An Island Microgrid Planning Method Based on Dynamic Games with Imperfect Information

Abstract

The present invention discloses a method for planning an island microgrid based on dynamic game with imperfect information. The steps are as follows: 1) Establish an optimization planning model for the to-be-built microgrid and an optimization operation model for the already-built microgrid; 2) Linearize the optimization planning model for the to-be-built microgrid and the optimization operation model for the already-built microgrid to obtain a linear optimization model for the to-be-built microgrid and a linear optimization model for the already-built microgrid; 3) Solve the linear optimization model for the to-be-built microgrid and the linear optimization model for the already-built microgrid to obtain a microgrid optimization planning scheme; 4) Establish a coordinated operation strategy model for the microgrid group; 5) Input the microgrid planning scheme into the coordinated operation strategy model for the microgrid group, and solve to obtain the optimal microgrid scheduling and electricity trading results. The present invention can be widely applied to the planning of island microgrid groups.

Description

Technical Field

[0001] The present invention relates to the field of island microgrid planning, and specifically to an island microgrid planning method based on dynamic game with imperfect information. Background Art

[0002] China has rich island resources, and many islands have extremely high economic and tourism value. Islands in hot areas such as the South China Sea are even in the front line of combat against the enemy for a long time. However, since many islands are far from the mainland, it is costly and difficult in engineering to supply power to the islands through the mainland power grid. Therefore, by building a microgrid to develop the rich wind, photovoltaic, and tidal current energy resources on the islands, reliable and economic power supply for the islands can be achieved. Compared with the onshore power grid, the new energy output power and load of the island microgrid have extremely strong seasonality, with greatly different powers in each season, and the daily peak-valley difference of the load is even larger, which poses challenges to the planning of the island microgrid. Summary of the Invention

[0003] The purpose of the present invention is to provide an island microgrid planning method based on dynamic game with imperfect information, including the following steps:

[0004] 1) Obtain the basic data.

[0005] The basic data includes the historical wind power output power data set P wind of the planned island, the historical photovoltaic output power data set P PV , the historical tidal current energy output power data set P TCF , the typical daily load curve P load , the typical daily electricity price curve The unit investment cost c i of equipment i, the unit penalty cost c em for carbon emission, the unit investment cost per unit length of the cable The unit penalty cost c LS for load shedding, the unit operation and maintenance cost of wind, light, and tidal power stations The unit operation and maintenance cost of battery energy storage The unit operation and maintenance cost of hydrogen energy storage The unit cost c f of diesel, the unit cost c CU for new energy curtailment, the unit operation and maintenance cost c DG of diesel generators, the cable length l gh between microgrids g and h, the discount rate r, the self-discharge rate Λ BS of battery energy storage, the self-discharge rate Λ HS of hydrogen energy storage, the microgrid scheduling time interval Δt, the maximum daily action times of hydrogen energy storage The minimum energy coefficient α BS of battery energy storage, the minimum energy coefficient α HS, the initial energy coefficient β of battery energy storage BS , the initial energy coefficient β of hydrogen energy storage HS , the carbon emission coefficient γ of diesel generators em , the equipment discount rate γ rv , the maximum load shedding coefficient ζ, the charge / discharge efficiency of battery energy storage the charge / discharge efficiency of hydrogen energy storage the ramp rate ρ of diesel generators DG , the power-capacity conversion coefficient τ of battery energy storage BS , the power-capacity conversion coefficient τ of hydrogen energy storage HS , the total number N of microgrids built on the island MG , the economic life L of equipment i i , the economic life L of the cable between microgrid g and microgrid h gh , the number N of scenarios in the new energy output power scenario set s , the number of inferences of the capacity of the g-th existing microgrid for the to-be-built microgrid the fuel consumption curve constant a of diesel generators DG and b DG .

[0006] 2) Determine the participant set N, action set A, payoff function set U, and inference set C in the game of island microgrids.

[0007] The participant set N = {0, 1, 2, …, N MG}}, where 0 represents the virtual participant "Nature" introduced to describe the uncertainty of the daily output power of wind / solar / tidal power stations. 1 to N MG - 1 represents the existing microgrids on the island. N MG is the to-be-built microgrid.

[0008] The action set A = × g∈N A g . Among them, the action set A g of participant g is as follows:

[0009]

[0010] In the formula, Ω N is the output power scenario set of wind, solar, and tidal power stations. is the output power action set of equipment i. is the load shedding action set. is the purchased power action set. is the sold power action set. is the plannable capacity action set of equipment i. The set of cable-plannable capacities for actions between microgrids g and h. The devices include new energy power stations, diesel generator equipment, battery energy storage equipment, and hydrogen energy storage equipment; h ∈ Ψ g represents the number of the microgrid connected to microgrid g through the cable, and Ψ g is the set of microgrids connected to microgrid g.

[0011] The set of payment functions U = {u1, u2, …, u NMG}. Among them, the element u g in the set of payment functions U is as follows:

[0012]

[0013] In the formula, C ED is the annual operating cost of the built microgrid, and C ING is the annual equivalent cost of the to-be-built microgrid.

[0014] The inference set C = × g＝1,2,…,NMG-1 Ω Cg . Ω Cg represents the inference set of the g-th built microgrid. The prior probability of each capacity in the inference set Ω Cg is ω g,c .

[0015] 3) Generate the set of daily output power scenarios for the island wind farm, photovoltaic power station, and tidal current power generation farm.

[0016] The steps to generate the set of daily output power scenarios for the wind farm, photovoltaic power station, and tidal current power generation farm include:

[0017] 3.1) Randomly select K daily curves X N from the set of historical daily output power curves of the wind farm / photovoltaic power station / tidal current power generation farm X = {X1, X2, …, X k} as the initial clustering centers of the corresponding power generation farms. k = 1, 2, …, K.

[0018] 3.2) Calculate the Euclidean distance ρ(X k' between each curve X k in the power data set X of the historical daily output power curves of the wind farm / photovoltaic power station / tidal current power generation farm and the clustering center curve X k' , X k ), that is:

[0019] ρ(X k' , X k ) = ||X k' , X k ||2, (k' = 1, 2,..., N) (3)

[0020] In the formula, ||X k’ , X k ||2 represents the two-norm between curve X k’ and X k .

[0021] 3.3) Take the category corresponding to the minimum Euclidean distance as the daily output power curve X k’ belongs to the category.

[0022] 3.4) For all curves in the same category, take the average value of the corresponding moment values, and use this curve as the updated clustering center. Determine whether the convergence condition is satisfied. If so, terminate the operation, and use the clustering center curve as the scenario in the microgrid optimization planning. Otherwise, return to step 3.2) to continue the iteration. The convergence condition is that the clustering center does not change for N max consecutive iterations.

[0023] 4) Establish an optimization planning model for the to-be-built microgrid and an optimized operation model for the existing microgrid.

[0024] The optimization objective of the to-be-built microgrid optimization planning model is min as follows:[[]]

[0025]

[0026] Among them, the annual equivalent investment cost of equipment The annual equivalent investment cost of cables Annual operation and maintenance cost Annual carbon emission penalty cost Annual load shedding penalty cost Annual power purchase cost Annual power sales revenue Annual equivalent residual revenue are respectively as follows:[[]]

[0027]

[0028]

[0029]

[0030]

[0031]

[0032]

[0033]

[0034]

[0035] In the formula, r represents the discount rate. L i is the economic life of device i. c i represents the unit investment cost of device i. Cap g,i is the planned capacity of device i in microgrid g. Φ g is the set of devices in microgrid g. L gh and l gh are the economic life and length of the cable between microgrids g and h. is the unit investment cost of the k-th type of cable. Cap gh is the planned capacity of the cable between microgrids g and h. Ψ g is the set of microgrids connected to microgrid g. D p is the number of days in a year. is the inferred quantity. The inferred quantity of the optimization planning model for the to-be-built microgrid N s is the number of scenarios. N t is the number of scheduling periods in a day. ω g,c is the probability of the c-th inference of the capacity of microgrid g for other microgrids. The probability ω of the optimization planning model for the to-be-built microgrid g,c = 1. ω s is the probability of scenario s. f i O&M (·) is the operation and maintenance cost function of device i. is the output power of device i. c em is the unit carbon emission penalty. γ em is the unit carbon emission of the diesel generator. a DG and b DG are the emission curve coefficients of the diesel generator. Δt is the microgrid scheduling time interval. is the rated capacity of the diesel generator. is the output power of the diesel generator. c LS is the unit load shedding cost. is the load shedding power. is the electricity price at time t. is the power purchase quantity from microgrid g to microgrid h, is the power sale quantity from microgrid g to microgrid h, γ rv is the residual coefficient.

[0036] Among them, the operation and maintenance cost function f of device i i O&M (·) includes the operation and maintenance cost function of the new energy power station the operation and maintenance cost function of the diesel generator the operation and maintenance cost function of the battery energy storage device and the operation and maintenance cost function of the hydrogen energy storage device The operation and maintenance cost functions are calculated as follows:

[0037]

[0038]

[0039]

[0040]

[0041] In the formula, is the output power of the new energy power station, is the unit operation and maintenance cost of the new energy power station, is the output power of the diesel generator, c f is the unit price of diesel, c DG is the unit operation and maintenance cost of the diesel generator, is the unit operation and maintenance cost of the battery energy storage, and are the charging and discharging powers of the battery energy storage in the t-th period under the inference c and scenario s in the microgrid g, is the unit operation and maintenance cost of the hydrogen energy storage, and are the charging and discharging powers of the hydrogen energy storage.

[0042] The constraint conditions of the to-be-built microgrid optimization planning model include the constraints related to the planning, the microgrid power flow constraints, the diesel generator operation constraints, the battery energy storage operation constraints, the hydrogen energy storage operation constraints, and the load shedding constraints.

[0043] Among them, the constraints related to the planning are as follows:

[0044] Cap g,DG +Cap g,BS >0 (17)

[0045]

[0046]

[0047] In the formula, Cap g,DG is the planned capacity of the diesel generator in the microgrid g. Cap g,BS is the planned capacity of the battery energy storage. Cap g,i is the planned capacity of the device i. is the upper limit of the planned capacity of the device i. Cap gh is the planned cable capacity between the microgrids g and h. is the upper limit of the planned cable capacity between the microgrids g and h.

[0048] The power flow constraints of the microgrid are as follows:

[0049]

[0050]

[0051]

[0052]

[0053] In the formula, is the purchased power from microgrid g to microgrid h during the t-th scheduling period in scenario s for inferring c. is the output power of device i. is the sold power from microgrid g to h. is the load power of microgrid g. is the load shedding power of microgrid g. Cap gh is the cable capacity between microgrids g and h.

[0054] The operating constraints of the diesel generator are as follows:

[0055]

[0056]

[0057] In the formula, and are the output powers of the diesel generators in the g-th microgrid during the t-th and t - 1-th scheduling periods in scenario s for inferring c. Cap g,DG is the planned capacity of the diesel generator in the g-th microgrid. ρ DG is the ramp rate of the diesel generator.

[0058] The operating constraints of the battery energy storage are as follows:

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066] In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in scenario s for inferring c. and are two-dimensional variables representing the charging state and discharging state of the battery energy storage. τ BS is the capacity-power conversion coefficient of the battery energy storage. ε is an infinitesimal constant. is the amount of electricity stored in the battery energy storage during the t-th scheduling period. is the amount of electricity stored in the battery energy storage during the (t - 1)-th scheduling period. Cap g,BS is the planned capacity of the battery energy storage. α BS is the lower limit coefficient of the battery energy storage electricity. β BS is the initial electricity coefficient of the battery energy storage. and are the initial electricity and final electricity of the battery energy storage within a scenario. Λ BS is the self-discharge rate of the battery energy storage. and are the charging efficiency and discharging efficiency of the battery energy storage.

[0067] The operating constraints of the hydrogen energy storage are as follows:

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077] In the formula, and are the charging power and discharging power of the hydrogen energy storage in the g-th microgrid during the t-th scheduling period in scenario s for inferring c. and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage. τ HS is the capacity-power conversion coefficient of the hydrogen energy storage. ε is an infinitesimal constant. The electricity stored in the hydrogen energy storage system during the t-th scheduling period. The electricity stored in the hydrogen energy storage system during the (t - 1)-th scheduling period. Cap g,HS The planned capacity of the hydrogen energy storage system. α HS The lower limit coefficient of the electricity in the hydrogen energy storage system. The initial electricity state of the first scenario of the hydrogen energy storage system. The initial electricity states of the s-th and (s - 1)-th scenarios of the hydrogen energy storage system. The final electricity state of the (s - 1)-th scenario of the hydrogen energy storage system. The final state of the last scenario. β HS The initial electricity coefficient of the hydrogen energy storage system. Λ HS The self-discharge rate of the battery energy storage system. and The charging efficiency and discharging efficiency of the battery energy storage system, The maximum number of times the charge and discharge state of the hydrogen energy storage system is allowed to change within a day. D p The number of days in a year. ω s-1 The probability of the (s - 1)-th scenario.

[0078] The load shedding constraint is as follows:

[0079]

[0080] In the formula, The load shedding power in the g-th microgrid at the t-th scheduling moment under the scenario s for inferring c. The load power. ζ is the maximum load shedding ratio coefficient.

[0081] The optimization objective of the established microgrid's optimal operation model is min as follows:

[0082]

[0083] In the formula, The annual operation and maintenance cost, The annual carbon emission penalty cost, The annual load shedding penalty cost, The annual electricity purchase cost, The annual electricity sales revenue.

[0084] The constraint conditions of the to-be-built microgrid optimization planning model include the microgrid power flow constraints (20)-(23), the diesel generator operation constraints (24)-(25), the battery energy storage operation constraints (26)-(32), the hydrogen energy storage operation constraints (33)-(41), and the load shedding constraint (42).

[0085] 5) Linearize the optimal planning model of the to-be-built microgrid and the optimal operation model of the existing microgrid to obtain the linear optimal model of the to-be-built microgrid and the linear optimal model of the existing microgrid.

[0086] The steps for linearizing the optimal planning model of the to-be-built microgrid and the optimal operation model of the existing microgrid include:

[0087] 5.1) Linearize the constraint conditions (24), (26)-(27), (33)-(34), and (39), where the non-linear terms in the constraint conditions are replaced as follows:

[0088]

[0089] In the formula, is the auxiliary variable for the operation constraint of the diesel generator in the g-th microgrid during the t-th scheduling period under the scenario s for inferring c, is the output power of the diesel generator, and are the auxiliary variables for the operation constraint of the battery energy storage, and are the two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the auxiliary variables for the operation constraint of the hydrogen energy storage, and y are the two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage.

[0090] 5.2) Add auxiliary constraint conditions to make the linearized model equivalent to the non-linear model.

[0091] Among them, the constraint condition (25) is replaced as follows:

[0092]

[0093]

[0094]

[0095]

[0096] In the formula, is the auxiliary variable for the operation constraint of the diesel generator in the g-th microgrid during the t-th scheduling period under the scenario s for inferring c, Cap g,DG is the planned capacity of the diesel generator, is the output power of the diesel generator, ρ DGis the ramp rate of the diesel generator, M DG is a positive number, and are auxiliary variables for the battery energy storage operation constraints.

[0097] Constraints (26)-(27), (33)-(34) are replaced as follows:

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110] In the formula, and are the charging power and discharging power of the battery energy storage in the gth microgrid during the tth scheduling period in the scenario s for inferring c, and are auxiliary variables for the battery energy storage operation constraints, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the charging power and discharging power of the hydrogen energy storage, and are auxiliary variables for the hydrogen energy storage operation constraints, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage, M BS and M HSAll are positive numbers.

[0111] The constraint condition (39) is replaced as follows:

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119] In the formula, is the auxiliary variable of the hydrogen energy storage operation constraint in the t-th and g-th microgrids under the scenario s of inferring c. is the maximum number of allowable charge and discharge state changes of the hydrogen energy storage within a day.

[0120] 5.3) Establish a linear optimization model for the to-be-built microgrid, that is:

[0121] obj.(4)-(16)

[0122]

[0123] Establish a linear optimization model for the existing microgrid, that is:

[0124] obj.(43), (7)-(16)

[0125]

[0126] 6) Solve the linear optimization model of the to-be-built microgrid and the linear optimization model of the existing microgrid to obtain the microgrid optimization planning scheme.

[0127] The methods for solving the linear optimization model of the to-be-built microgrid and the linear optimization model of the existing microgrid include the mixed integer programming method.

[0128] The microgrid optimization planning scheme includes the planning, operation, and power trading results of the to-be-built microgrid, and the operation and power trading results of the existing microgrid.

[0129] 7) Establish a coordinated operation strategy model for the microgrid group.

[0130] The steps for establishing a coordinated operation strategy model for the microgrid group include:

[0131] 7.1) Establish the optimization objective of the microgrid group operation strategy coordination model, i.e.:

[0132]

[0133] In the formula, E loss is the total amount of new energy curtailment and load shedding in the microgrid group, D p is the number of days in a year, N MG is the total number of microgrids in the microgrid group, is the total number of inferences of the g-th microgrid, N s is the total number of scenarios, N t is the number of scheduling periods in a day, ω g,c is the probability of the c-th inference of the g-th microgrid for the capacity of the to-be-built microgrid, ω s is the probability of the s-th scenario, is the new energy curtailment power of the g-th microgrid at the t-th scheduling period under the s-th scenario of the c-th inference, is the load shedding power, and Δt is the microgrid scheduling period interval.

[0134] 7.2) Establish the power flow constraint conditions of the microgrid group operation strategy coordination model, i.e.:

[0135]

[0136] In the formula, is the power purchased by the g-th microgrid from the h-th microgrid at the t-th scheduling period under the s-th scenario of the c-th inference, is the output power of the i-th device, is the new energy curtailment power of the g-th microgrid, is the power sold by the g-th microgrid to the h-th microgrid, is the load power, is the load shedding power, Ψ g is the set of microgrids connected to the g-th microgrid by cables, Φ g is the set of devices in the g-th microgrid.

[0137] 7.3) Establish the microgrid group operation strategy coordination model, i.e.:

[0138] obj.(70)

[0139]

[0140] 8) Input the microgrid planning scheme into the microgrid group operation strategy coordination model, and solve to obtain the optimal microgrid scheduling and power trading results.

[0141] The technical effects of the present invention are beyond doubt. The present invention studies the planning problem of island microgrid groups belonging to multiple stakeholders, takes into account the strong seasonality of new energy and load in island microgrids, and uses imperfect information dynamic game to describe the competition relationship between microgrids. The present invention can be widely applied to the planning of island microgrid groups, and can optimize the planning and analysis of microgrids based on the game relationship between microgrids. BRIEF DESCRIPTION OF THE DRAWINGS

[0142] Figure 1 is a flow chart of the present invention;

[0143] Figure 2 are the typical daily load curve and typical daily electricity price curve of the microgrid used in the simulation analysis; Figure 2 (a) is the typical daily load curve of MG1; Figure 2 (b) is the typical daily load curve of MG2 in different seasons; Figure 2 (c) is the typical daily load curve of MG3 in different seasons; Figure 2 (d) is the typical daily electricity price curve;

[0144] Figure 3 is the island new energy output power scenario used in the simulation analysis; Figure 3 (a) is the set of wind power output power scenarios; Figure 3 (b) is the set of photovoltaic output power scenarios; Figure 3 (c) is the set of tidal current energy output power scenarios. DETAILED DESCRIPTION OF THE INVENTION

[0145] The present invention will be further described below in conjunction with embodiments, but it should not be understood that the above-mentioned subject matter scope of the present invention is limited to the following embodiments. Without departing from the above-mentioned technical idea of the present invention, various substitutions and changes made according to ordinary technical knowledge and customary means in the art should be included within the protection scope of the present invention.

[0146] Embodiment 1:

[0147] Refer to Figure 1 、 Figure 2 、 Figure 3 , a method for planning an island microgrid based on imperfect information dynamic game, comprising the following steps:

[0148] 1) Obtain the basic data.

[0149] The basic data includes the historical wind power output power data set P wind of the planned island, the historical photovoltaic output power data set P PV , the historical tidal current energy output power data set P TCF , the typical daily load curve P load , the typical daily electricity price curve The unit investment cost c of device i i , the unit penalty cost c for carbon emissions em , the investment cost per unit length of the cable The unit penalty cost c for load shedding LS , the unit operation and maintenance cost of wind, solar, and tidal power stations The unit operation and maintenance cost of battery energy storage The unit operation and maintenance cost of hydrogen energy storage The unit cost c of diesel f , the unit cost c of new energy curtailment CU , the unit operation and maintenance cost c of diesel generators DG , the cable length l between microgrids g and h gh , the discount rate r, the self-discharge rate Λ of battery energy storage BS , the self-discharge rate Λ of hydrogen energy storage HS , the microgrid scheduling time interval Δt, the maximum daily operation times of hydrogen energy storage The minimum energy coefficient α of battery energy storage BS , the minimum energy coefficient α of hydrogen energy storage HS , the initial energy coefficient β of battery energy storage BS , the initial energy coefficient β of hydrogen energy storage HS , the carbon emission coefficient γ of diesel generators em , the device discount rate γ rv , the maximum load shedding coefficient ζ, the charge / discharge efficiency of battery energy storage The charge / discharge efficiency of hydrogen energy storage The ramp rate ρ of diesel generators DG , the power-capacity conversion coefficient τ of battery energy storage BS , the power-capacity conversion coefficient τ of hydrogen energy storage HS , the total number N of microgrids built on the island MG , the economic life L of device i i , the economic life L of the cable between microgrids g and h gh , the number of scenarios N in the new energy output power scenario set s , the inferred number of the capacity of the to-be-built microgrid by the g-th built microgrid The fuel consumption curve constant a of diesel generators DG 、b DG 。

[0150] 2) Determine the participant set N, action set A, payment function set U, and inference set C in the island microgrid game. The participant set N = {0, 1, 2, …, N MG},where 0 represents the virtual participant "Nature" introduced to describe the uncertainty of the daily output power of wind / solar / tidal power stations. 1 to N MG-1 represents the microgrid already built on the island. N MG is the microgrid to be built.

[0151] The action set A = × g∈N A g . Among them, the action set A of participant g g is as follows:

[0152]

[0153] In the formula, Ω N is the set of output power scenarios of wind, light, and tidal power stations. is the output power action set of device i. is the load shedding action set. is the purchased electricity power action set. is the sold electricity power action set. is the plannable capacity action set of device i. is the plannable capacity action set of the cable between microgrids g and h. The devices include new energy power stations, diesel generator sets, battery energy storage devices, and hydrogen energy storage devices. h ∈ Ψ g represents the number of the microgrid connected to microgrid g through the cable, and Ψ g is the set of microgrids connected to microgrid g.

[0154] The payment function set U = {u1, u2, …, u NMG}. Among them, the element u in the payment function set U g is as follows:

[0155]

[0156] In the formula, C ED is the annual operating cost of the already built microgrid, and C ING is the annual equivalent cost of the microgrid to be built.

[0157] The inference set C = × g＝1,2,…,NMG-1 Ω Cg . Ω Cg represents the inference set of the g-th already built microgrid. The prior probability of each capacity in the inference set Ω Cg is ω g,c .

[0158] 3) Generate the daily output power scenario set of the island's wind farm, photovoltaic power station, and tidal current power generation farm.

[0159] The steps to generate the daily output power scenario set of the wind farm, photovoltaic power station, and tidal current power generation farm include:

[0160] 3.1) From the set X = {X1, X2, …, X N} of historical daily output power curves of a wind farm / photovoltaic power station / tidal current power farm, randomly select K daily curves X k respectively as the initial clustering centers of the corresponding power farm. k = 1, 2, …, K.

[0161] 3.2) Calculate the Euclidean distance ρ(X k’ between each curve X in the set of power data of the historical daily output power curves of the wind farm / photovoltaic power station / tidal current power farm and the clustering center curve X k , that is: k' ,X k ) = ||X

[0162] ρ(X k' ,X k ) = ||X k' ,X k ||2, (k' = 1, 2,..., N) (3)

[0163] In the formula, ||X k’ ,X k ||2 represents the two-norm between the curves X k’ and X k .

[0164] 3.3) Take the category corresponding to the minimum Euclidean distance as the category to which the daily output power curve X k’ belongs.

[0165] 3.4) For all curves in the same category, take the average value of the corresponding moment values, and use this curve as the updated clustering center, and judge whether the convergence condition is satisfied. If so, terminate the operation, and use the clustering center curve as the scenario in the microgrid optimization planning. Otherwise, return to step 3.2) to continue the iteration. The convergence condition is that the clustering center does not change for N max consecutive iterations.

[0166] 4) Establish an optimization planning model for the to-be-built microgrid and an optimized operation model for the existing microgrid.

[0167] The optimization objective min of the to-be-built microgrid optimization planning model is as follows:

[0168]

[0169] Among them, the annual equivalent investment cost of equipment The annual equivalent investment cost of cables The annual operation and maintenance cost The annual carbon emission penalty cost The annual load shedding penalty cost The annual power purchase cost Annual electricity sales revenue Annual equivalent residual revenue Are respectively as follows:

[0170]

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178] In the formula, r represents the discount rate. L i Is the economic life of equipment i. c i Represents the unit investment cost of equipment i. Cap g,i Is the planned capacity of equipment i in microgrid g. Φ g Is the set of equipment in microgrid g. L gh And l gh Are the economic life and length of the cable between microgrids g and h. Is the unit investment cost of the kth type of cable. Cap gh Is the planned capacity of the cable between microgrids g and h. Ψ g Is the set of microgrids connected to microgrid g. D p Is the number of days in a year. Is the inference quantity. The inference quantity of the microgrid optimization planning model to be built N s Is the number of scenarios. N t Is the number of scheduling periods in a day. ω g,c Is the probability of the cth inference of the capacity of microgrid g for other microgrids. The probability ω of the microgrid optimization planning model to be built g,c = 1. ω s Is the probability of scenario s. f i O&M (·) is the operation and maintenance cost function of equipment i. Is the output power of equipment i. c em Is the unit carbon emission penalty. γ em Is the unit carbon emission of the diesel generator. a DG And b DGis the emission curve coefficient of the diesel generator. Δt is the microgrid scheduling time interval. is the rated capacity of the diesel generator. is the output power of the diesel generator. c LS is the unit load shedding cost. is the load shedding power. is the electricity price at time t. is the power purchase from microgrid g to microgrid h, is the power sale from microgrid g to microgrid h, γ rv is the residual coefficient.

[0179] Among them, the operation and maintenance cost function f of device i i O&M (·) includes the operation and maintenance cost function of the new energy power station the operation and maintenance cost function of the diesel generator the operation and maintenance cost function of the battery energy storage device and the operation and maintenance cost function of the hydrogen energy storage device are calculated respectively as follows:

[0180]

[0181]

[0182]

[0183]

[0184] In the formula, is the output power of the new energy power station, is the unit operation and maintenance cost of the new energy power station, is the output power of the diesel generator, c f is the unit price of diesel, c DG is the unit operation and maintenance cost of the diesel generator, is the unit operation and maintenance cost of the battery energy storage, and are the charging and discharging powers of the battery energy storage in the gth microgrid at the tth period under the inference c and scenario s, is the unit operation and maintenance cost of the hydrogen energy storage, and are the charging and discharging powers of the hydrogen energy storage.

[0185] The constraint conditions of the to-be-built microgrid optimization planning model include the constraints related to planning, microgrid power flow constraints, diesel generator operation constraints, battery energy storage operation constraints, hydrogen energy storage operation constraints, and load shedding constraints.

[0186] Among them, the constraints related to the planning are as follows:

[0187] Cap g,DG +Cap g,BS >0 (17)

[0188]

[0189]

[0190] In the formula, Cap g,DG is the planned capacity of the diesel generator in the microgrid g. Cap g,BS is the planned capacity of the battery energy storage. Cap g,i is the planned capacity of equipment i. is the upper limit of the planned capacity of equipment i. Cap gh is the planned cable capacity between the microgrids g and h. is the upper limit of the planned cable capacity between the microgrids g and h.

[0191] The microgrid power flow constraints are as follows:

[0192]

[0193]

[0194]

[0195]

[0196] In the formula, is the power purchase power from the microgrid g to the microgrid h at the t-th scheduling period in the scenario s of the inference c. is the output power of equipment i. is the power selling power from the microgrid g to h. is the load power of the microgrid g. is the load shedding power of the microgrid g. Cap gh is the cable capacity between the microgrids g and h.

[0197] The operating constraints of the diesel generator are as follows:

[0198]

[0199]

[0200] In the formula, and are the output powers of the diesel generator in the g-th microgrid at the t-th and t-1-th scheduling periods in the scenario s of the inference c. Cap g,DGis the planned capacity of the diesel generator in the g-th microgrid. ρ DG is the ramp rate of the diesel generator.

[0201] The operating constraints of the battery energy storage are as follows:

[0202]

[0203]

[0204]

[0205]

[0206]

[0207]

[0208]

[0209] In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period under the scenario s of inferring c. and are two-dimensional variables representing the charging state and discharging state of the battery energy storage. τ BS is the capacity-power conversion coefficient of the battery energy storage. ε is an infinitesimal constant. is the amount of electricity stored in the battery energy storage during the t-th scheduling period. is the amount of electricity stored in the battery energy storage during the (t - 1)-th scheduling period. Cap g,BS is the planned capacity of the battery energy storage. α BS is the lower limit coefficient of the battery energy storage power. β BS is the initial power coefficient of the battery energy storage. and are the initial power and final power of the battery energy storage in a scenario. Λ BS is the self-discharge rate of the battery energy storage. and are the charging efficiency and discharging efficiency of the battery energy storage.

[0210] The operating constraints of the hydrogen energy storage are as follows:

[0211]

[0212]

[0213]

[0214]

[0215]

[0216]

[0217]

[0218]

[0219]

[0220] In the formula, and are the charging power and discharging power of the hydrogen energy storage in the gth microgrid during the tth scheduling period in the scenario s for inferring c. and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage. τ HS is the capacity-power conversion coefficient of the hydrogen energy storage. ε is an infinitesimal constant. is the amount of electricity stored in the hydrogen energy storage during the tth scheduling period. is the amount of electricity stored in the hydrogen energy storage during the (t - 1)th scheduling period. Cap g,HS is the planned capacity of the hydrogen energy storage. α HS is the lower limit coefficient of the hydrogen energy storage electricity. is the initial electricity state of the hydrogen energy storage in the first scenario. are the initial electricity states of the hydrogen energy storage in the s-th and (s - 1)-th scenarios. is the final electricity state of the hydrogen energy storage in the (s - 1)-th scenario. is the final state of the last scenario. β HS is the initial electricity coefficient of the hydrogen energy storage. Λ HS is the self-discharge rate of the battery energy storage. and are the charging efficiency and discharging efficiency of the battery energy storage, is the maximum number of times the charge and discharge state of the hydrogen energy storage is allowed to change within a day. D p is the number of days in a year. ω s-1 is the probability of the (s - 1)-th scenario.

[0221] The load shedding constraint is as follows:

[0222]

[0223] In the formula, is the load shedding power in the gth microgrid at the tth scheduling moment in the scenario s for inferring c. is the load power. ζ is the maximum load shedding ratio coefficient.

[0224] The optimization objective min of the optimization operation model of the established microgrid is as follows: As follows:

[0225]

[0226] In the formula, is the annual operation and maintenance cost, is the annual carbon emission penalty cost, is the annual load shedding penalty cost, is the annual power purchase cost, is the annual power selling revenue.

[0227] The constraint conditions of the optimization planning model of the to-be-built microgrid include the microgrid power flow constraints (20)-(23), the diesel generator operation constraints (24)-(25), the battery energy storage operation constraints (26)-(32), the hydrogen energy storage operation constraints (33)-(41), and the load shedding constraints (42).

[0228] 5) Linearize the optimization planning model of the to-be-built microgrid and the optimization operation model of the established microgrid to obtain the linear optimization model of the to-be-built microgrid and the linear optimization model of the established microgrid.

[0229] The steps of linearizing the optimization planning model of the to-be-built microgrid and the optimization operation model of the established microgrid include:

[0230] 5.1) Linearize the constraint conditions (24), (26)-(27), (33)-(34), and (39). Among them, the non-linear terms in the constraint conditions are replaced as follows:

[0231]

[0232] In the formula, is the auxiliary variable of the diesel generator operation constraint in the g-th microgrid during the t-th scheduling period under the scenario s for inferring c, is the output power of the diesel generator, and are the auxiliary variables of the battery energy storage operation constraint, and are two-dimensional variables representing the charge state and discharge state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the auxiliary variables of the hydrogen energy storage operation constraint, and y are two-dimensional variables representing the charge state and discharge state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage.

[0233] 5.2) Add auxiliary constraint conditions to make the linearized model equivalent to the non-linear model.

[0234] Among them, the constraint condition (25) is replaced as follows:

[0235]

[0236]

[0237]

[0238]

[0239] In the formula, is the auxiliary variable for the operation constraint of the diesel generator in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, Cap g,DG is the planned capacity of the diesel generator, is the output power of the diesel generator, ρ DG is the ramp rate of the diesel generator, M DG is a positive number, and are the auxiliary variables for the operation constraint of the battery energy storage.

[0240] The constraint conditions (26)-(27), (33)-(34) are replaced as follows:

[0241]

[0242]

[0243]

[0244]

[0245]

[0246]

[0247]

[0248]

[0249]

[0250]

[0251]

[0252]

[0253] In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, and are auxiliary variables for the battery energy storage operation constraints, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the charging power and discharging power of the hydrogen energy storage, and are auxiliary variables for the hydrogen energy storage operation constraints, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage, M BS and M HS are both positive numbers.

[0254] The constraint condition (39) is replaced as follows:

[0255]

[0256]

[0257]

[0258]

[0259]

[0260]

[0261]

[0262] wherein, is the auxiliary variable for the hydrogen energy storage operation constraint in the g-th microgrid during the t-th in the scenario s for inferring c. is the maximum allowable number of charge and discharge state changes of the hydrogen energy storage within a day.

[0263] 5.3) Establish the linear optimization model of the to-be-built microgrid, that is:

[0264] obj.(4)-(16)

[0265]

[0266] Establish the linear optimization model of the built microgrid, that is:

[0267] obj.(43), (7)-(16)

[0268]

[0269] 6) Solve the linear optimization models of the to-be-built microgrid and the existing microgrid to obtain the optimal planning scheme for the microgrid.

[0270] The method for solving the linear optimization models of the to-be-built microgrid and the existing microgrid includes the mixed integer programming method.

[0271] The optimal planning scheme for the microgrid includes the planning, operation, and power trading results of the to-be-built microgrid, and the operation and power trading results of the existing microgrid.

[0272] 7) Establish a coordinated operation strategy model for the microgrid cluster.

[0273] The steps for establishing a coordinated operation strategy model for the microgrid cluster include:

[0274] 7.1) Establish the optimization objective of the coordinated operation strategy model for the microgrid cluster, that is:

[0275]

[0276] In the formula, E loss is the total amount of new energy abandonment and load shedding in the microgrid cluster, D p is the number of days in a year, N MG is the total number of microgrids in the microgrid cluster, is the total inference of the g-th microgrid, N s is the total number of scenarios, N t is the number of scheduling periods in a day, ω gc is the probability of the c-th inference of the g-th microgrid for the capacity of the to-be-built microgrid, ω s is the probability of the s-th scenario, is the new energy generator tripping power of the g-th microgrid at the t-th scheduling period under the scenario s of the inference c, is the load shedding power, and Δt is the microgrid scheduling period interval.

[0277] 7.2) Establish the power flow constraint conditions of the coordinated operation strategy model for the microgrid cluster, that is:

[0278]

[0279] In the formula, is the power purchased by the g-th microgrid from the h-th microgrid at the t-th scheduling period under the scenario s of the inference c, is the output power of the i-th device, is the new energy generator tripping power of the g-th microgrid, is the power sold from the g-th microgrid to the h-th microgrid, is the load power, is the load shedding power, Ψ g is the set of microgrids connected to the g-th microgrid via cables, Φ g is the set of devices in the g-th microgrid.

[0280] 7.3) Establish a coordinated operation strategy model for the microgrid group, that is:

[0281] obj.(70)

[0282]

[0283] 8) Input the microgrid planning scheme into the coordinated operation strategy model of the microgrid group, and solve to obtain the optimal microgrid scheduling and power trading results.

[0284] Embodiment 2:

[0285] See Figure 1 、 Figure 2 , a method for planning an island microgrid based on dynamic game with imperfect information, includes the following steps:

[0286] 1) Obtain the basic data;

[0287] 2) Determine the set of participants N, the set of actions A, the set of payoff functions U, and the set of inferences C in the game;

[0288] 2.1) The set of participants N = {0, 1, 2,..., N MG}, where 0 represents the virtual participant "Nature" introduced to describe the uncertainty of the daily output power of the wind / solar / tidal power station, and 1 to N MG -1 represent the microgrids already built on the island, and N MG is the microgrid to be built;

[0289] 2.2) The set of actions A: According to the different types of participants, its set of actions A g is also different. The set of actions of the game A = × g∈N A g .

[0290] 2.3) The set of payoff functions U: For the built microgrid and the microgrid to be built, the set of payoff functions U = {u1, u2,..., u NMG}, and the payoff function of the participant "Nature" is not considered.

[0291] 2.4) Inference set C: For the already built microgrid, the equipment capacity of the to-be-built microgrid is unknown. Therefore, the already built microgrid needs to infer the location capacity to determine its own operation mode after the access of the to-be-built microgrid; according to prior knowledge, the inference set of the g-th already built microgrid is Ω Cg , and the prior probability of each capacity in the set is ω g,c ; therefore, the inference set C = × g＝1,2,…,NMG-1 Ω Cg .

[0292] 3) Generate the daily output power scenario set of the island wind farm, photovoltaic power station, and tidal current power generation farm;

[0293] 4) Establish the optimization planning model of the to-be-built microgrid and the optimization operation model of the already built microgrid;

[0294] The optimization planning model of the to-be-built microgrid includes the following optimization objectives and constraints:

[0295] 4.1) Optimization planning model of the to-be-built microgrid

[0296] 4.1.1) To minimize the annual equivalent cost of the to-be-built microgrid, the optimization objective of the to-be-built microgrid is:

[0297] 5) Linearize the two types of optimization models constructed and solve them based on the mixed-integer programming algorithm;

[0298] The linear optimization models of each microgrid can all be solved by the mixed-integer programming algorithm to obtain the planning, operation, and power trading results of the to-be-built microgrid, as well as the operation and power trading results of the already built microgrid.

[0299] 6) Under the microgrid planning scheme, coordinate the operation and power trading strategies obtained by each microgrid to obtain reasonable microgrid scheduling and power trading results;

[0300] 6.4) Substitute the planning scheme solved by equations (68) and (69), as well as the optimization operation results of the diesel generator, battery energy storage, and hydrogen energy storage, into the optimization model given by equation (72), and take the above optimization results as fixed values. Solve the values of other optimization variables through the mixed-integer programming algorithm to obtain the final microgrid power trading, load shedding, and new energy curtailment strategies.

[0301] Embodiment 3:

[0302] On the basis of considering the historical output power data of the wind, light, and tidal power stations, when two microgrids belonging to different interest subjects have been built on a large island, another interest subject plans the third microgrid and connects it to the first two microgrids through cables, so as to verify the island microgrid planning method based on imperfect information dynamic game. The specific implementation steps are as follows:

[0303] 1) Obtain the basic data;

[0304] The equipment capacities and peak loads of the two existing microgrids (named MG1 and MG2), and the peak load of the to-be-built microgrid (named MG3) are shown in Table 1:

[0305] Table 1 Microgrid Equipment Capacities and Peak Loads in the Test System

[0306] MG1 MG2 MG3 Wind farm (kW) 270 180 — Photovoltaic power station (kW) 680 560 — Tidal current power generation farm (kW) 0 0 — Diesel generator (kW) 400 300 — Battery energy storage (kWh) 800 650 — Hydrogen energy storage (kWh) 0 0 — Peak load (kW) 699 630 702

[0307] The historical wind power output power data set P of the planned island wind , the historical photovoltaic output power data set P PV and the historical tidal current energy output power data set P TCF are all collected from Orkney Island in Scotland. The sampling time is from 1996 to 2005, and the sampling interval is 1 hour; the typical daily load curve P of the microgrid load and the typical daily electricity price curve are shown by Figure 2 , where the loads of MG2 and MG3 are seasonal, while the load of MG1 is not seasonal; the unit investment cost c of equipment i i is shown in Table 2, the unit penalty cost c for carbon emissions em = 30 $ / t, the unit length investment cost of the cable The unit penalty cost c for load shedding LS = 710 $ / MWh, the unit operation and maintenance cost of wind, light, and tidal power stations The unit operation and maintenance cost of battery energy storage The unit operation and maintenance cost of hydrogen energy storage The unit cost c of diesel f = 0.43 $ / L, the unit cost c for new energy curtailment CU = 110 $ / MWh, the unit operation and maintenance cost c of diesel generators DG = 0.05 $ / h, the cable length l between microgrids g and h gh = 2000 m, the discount rate r = 0.07, the self-discharge rate Λ of battery energy storage BS = 0.04, the self-discharge rate Λ of hydrogen energy storage HS = 0.001, the microgrid scheduling time interval Δt = 1 h, the maximum daily operation times of hydrogen energy storage The minimum energy coefficient α of battery energy storage BS = 0.1, the minimum energy coefficient α of hydrogen energy storage HS = 0.1, the initial energy coefficient β of battery energy storage BS = 0.5, the initial energy coefficient β of hydrogen energy storage HS = 0.5, the carbon emission coefficient γ of diesel generatorsem = 0.0028t / L, equipment discount rate γ rv = 0.05, maximum load shedding coefficient ζ = 0.02, battery energy storage charge / discharge efficiency Hydrogen energy storage charge / discharge efficiency Diesel generator ramp rate ρ DG = 0.9, battery energy storage power-capacity conversion coefficient τ BS = 0.4, hydrogen energy storage power-capacity conversion coefficient τ HS = 0.2, total number of microgrids N built on the island MG = 3, economic life L of equipment i i = 20 years, economic life L of the cable between microgrid g and microgrid h gh = 20 years, number of scenarios N in the new energy output power scenario set s = 12, number of inferences of the capacity of the g-th existing microgrid for the to-be-built microgrid Diesel generator fuel consumption curve constant a DG = 246L / MWh, b DG = 84.15L / MWh.

[0308] Table 2 Unit investment cost of microgrid access equipment

[0309] Unit investment cost Wind farm 2240 $ / kW Photovoltaic power station 1000 $ / kW Tidal current power generation farm 3500 $ / kW Diesel generator 210 $ / kW Battery energy storage 342 $ / kWh Hydrogen energy storage 1570 $ / kWh

[0310] 2) Determine the set of participants N, the set of actions A, the set of payoff functions U, and the set of inferences C in the game;

[0311] 2.1) The set of participants N = {0, 1, 2, 3}, where 0 represents the virtual participant "Nature" introduced to describe the uncertainty of the daily output power of the wind / solar / tidal power station, 1 - 2 represent the existing microgrids on the island, and 3 is the to-be-built microgrid;

[0312] 2.2) The set of actions A: According to the different types of participants, its set of actions A g is also different. The set of actions A of the game = × g∈N A g .

[0313]

[0314] In the formula, Ω N is the set of wind, solar, and tidal power station output power scenarios, is the output power action set of equipment i, is the load shedding action set, is the power purchase action set, is the power sale action set, The set of schedulable capacity actions for device i The set of schedulable capacity actions for the cable between microgrids g and h; h ∈ Ψ g Denotes the number of the microgrid connected to microgrid g through the cable, and Ψ g Is the set of microgrids connected to microgrid g;

[0315] 2.3) Set of payment functions U: For the existing microgrids and the to-be-built microgrids, their payment functions u g Can be expressed by the following formulas respectively. The set of payment functions U = {u1, u2, u3}, and the payment function of the participant "Nature" is not considered:

[0316]

[0317] In the formula, C ED Is the annual operating cost of the existing microgrid, and C ING Is the annual equivalent cost of the to-be-built microgrid;

[0318] 2.4) Inference set C: For the existing microgrids, the equipment capacity of the to-be-built microgrids is unknown. Therefore, the existing microgrids need to make inferences about the location capacity to determine their own operating modes after the connection of the to-be-built microgrids; According to prior knowledge, the inference set of the g-th existing microgrid is Ω Cg , and the prior probability of each capacity in the set is ω g,c ; Therefore, the inference set C = × g＝1,2 Ω Cg ;

[0319] 2.5) Public information and confidential information of existing microgrids and to-be-built microgrids: According to the different types of MG1, MG2, and MG3, their public information and confidential information are shown in Table 3:

[0320] Table 3 Public Information and Confidential Information of MG1, MG2, and MG3

[0321]

[0322] 3) Generate the set of daily output power scenarios for the island wind farm, photovoltaic power station, and tidal current power generation farm;

[0323] The steps to generate the set of daily output power scenarios include:

[0324] 3.1) Randomly select 12 daily curves X 3653 from the set of historical daily output power curves of the wind farm / photovoltaic power station / tidal current power generation farm X = {X1, X2,..., X k} as the initial clustering centers for the corresponding power generation farms (k = 1, 2,..., 12);

[0325] 3.2) Calculate the Euclidean distance ρ(X k’ between each curve X in the power data set X of the historical daily output power curve of the wind farm / photovoltaic power station / tidal current power generation farm k and the clustering center curve X k' , that is: k ) as follows:

[0326] ρ(X k' ,X k ) = ||X k' ,X k ||2, (k' = 1, 2,..., 3653) (3)

[0327] where ||X k’ ,X k ||2 represents the two-norm between curve X k’ and X k ;

[0328] 3.3) Take the category corresponding to the minimum Euclidean distance as the category to which the daily output power curve X k’ belongs;

[0329] 3.4) For all curves in the same category, take the average value of the corresponding moment values, and use this curve as the updated clustering center, and judge whether the convergence condition is satisfied; if so, terminate the operation, and use the clustering center curve as the scenario in the microgrid optimization planning; otherwise, return to step 3.2) to continue the iteration; the convergence condition is that the clustering center does not change for 5 consecutive iterations. The output power scenario sets of the wind, light, and tidal power stations obtained after clustering are as Figure 2 shown.

[0330] 4) Establish an optimization planning model for the to-be-built microgrid and an optimization operation model for the existing microgrid;

[0331] The optimization planning model for the to-be-built microgrid includes the following optimization objectives and constraints:

[0332] 4.1) Optimization planning model for the to-be-built microgrid

[0333] 4.1.1) In order to minimize the annual equivalent cost of the to-be-built microgrid, the optimization objective of the to-be-built microgrid is:

[0334]

[0335] where is the annual equivalent investment cost of the equipment, is the annual equivalent investment cost of the cable, is the annual operation and maintenance cost, is the annual carbon emission penalty cost, is the annual load shedding penalty cost, is the annual electricity purchase cost, is the annual electricity sales revenue, is the annual equivalent residual revenue; each cost / revenue in the optimization objective is calculated by the following formulas respectively:

[0336]

[0337]

[0338]

[0339]

[0340]

[0341]

[0342]

[0343]

[0344] In the formula, r = 0.07 represents the discount rate, L i = 20 years is the economic life of equipment i, c i represents the unit investment cost of equipment i (as shown in Table 2), Cap g,i is the planned capacity of equipment i in microgrid g, Φ g = {wind farm, photovoltaic power station, tidal current power generation farm, diesel generator, battery energy storage, hydrogen energy storage} is the equipment set of microgrid g, L gh = 20 years and l gh = 2000 m is the economic life and length of the cable between microgrids g and h, is the unit investment cost of the cable, Cap gh is the planned capacity of the cable between microgrids g and h, Ψ g is the set of microgrids connected to microgrid g, D p = 365 is the number of days in a year, is the inferred quantity (fixed to 1 for the to-be-built microgrid), N s = 12 is the number of scenarios, N t = 24 is the number of dispatching periods in a day, ω g,c is the probability of the c-th inference of the capacity of microgrid g for other microgrids (fixed to 1 for the to-be-built microgrid), ω s is the probability of scenario s, f i O&M (·) is the operation and maintenance cost function of equipment i, which can be calculated by formulas (13)-(16), is the output power of equipment i, cem = 30 $ / t is the unit carbon emission penalty, γ em = 0.0028 t / L is the unit carbon emission of the diesel generator, a DG = 246 L / MWh and b DG = 84.15 L / MWh are the emission curve coefficients of the diesel generator, Δt = 1 h is the microgrid scheduling time interval, is the rated capacity of the diesel generator, c LS = 710 $ / MWh is the unit load shedding cost, is the load shedding power, is the electricity price at time t, as Figure 2 (d) shows, is the power purchase from microgrid g to microgrid h, is the power sale from microgrid g to microgrid h, γ rv = 0.05 is the residual coefficient;

[0345] The operation and maintenance cost functions of new energy power stations, diesel generators, battery energy storage, and hydrogen energy storage are calculated as follows:

[0346]

[0347]

[0348]

[0349]

[0350] In the formula, is the output power of the new energy power station, is the unit operation and maintenance cost of the new energy power station, is the output power of the diesel generator, c f = 0.43 $ / L is the unit price of diesel, c DG = 0.05 $ / h is the unit operation and maintenance cost of the diesel generator, is the unit operation and maintenance cost of the battery energy storage, and are the charging and discharging powers of the battery energy storage in microgrid g at the t-th time period under the inference c and scenario s, is the unit operation and maintenance cost of the hydrogen energy storage, and are the charging and discharging powers of the hydrogen energy storage;

[0351] 4.1.2) The optimization planning model of the to-be-built microgrid includes the following planning-related constraint conditions:

[0352] Cap g,DG + Cap g,BS>0 (17)

[0353]

[0354]

[0355] In the formula, Cap g,DG is the planned capacity of the diesel generator in microgrid g, Cap g,BS is the planned capacity of the battery energy storage, Cap g,i is the planned capacity of device i, is the upper limit of the planned capacity of device i, Cap gh is the planned capacity of the cable between microgrids g and h, is the upper limit of the planned capacity of the cable between microgrids g and h;

[0356] 4.1.3) The optimization planning model of the to-be-built microgrid includes the following constraints related to the power flow between microgrids:

[0357]

[0358]

[0359]

[0360]

[0361] In the formula, is the power purchase from microgrid g to microgrid h during the t-th scheduling period under scenario s of inferring c, is the output power of device i, is the power sold from microgrid g to h, is the load power of microgrid g, is the load shedding power of microgrid g, Cap gh is the cable capacity between microgrids g and h;

[0362] 4.1.4) The optimization planning model of the to-be-built microgrid includes the following constraints related to the operation of the diesel generator:

[0363]

[0364]

[0365] In the formula, and are the output powers of the diesel generators in the g-th microgrid during the t-th and t-1-th scheduling periods under scenario s of inferring c, Cap g,DG is the planned capacity of the diesel generator in the g-th microgrid, ρ DG= 0.9 is the ramp rate of the diesel generator;

[0366] 4.1.5) The optimization planning model of the to-be-built microgrid includes the following constraint conditions related to the operation of the battery energy storage:

[0367]

[0368]

[0369]

[0370]

[0371]

[0372]

[0373]

[0374] In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, τ BS = 0.4 is the capacity-power conversion coefficient of the battery energy storage, ε = 1×10 -8 is an infinitesimal constant, is the electricity stored in the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, α BS = 0.1 is the lower limit coefficient of the battery energy storage electricity, β BS = 0.5 is the initial electricity coefficient of the battery energy storage, and are the initial electricity and final electricity of the battery energy storage within a scenario, Λ BS = 0.04 is the self-discharge rate of the battery energy storage, and are the charging efficiency and discharging efficiency of the battery energy storage;

[0375] 4.1.6) The optimization planning model of the to-be-built microgrid includes the following constraint conditions related to the operation of the hydrogen energy storage:

[0376]

[0377]

[0378]

[0379]

[0380]

[0381]

[0382]

[0383]

[0384]

[0385] In the formula, and are the charging power and discharging power of the hydrogen energy storage in the gth microgrid during the tth scheduling period in the scenario s for inferring c, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, τ HS = 0.2 is the capacity-power conversion coefficient of the hydrogen energy storage, ε = 1×10 -8 is an infinitesimal constant, is the electric quantity stored in the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage, α HS = 0.1 is the lower limit coefficient of the hydrogen energy storage electric quantity, is the initial electric quantity state of the first scenario of the hydrogen energy storage, is the final state of the last scenario, β HS is the initial electric quantity coefficient of the hydrogen energy storage, Λ HS is the self-discharge rate of the battery energy storage, and are the charging efficiency and discharging efficiency of the battery energy storage, is the maximum number of times the charging and discharging state of the hydrogen energy storage is allowed to change within a day, D p = 365 is the number of days in a year, ω s-1 is the probability of the (s - 1)th scenario;

[0386] 4.1.7) The optimization planning model of the to-be-built microgrid includes the following load-shedding related constraint conditions:

[0387]

[0388] In the formula, is the load-shedding power in the gth microgrid during the tth in the scenario s for inferring c, is the load power, ζ = 0.02 is the maximum load-shedding ratio coefficient;

[0389] 4.2) The optimization operation model of the existing microgrid

[0390] 4.2.1) To minimize the annual operating cost of the established microgrid, the optimization objective of the established microgrid is as follows:

[0391]

[0392] where, is the annual operation and maintenance cost, is the annual carbon emission penalty cost, is the annual load shedding penalty cost, is the annual power purchase cost, is the annual power selling revenue. The calculation methods of the above costs or revenues are shown in (7)-(11);

[0393] 4.2.2) In the optimization planning model of the to-be-built microgrid, it includes the microgrid power flow constraints (20)-(23), the diesel generator operation constraints (24)-(25), the battery energy storage operation constraints (26)-(32), the hydrogen energy storage operation constraints (33)-(41), and the load shedding constraints (42).

[0394] 5) Linearize the two types of optimization models constructed and solve them based on the mixed integer programming algorithm;

[0395] 5.1) In the optimization model constructed in step 4), the formulas (24), (26)-(27), (33)-(34), and (39) are non-linear constraint conditions, which need to be linearized. The non-linear terms in the constraint conditions are replaced as shown by the formula (44):

[0396]

[0397] where, is the auxiliary variable of the diesel generator operation constraint in the g-th microgrid during the t-th scheduling period under the scenario s for inferring c, is the output power of the diesel generator, and are the auxiliary variables of the battery energy storage operation constraint, and are the two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the auxiliary variables of the hydrogen energy storage operation constraint, and y are the two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage;

[0398] 5.2) After linearizing the non - linear constraint conditions in the original model based on the variable substitution in formula (44), the following auxiliary constraint conditions need to be added to the original model to make the linearized model equivalent to the non - linear model;

[0399] Constraint condition (25) can be replaced by the following formula:

[0400]

[0401]

[0402]

[0403]

[0404] In the formula, is the auxiliary variable for the operation constraint of the diesel generator in the g - th micro - grid during the t - th scheduling period in the scenario s for inferring c, Cap g,DG is the planned capacity of the diesel generator, is the output power of the diesel generator, ρ DG = 0.9 is the ramp rate of the diesel generator, M DG is a sufficiently large positive number, and are the auxiliary variables for the operation constraint of the battery energy storage;

[0405] Constraint conditions (26) - (27), (33) - (34) can be replaced by the following formula:

[0406]

[0407]

[0408]

[0409]

[0410]

[0411]

[0412]

[0413]

[0414]

[0415]

[0416]

[0417]

[0418] In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, and are auxiliary variables for the operation constraints of the battery energy storage, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the charging power and discharging power of the hydrogen energy storage, and are auxiliary variables for the operation constraints of the hydrogen energy storage, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage, M BS and M HS are both sufficiently large positive numbers;

[0419] Constraint condition (39) can be replaced by the following formula:

[0420]

[0421]

[0422]

[0423]

[0424]

[0425]

[0426]

[0427] In the formula, and are auxiliary variables for the operation constraints of the hydrogen energy storage in the g-th microgrid during the t-th in the scenario s for inferring c; is the maximum number of allowable changes in the charging and discharging state of the hydrogen energy storage within a day;

[0428] 6) Under the microgrid planning scheme, coordinate the operation and power trading strategies obtained from each microgrid to obtain reasonable microgrid scheduling and power trading results;

[0429] 6.1) The optimization objectives of the microgrid group operation strategy coordination model are as follows:

[0430]

[0431] In the formula, E loss is the total amount of new energy abandonment and load shedding in the microgrid group, D p = 365 is the number of days in a year, N MG = 3 is the total number of microgrids in the microgrid group, is the total inference of the g-th microgrid (let ), N s is the total number of scenarios, N t = 24 is the number of scheduling periods in a day, ω g,c is the probability of the c-th inference of the g-th microgrid for the capacity of the to-be-built microgrid, ω s is the probability of the s-th scenario, is the new energy curtailment power of the g-th microgrid at the t-th scheduling period under the scenario s of the inference c, is the load shedding power, Δt = 1h is the microgrid scheduling period interval;

[0432] 6.2) The power flow constraint conditions in the microgrid group operation strategy coordination model are:

[0433]

[0434] In the formula, is the power purchased by the g-th microgrid from the h-th microgrid at the t-th scheduling period under the scenario s of the inference c, is the output power of the i-th device, is the new energy curtailment power of the g-th microgrid, is the power sold by the g-th microgrid to the h-th microgrid, is the load power, is the load shedding power, Ψ g is the set of microgrids connected to the g-th microgrid by cables, Φ g is the set of devices in the g-th microgrid;

[0435] 6.3) Based on the above objective function and power flow constraints, the microgrid group operation strategy coordination model can be summarized as the following formula:

[0436] obj.(70)

[0437]

[0438] 6.4) Substitute the planning solutions obtained from equations (68) and (69), as well as the optimal operation results of the diesel generator, battery energy storage, and hydrogen energy storage, into the optimization model given by equation (72). Treat the above optimization results as fixed values and solve for the values of other optimization variables through the mixed-integer programming algorithm to obtain the final microgrid power trading, load shedding, and new energy curtailment strategies. The cost lists of MG1, MG2, and MG3 obtained hereby are shown in Tables 4, 5, and 6 respectively.

[0439] Table 4 Cost List of MG1

[0440]

[0441] Table 5 Cost List of MG2

[0442]

[0443] Table 6 Cost List of MG3

[0444]

Claims

1. A method for planning a microgrid on an island based on dynamic games with imperfect information, characterized in that It includes the following steps: 1) Obtain basic data; 2) Determine the set of participants N, the set of actions A, the set of payment functions U, and the set of inferences C in the game of island microgrid; 3) Generate the set of daily output power scenarios of the island wind farm, photovoltaic power station, and tidal current energy power station; 4) Establish the optimization planning model of the to-be-built microgrid and the optimization operation model of the existing microgrid; 5) Linearize the optimization planning model of the to-be-built microgrid and the optimization operation model of the existing microgrid to obtain the linear optimization model of the to-be-built microgrid and the linear optimization model of the existing microgrid; 6) Solve the linear optimization model of the to-be-built microgrid and the linear optimization model of the existing microgrid to obtain the microgrid optimization planning scheme; 7) Establish the operation strategy coordination model of the microgrid group; 8) Input the microgrid planning scheme into the operation strategy coordination model of the microgrid group, and solve to obtain the optimal microgrid scheduling and power trading results.

2. The method for planning an island microgrid based on dynamic game with imperfect information according to claim 1, wherein The basic data includes the historical wind power output power data set P of the planned island wind , the historical photovoltaic output power data set P PV , the historical tidal current energy output power data set P TCF , the typical daily load curve P load , the typical daily electricity price curve The unit investment cost c of equipment i i , the unit penalty cost c for carbon emissions em , the unit investment cost per unit length of cable The unit penalty cost c for load shedding LS , the unit operation and maintenance cost of wind, light, and tidal power stations The unit operation and maintenance cost of battery energy storage The unit operation and maintenance cost of hydrogen energy storage The unit cost c of diesel f , the unit cost c for new energy curtailment CU , the unit operation and maintenance cost c of diesel generators DG , the cable length l between microgrids g and h gh , the discount rate r, the self-discharge rate Λ of battery energy storage BS , the self-discharge rate Λ of hydrogen energy storage HS , the microgrid scheduling time interval Δt, the maximum daily operation times of hydrogen energy storage The minimum energy coefficient α of battery energy storage BS , the minimum energy coefficient α of hydrogen energy storage HS , the initial energy coefficient β of battery energy storage BS , the initial energy coefficient β of hydrogen energy storage HS , the carbon emission coefficient γ of diesel generators em , the equipment discount rate γ rv , the maximum load shedding coefficient ζ, the charge / discharge efficiency of battery energy storage The charge / discharge efficiency of hydrogen energy storage The ramp rate ρ of diesel generators DG , the power-capacity conversion coefficient τ of battery energy storage BS , the power-capacity conversion coefficient τ of hydrogen energy storage HS , the total number N of microgrids built on the island MG , the economic life L of equipment i i , the economic life L of the cable between microgrids g and h gh , the number of scenarios N in the new energy output power scenario set s , the inferred number of the capacity of the to-be-built microgrid by the g-th built microgrid The fuel consumption curve constant a of diesel generators DG 、b DG .

3. The method for planning an island microgrid based on dynamic game with imperfect information according to claim 1, wherein The set of participants \(N = \{0, 1, 2, \ldots, N\}\) MG}, where 0 represents the virtual participant "Nature" introduced to describe the uncertainty of the daily output power of wind / solar / tidal power plants; 1 to \(N - 1\) MG represent the microgrids already built on the island; \(N\) MG is the microgrid to be built. The set of actions \(A = \times\) g∈N A g ; among which, the set of actions \(A\) of participant \(g\) g is as follows: where, Ω N is the set of output power scenarios of wind, solar, and tidal power stations; is the set of output power actions of device i; is the set of load shedding actions; is the set of power purchase actions; is the set of power selling actions; is the set of plannable capacity actions of device i; is the set of plannable capacity actions of the cable between microgrids g and h; the devices include new energy power stations, diesel generator sets, battery energy storage devices, and hydrogen energy storage devices; h ∈ Ψ g represents the microgrid number connected to microgrid g through a cable, and Ψ g is the set of microgrids connected to microgrid g; The set of payment functions \(U = \{u_1, u_2, \ldots, u\}\) NMG ; where the element \(u\) in the set of payment functions \(U\) g is as follows: Where, C ED is the annual operating cost of the built microgrid, and C ING is the annual equivalent cost of the to-be-built microgrid; The inference set $C = \times$ g＝1,2,…,NMG-1 $\Omega$ Cg ; $\Omega$ Cg represents the inference set of the $g$-th established microgrid; the prior probability of each capacity in the inference set $\Omega$ Cg is $\omega$ g,c .

4. The method for planning an island microgrid based on imperfect information dynamic game according to claim 1, characterized in that, The steps of generating the set of daily output power scenarios of the wind farm, photovoltaic power station, and tidal current energy power station include: 1) From the set X = {X1, X2, …, X N} of historical daily output power curves of a wind farm / photovoltaic power station / tidal current power generation farm, randomly select K daily curves X k respectively as the initial clustering centers of the corresponding power generation farm; k = 1, 2, …, K; 2) Calculate the Euclidean distance ρ(X k' of each curve X in the power data set X of the historical daily output power curve of the wind farm / photovoltaic power station / tidal current power generation farm k from the cluster center curve X k' ,X k ), that is: ρ(X k' ,X k )=||X k' ,X k ||2,(k'=1,2,...,N) (3) where ||X k’ ,X k ||2 represents the two - norm between the curves X k’ and X k ; 3) The category corresponding to the minimum Euclidean distance is the daily output power curve X k’ the category to which it belongs; 4) For all curves in the same category, calculate the mean value of the corresponding moment values, and use this curve as the updated clustering center. Then, determine whether the convergence condition is met. If so, terminate the operation and use the clustering center curve as the scenario in the microgrid optimal planning. Otherwise, return to step 2) to continue the iteration. The convergence condition is that the clustering center does not change for N max consecutive iterations.

5. The method for planning a microgrid on an island based on dynamic game with imperfect information according to claim 1, wherein The optimization objectives of the to-be-built microgrid optimization planning model are as follows: Among them, the annual equivalent investment cost of the equipment The annual equivalent investment cost of the cable The annual operation and maintenance cost The annual carbon emission penalty cost The annual load shedding penalty cost The annual electricity purchase cost The annual electricity selling revenue The annual equivalent residual revenue They are respectively as follows: where r represents the discount rate; L i is the economic life of equipment i; c i represents the unit investment cost of equipment i; Cap g,i is the planned capacity of equipment i in microgrid g; Φ g is the set of equipment in microgrid g; L gh and l gh are the economic life and length of the cable between microgrids g and h; is the unit investment cost of the k-th type of cable; Cap gh is the planned capacity of the cable between microgrids g and h; Ψ g is the set of microgrids connected to microgrid g; D p is the number of days in a year; is the inferred quantity; the inferred quantity of the optimization planning model for the to-be-built microgrid N s is the number of scenarios; N t is the number of scheduling periods in a day; ω g,c is the probability of the c-th inference of the capacity of microgrid g for other microgrids; the probability ω of the optimization planning model for the to-be-built microgrid g,c = 1; ω s is the probability of scenario s; f i O&M (·) is the operation and maintenance cost function of equipment i; is the output power of equipment i; c em is the unit carbon emission penalty; γ em is the unit carbon emission of the diesel generator; a DG and b DG are the emission curve coefficients of the diesel generator; Δt is the microgrid scheduling time interval; is the rated capacity of the diesel generator; is the output power of the diesel generator; c LS is the unit load shedding cost; is the load shedding power; is the electricity price at time t; is the power purchase quantity from microgrid g to microgrid h, is the power sale quantity from microgrid g to microgrid h, γ rv is the residual coefficient; Among them, the operation and maintenance cost function f i O&M (·) includes the operation and maintenance cost function of new energy power stations the operation and maintenance cost function of diesel generators the operation and maintenance cost function of battery energy storage devices and the operation and maintenance cost function of hydrogen energy storage devices are calculated respectively by the following formulas: Wherein, is the output power of the new energy power station, is the unit operation and maintenance cost of the new energy power station, is the output power of the diesel generator, c f is the unit price of diesel, c DG is the unit operation and maintenance cost of the diesel generator, is the unit operation and maintenance cost of the battery energy storage, and are the charging and discharging powers of the battery energy storage in the microgrid g at the t-th time period under the inference c and scenario s, is the unit operation and maintenance cost of the hydrogen energy storage, and are the charging and discharging powers of the hydrogen energy storage; The constraint conditions of the optimization planning model of the to-be-built microgrid include the constraints related to planning, microgrid power flow constraints, diesel generator operation constraints, battery energy storage operation constraints, hydrogen energy storage operation constraints, and load shedding constraints; Among them, the constraints related to planning are as follows: Cap g,DG +Cap g,BS >0 (17) where Cap g,DG is the planned capacity of the diesel generator in microgrid g; Cap g,BS is the planned capacity of the battery energy storage; Cap g,i is the planned capacity of device i; is the upper limit of the planned capacity of device i; Cap gh is the planned cable capacity between microgrids g and h; is the upper limit of the planned cable capacity between microgrids g and h; The microgrid power flow constraints are as follows: Wherein, is the power purchase from microgrid g to microgrid h during the t-th scheduling period in the scenario s for inferring c; is the output power of device i; is the power sold from microgrid g to h; is the load power of microgrid g; is the load shedding power of microgrid g; Cap gh is the cable capacity between microgrids g and h; The diesel generator operation constraints are as follows: In the formula, and are the output powers of the diesel generators in the g-th microgrid during the t-th and (t-1)-th scheduling periods in the scenario s for inferring c; Cap g,DG is the planned capacity of the diesel generators in the g-th microgrid; ρ DG is the ramp rate of the diesel generators. The battery energy storage operation constraints are as follows: In the formula, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c; and are two-dimensional variables representing the charging state and discharging state of the battery energy storage; τ BS is the capacity-power conversion coefficient of the battery energy storage; ε is an infinitesimal constant; is the amount of electricity stored in the battery energy storage during the t-th scheduling period; is the amount of electricity stored in the battery energy storage during the (t - 1)-th scheduling period; Cap g,BS is the planned capacity of the battery energy storage; α BS is the lower limit coefficient of the battery energy storage power; β BS is the initial power coefficient of the battery energy storage; and are the initial power and final power of the battery energy storage in a scenario; Λ BS is the self-discharge rate of the battery energy storage; and are the charging efficiency and discharging efficiency of the battery energy storage; The hydrogen energy storage operation constraints are as follows: Wherein, and are the charging power and discharging power of the hydrogen energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c; and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage; τ HS is the capacity-power conversion coefficient of the hydrogen energy storage; ε is an infinitesimal constant; is the electric quantity stored in the hydrogen energy storage during the t-th scheduling period; is the electric quantity stored in the hydrogen energy storage during the (t - 1)-th scheduling period; Cap g,HS is the planned capacity of the hydrogen energy storage; α HS is the lower limit coefficient of the hydrogen energy storage electric quantity; is the initial electric quantity state of the hydrogen energy storage in the first scenario; are the initial electric quantity states of the hydrogen energy storage in the s-th and (s - 1)-th scenarios; is the final electric quantity state of the hydrogen energy storage in the (s - 1)-th scenario; is the final state of the last scenario; β HS is the initial electric quantity coefficient of the hydrogen energy storage; Λ HS is the self-discharge rate of the battery energy storage; and are the charging efficiency and discharging efficiency of the battery energy storage, is the maximum number of times the charge-discharge state of the hydrogen energy storage is allowed to change within a day; D p is the number of days in a year; ω s-1 is the probability of the (s - 1)-th scenario; The load shedding constraints are as follows: In the formula, is the load shedding power in the g-th microgrid at the t-th scheduling moment in the scenario s for inferring c; is the load power; ζ is the maximum load shedding ratio coefficient.

6. The method for planning an island microgrid based on imperfect information dynamic game according to claim 5, wherein Optimization objectives of the established microgrid optimal operation model Are as follows: Wherein, is the annual operation and maintenance cost, is the annual carbon emission penalty cost, is the annual load shedding penalty cost, is the annual power purchase cost, is the annual electricity selling revenue; The constraint conditions of the optimization planning model of the to-be-built microgrid include microgrid power flow constraints (20)-(23), diesel generator operation constraints (24)-(25), battery energy storage operation constraints (26)-(32), hydrogen energy storage operation constraints (33)-(41), and load shedding constraints (42).

7. The method for planning an island microgrid based on imperfect information dynamic game according to claim 6, wherein The steps of linearizing the optimization planning model of the to-be-built microgrid and the optimization operation model of the existing microgrid include: 1) Linearize the constraint conditions (24), (26)-(27), (33)-(34), and (39), where the non-linear terms in the constraint conditions are replaced as follows: In the formula, is the auxiliary variable for the operation constraint of the diesel generator in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, is the output power of the diesel generator, and are the auxiliary variables for the operation constraint of the battery energy storage, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the auxiliary variables for the operation constraint of the hydrogen energy storage, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage; 2) Add auxiliary constraint conditions to make the linearized model equivalent to the non-linear model; Among them, the constraint condition (25) is replaced as follows: In the formula, is the auxiliary variable of the operation constraint of the diesel generator in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, Cap g,DG is the planned capacity of the diesel generator, is the output power of the diesel generator, ρ DG is the ramp rate of the diesel generator, M DG is a positive number, and are the auxiliary variables of the operation constraint of the battery energy storage; The constraint conditions (26)-(27), (33)-(34) are replaced as follows: Wherein, and are the charging power and discharging power of the battery energy storage in the g-th microgrid during the t-th scheduling period in the scenario s for inferring c, and are auxiliary variables for the operating constraints of the battery energy storage, and are two-dimensional variables representing the charging state and discharging state of the battery energy storage, Cap g,BS is the planned capacity of the battery energy storage, and are the charging power and discharging power of the hydrogen energy storage, and are auxiliary variables for the operating constraints of the hydrogen energy storage, and are two-dimensional variables representing the charging state and discharging state of the hydrogen energy storage, Cap g,HS is the planned capacity of the hydrogen energy storage, M BS and M HS are all positive numbers; The constraint condition (39) is replaced as follows: In the formula, is the auxiliary variable of the hydrogen energy storage operation constraint in the t-th and g-th microgrids under the scenario s of inferring c; is the maximum number of times the charge and discharge state of the hydrogen energy storage is allowed to change within a day; 3) Establish the linear optimization model of the to-be-built microgrid, that is: Establish the linear optimization model of the existing microgrid, that is:

8. The method for planning an island microgrid based on imperfect information dynamic game according to claim 1, wherein The method for solving the linear optimization model of the to-be-built microgrid and the linear optimization model of the existing microgrid includes the mixed integer programming method; The microgrid optimization planning scheme includes the planning, operation, and power trading results of the to-be-built microgrid, and the operation and power trading results of the existing microgrid.

9. The method for planning a microgrid on an island based on dynamic game with imperfect information according to claim 1, wherein The steps of establishing the operation strategy coordination model of the microgrid group include: 1) Establish the optimization objective of the operation strategy coordination model of the microgrid group, that is: Where, E loss is the total amount of new energy curtailment and load shedding in the microgrid cluster, D p is the number of days in a year, N MG is the total number of microgrids in the microgrid cluster, is the total number of inferences for the g-th microgrid, N s is the total number of scenarios, N t is the number of scheduling periods in a day, ω g,c is the probability of the c-th inference for the capacity of the to-be-built microgrid of the g-th microgrid, ω s is the probability of the s-th scenario, is the new energy curtailment power of the g-th microgrid at the t-th scheduling period under scenario s of inference c, is the load shedding power, and Δt is the microgrid scheduling period interval; 2) Establish the power flow constraint conditions of the operation strategy coordination model of the microgrid group, that is: In the formula, is the power purchased by the g-th microgrid from the h-th microgrid during the t-th scheduling period in the scenario s for inferring c. is the output power of the i-th device. is the new energy shedding power of the g-th microgrid. is the power sold by the g-th microgrid to the h-th microgrid. is the load power. is the load shedding power, Ψ g is the set of microgrids connected to the g-th microgrid through cables, Φ g is the set of devices in the g-th microgrid; 3) Establish the operation strategy coordination model of the microgrid group, that is:

Images