbine power; and is the behavior variable of the multi-microgrid alliance purchasing and selling electricity from ADN. When it is 1, it means that the multi-microgrid alliance purchases electricity from ADN, and when it is 0, it means that the multi-microgrid alliance sells electricity to ADN. Shared energy storage charging and discharging balance constraints: Where: The discharge power and charging power of the shared energy storage station; Gas turbine output upper and lower limit constraints: Where: and are the upper and lower limits of gas turbine power; Power exchange constraints between microgrids: In the formula, Respectively represent the conversion coefficients between different energy forms; They represent the remaining energy capacity between microgrids and the maximum power exchange after comprehensive consideration.

6. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: The step S4 allocates the income of the multi-microgrid alliance based on the Shapley value, specifically: The multi-microgrid alliance adheres to the following principles: the overall operating cost shall not exceed the total cost of each microgrid when it operates independently; and the cost saved through cooperation, when allocated to each microgrid, must be lower than the cost of each microgrid operating independently; The Shapley value distribution method is expressed as: Where: x i is the remaining cooperation amount allocated to alliance member i; e(i) is the benefit obtained by alliance member i; s i is the sub-alliance composed of member i; ω(|s|) is the weight coefficient; e(s) and e(s / i) are the remainder of sub-alliance s and the remainder after excluding member i; |s| is the number of members in the alliance.

7. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: The step S3 uses a fast non-dominated sorting genetic algorithm with an elite retention strategy to solve the asymmetric Nash equilibrium game problem, specifically: After obtaining the uniformly distributed Pareto optimal solution through the fast non-dominated sorting genetic algorithm with elite retention strategy, the grey system membership function is established, the compromise solution is screened by the grey system membership function, and the optimal charging and discharging strategy of each microgrid is recorded.

8. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 7, characterized in that: The membership function of the grey system is expressed as: In the formula: μ(x) is the degree of membership; x is the independent variable; a1 and a2 are constants.

9. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 1, characterized in that: The master-slave game solution model of step S3 uses the CPLEX solver to solve the lower model, specifically: The optimization model of the master-slave game is expressed as: G={M;ρ j,t ;{ξ B ,x S ,i B ,i S };I ADN ;U MGA } Participant set: ADN and MGA with RSES constitute the participant set M, where ADN is the main participant and MGA with RSES is the slave participant; Decision variables: ADN decision variables are in the form of time period electricity price ρ j,t The decision variables of the MGA with RSES are the power of each microgrid to purchase and sell electricity to the main grid. And the power of shared energy storage to purchase and sell electricity to the main grid The master-slave equilibrium condition of the Stackelberg game is: In the formula, the superscript * indicates the equilibrium solution after solving; then The master-slave game is solved.

10. The method for optimizing the dispatch of a microgrid alliance and a distribution network with shared energy storage based on hybrid game according to claim 9, characterized in that: The upper layer ADN of the master-slave game optimization model is solved by the whale optimization algorithm, and the whale optimization algorithm is used to calculate and update the ADN time-of-use electricity price.

Images