Coordinated interactive transaction method of microgrid with long-short period hybrid energy storage

Abstract

The present application belongs to the technical field of electric energy transaction, and particularly relates to a distribution-microgrid collaborative interactive transaction method containing long-short cycle hybrid energy storage, steps of which comprise: constructing a distribution-microgrid collaborative architecture containing long-short cycle hybrid energy storage; taking the maximum profit of distribution network agent as an optimization target, constructing a target function and constraint conditions of an upper layer model; establishing a joint price uncertainty set; taking the minimum operation cost of microgrid as an optimization target, constructing a target function and constraint conditions of a lower layer model; converting the lower layer model into additional constraint conditions of the upper layer model, linearizing the target function of the upper layer model, and in the master-slave game process of the upper and lower layer models, optimizing the linearized target function based on the joint price uncertainty set and solving, to obtain an optimal distribution-microgrid collaborative interactive transaction scheme. Through the long-short cycle hybrid energy storage and dynamic pricing mechanism, the present application solves the problems of unstable new energy and market price fluctuation, and significantly improves the economy of the collaborative operation of distribution-microgrid.

Description

Technical Field

[0001] This invention belongs to the field of power trading technology, specifically relating to a method for collaborative and interactive trading of distribution-microgrid systems that incorporates hybrid energy storage with both long and short lifecycles. Background Technology

[0002] Currently, the coordinated operation of distribution and microgrids mainly relies on energy storage. Distribution networks can interact with microgrids by installing energy storage, meeting the source-load matching needs of microgrids, avoiding the costs of wind and solar curtailment for microgrids, and allowing surplus electricity to be sold to the electricity market, thus achieving price transfer. Furthermore, distribution networks can formulate real-time dynamic pricing strategies based on microgrid energy consumption, and microgrids can optimize their own electricity consumption strategies under the guidance of the distribution network's pricing strategy. This real-time dynamic optimization between the two sides achieves a win-win situation in a master-slave game.

[0003] However, traditional single electrochemical energy storage is ill-suited to the long-term source-load matching needs under the background of high renewable energy penetration, and it is difficult to cope with the decline in the economics of distribution-microgrid collaborative systems under the condition of fluctuating electricity market prices. Furthermore, existing research focuses only on improving the system's source-load matching capability, neglecting the impact of energy storage on the system's economics under the condition of fluctuating electricity market transaction prices. The potential of long-term and short-term hybrid energy storage to mitigate the risks of external electricity market price fluctuations and balance the interests of both parties in the transaction is not fully explored. Existing research has not fully considered the conditions of market price fluctuations and has ignored the uncertainty of market price fluctuations. The leader pricing mechanism focuses on the pricing principle of intraday short-term transaction electricity, and has failed to give full play to the role of improving the average margin of pricing over long time scales. Existing research implements fixed pricing or quantity-based pricing for interruption load compensation prices, ignoring the impact of multiple price variables on the demand side under the hierarchical pricing feedback mechanism. The flexibility of demand-side participation in electricity trading is insufficient, and the potential for two-way interaction between user willingness and electricity resources is not fully explored. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide a distribution-microgrid collaborative interactive trading method with hybrid energy storage of long and short cycles. By combining hybrid energy storage of long and short cycles with a dynamic pricing mechanism, the method solves the problems of instability of new energy sources and market price fluctuations, and significantly improves the economic efficiency of distribution-microgrid collaborative operation.

[0005] To achieve the above objectives, this invention provides a distribution-microgrid collaborative interaction trading method incorporating hybrid long- and short-cycle energy storage, comprising the following steps:

[0006] S1. Construct a distribution-microgrid collaborative architecture that includes hybrid energy storage with both long and short cycles. This includes an upper layer where the distribution network is the main stakeholder and a lower layer where the microgrid is the main stakeholder. The upper layer, the distribution network, includes distribution network agents, heat recovery devices, short-cycle energy storage, and long-cycle energy storage. The lower layer, the microgrid, includes wind power, users, and photovoltaics.

[0007] S2. Considering the interaction between the distribution network and the electricity market, heat market, and microgrid, and taking the maximization of distribution network agent profits as the optimization objective, construct the objective function and its constraints for the upper-level model.

[0008] S3. Establish a joint price uncertainty set based on Wasserstein distance to characterize changes in electricity market prices;

[0009] S4. With the goal of minimizing the microgrid operating cost, construct the objective function and its constraints for the lower-level model.

[0010] S5. Perform non-convex term transformation and constraint standardization on the lower-level model, then transform the lower-level model into additional constraints of the upper-level model, linearize the objective function of the upper-level model, optimize the linearized objective function based on the joint price uncertainty set during the master-slave game between the upper and lower-level models, and solve the optimized objective function to obtain the optimal distribution-microgrid collaborative interaction trading scheme.

[0011] As a preferred embodiment of the present invention, in S1, the short-cycle energy storage of the power distribution network includes electrochemical energy storage and thermal energy storage, while the long-cycle energy storage includes electrolyzers, hydrogen storage tanks, and fuel cells.

[0012] In a preferred embodiment of the present invention, in S2, the objective function of the upper-level model is:

[0013] (1);

[0014] In the formula, For revenue generated from transactions between the distribution network and the microgrid; To generate revenue from real-time electricity market transactions between the distribution network and the power market; For revenue generated from transactions between the power distribution network and the heat market; The day-ahead electricity purchase cost for the distribution network and the electricity market; For energy storage operating costs;

[0015] The constraints of the upper-level model include energy balance constraints, electricity purchase and sale price constraints between the distribution network and the microgrid, and electricity market transaction constraints between the distribution network and the power market. Short-cycle energy storage device models, long-cycle energy storage models, and heat recovery device models are also established.

[0016] As a preferred embodiment of the present invention Represented as:

[0017] (2);

[0018] In the formula, , These represent the electricity sales and purchase prices from the distribution network to the microgrid during dispatch period t on day d, where D is the total number of days and T is the total number of dispatch periods. , These represent the power sold and purchased by the distribution network to the microgrid during the dispatching period t on day d. For time intervals;

[0019] Represented as: (3);

[0020] In the formula, , These represent the real-time electricity sales and purchase prices from the distribution network to the electricity market during the dispatching period on day d, respectively. , These represent the real-time power sold and purchased by the distribution network to the electricity market during the dispatching period t on day d.

[0021] Represented as: (4);

[0022] In the formula, The real-time heat price for the distribution network to sell heat to the heat market during the dispatch period on day d; This represents the real-time heat sales capacity of the distribution network to the heat market during the dispatching period on day d.

[0023] Represented as: (5);

[0024] In the formula, The day-ahead contract price for the dispatch period t on day d; The day-ahead contract power for the scheduling period t on day d;

[0025] Represented as: (6);

[0026] In the formula, , , , , These are the operating costs of electrochemical energy storage, hydrogen energy storage, thermal energy storage, fuel cells, and electrolyzers, respectively.

[0027] As a preferred embodiment of the present invention, the energy balance constraint is:

[0028] In the scenario of power distribution network selling electricity to microgrid, the power purchased by the power distribution network from the electricity market on a day-ahead basis, the power purchased by the power distribution network from the electricity market in real time, the power purchased by the power distribution network from the microgrid, the power discharged by short-cycle energy storage, and the power generated by fuel cells together constitute the power input. The power sold by the power distribution network to the microgrid, the power charged by short-cycle energy storage, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal.

[0029] In the scenario of the distribution network selling electricity to the electricity market, the power purchased by the distribution network from the electricity market on the day-ahead, the power purchased by the distribution network from the electricity market in real time, the power purchased by the distribution network from the microgrid, the power of short-cycle energy storage discharge, and the power generated by fuel cells together constitute the power input. The power sold by the distribution network to the electricity market, the power of short-cycle energy storage charging, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal.

[0030] The sum of hydrogen production from the electrolyzer and hydrogen release from long-cycle hydrogen storage equals the sum of hydrogen charging from long-cycle hydrogen storage and hydrogen consumption from fuel cells.

[0031] The sum of the heat release power of the heat recovery device and the heat release power of the thermal energy storage is equal to the sum of the heat charging power of the thermal energy storage and the heat selling power to the heat market.

[0032] The price constraint for electricity purchase and sale from the distribution network to the microgrid is:

[0033] (7);

[0034] (8);

[0035] (9);

[0036] (10);

[0037] In the formula, , These represent the upper and lower limits of the electricity price sold from the distribution network to the microgrid during the dispatching period t on day d. , These represent the upper and lower limits of the electricity purchase price from the microgrid by the distribution network during the t-period dispatch on day d; , These are the power sales and purchase flags for the distribution network during the dispatching period t on day d. When the distribution network sells electricity to the microgrid during the dispatching period t on day d... Otherwise, it equals 0. Similarly;

[0038] The sum of the difference between the total revenue from the real-time sale of electricity from the distribution network to the electricity market and the total cost of purchasing electricity from the microgrid is greater than or equal to 0.

[0039] The constraints on power distribution network and electricity market transactions are:

[0040] (11);

[0041] (12);

[0042] (13);

[0043] (14);

[0044] In the formula, This refers to the maximum daily power purchase capacity of the distribution network under contract with the electricity market. , This refers to the maximum real-time power sales and purchase capacity of the distribution network and the electricity market. , These are the power sales and purchase flags for the dispatching period t on day d, respectively. When the distribution network sells electricity to the electricity market during the dispatching period t on day d... Otherwise, it equals 0. Similarly.

[0045] As a preferred embodiment of the present invention, the short-cycle energy storage device model includes an electrochemical energy storage model and a thermal energy storage model, wherein the electrochemical energy storage model is expressed as follows:

[0046] (15);

[0047] (16);

[0048] In the formula, , These represent the stored energy of electrochemical energy storage during scheduling periods t and t-1 on day d, respectively. The self-discharge rate of electrochemical energy storage; To improve the charge and discharge efficiency of electrochemical energy storage; , These represent the electrochemical energy storage charging and discharging power during the t-period scheduling on day d, respectively. , These represent the stored energy of electrochemical energy storage during scheduling periods 1 and 24 on day d, respectively. , These represent the charging and discharging power of electrochemical energy storage during scheduling period 1 on day d; similarly, a thermal energy storage model is constructed.

[0049] Long-term energy storage models include: hydrogen energy storage device models:

[0050] (17);

[0051] (18);

[0052] (19);

[0053] In the formula, , These represent the hydrogen storage capacity during the hydrogen storage period on day d and t-1, respectively. , These represent the hydrogen storage charging and discharging amounts during the t-period scheduling time on day d. To improve the efficiency of hydrogen charging and discharging for hydrogen energy storage; This represents the amount of hydrogen stored in the hydrogen storage system during the first scheduling period on day d+1. This represents the amount of hydrogen stored in the hydrogen storage system during the 24-hour scheduling period on day d. , Similarly; , These represent the hydrogen storage charging and discharging amounts for the first scheduling period on day d+1. , Similarly;

[0054] Electrolyzer-Fuel Cell Model: (20);

[0055] (twenty one);

[0056] In the formula, The amount of hydrogen produced by the electrolyzer during the scheduling period t on day d; The power consumption of the electrolytic cell during the scheduling period t on day d; The fuel cell power generation during the t-th scheduling period on day d; This represents the hydrogen consumption of the fuel cell during the t-day scheduling period on day d. , These are the efficiencies of the electrolyzer and the fuel cell, respectively.

[0057] The heat recovery device model is represented as follows: (twenty two);

[0058] In the formula, The heat recovered by the heat recovery device during the t-day scheduling period on day d; , These refer to the efficiency of the heat recovery device in recovering waste heat from the electrolyzer and hydrogen fuel cell, respectively.

[0059] As a preferred embodiment of the present invention, in S3, the process of establishing the joint price uncertainty set based on Wasserstein distance is as follows:

[0060] S3.1, Establish based on , Joint price support set Historical price data support set :

[0061] (twenty three);

[0062] (twenty four);

[0063] In the formula, c represents a price pair containing , ; , These represent the upper and lower limits of the day-ahead contract power purchase price of the distribution network and the electricity market during the dispatching period on day d; , These represent the upper and lower limits of the real-time electricity sales price from the distribution network to the electricity market during the dispatching period on day d; This represents the l-th price sample of the distribution network during dispatch period t, including , , This represents the l-th price sample of the day-ahead contract power purchase between the distribution network and the electricity market during the t-schedule period. The l-th price sample for the real-time electricity sale from the distribution network to the electricity market during the t-schedule period;

[0064] S3.2. Obtain the empirical probability distribution based on the N sets of historical data. To estimate the true distribution Q, where To concentrate on The Dirac measure; described using 1-Wasserstein distance based on the joint price distribution. Distance between Q and :

[0065] (25);

[0066] In the formula, inf represents the infimum; Represents the 1-norm; Indicates based on and Q distribution The joint distribution of c;

[0067] S3.3, Define the joint price uncertainty set as: (26);

[0068] In the formula, express All probability distributions on; Indicated by With the center of the ball, A Wasserstein sphere with radius .

[0069] The method for determining it is as follows: (27);

[0070] (28);

[0071] In the formula, G is a coefficient related to the data distribution; Confidence level; This is the adjustment coefficient; express The mean of the sample; e is the natural constant; n is the sample index.

[0072] In a preferred embodiment of the present invention, in S4, the objective function of the lower-level model is:

[0073] (29);

[0074] In the formula, Compensation payments made to users for load interruptions in microgrids; Costs associated with curtailing wind and solar power;

[0075] Represented as: (30);

[0076] In the formula, The unit electricity compensation price paid by the microgrid to users during the t-schedule period on day d; The interrupted user load power during the t-period scheduling on day d;

[0077] Represented as: (31);

[0078] In the formula, Cost per unit of wind and solar curtailment; The power of wind and solar power curtailed by the microgrid during the t-th dispatch period on day d;

[0079] The constraints of the lower-level model include microgrid power balance constraints, microgrid power purchase and sale to the distribution network constraints, microgrid interruptible load constraints, and microgrid wind and solar power curtailment constraints.

[0080] As a preferred embodiment of the present invention, in S5, the objective function for optimizing linearization based on the joint price uncertainty set is specifically as follows: in the master-slave game process of the upper and lower level models, after obtaining the first stage optimization variables, the second stage obtains the price distribution under the extreme scenario by optimizing the expectation under the uncertainty scenario, and optimizes the objective function for the price distribution under the extreme scenario.

[0081] The price distribution in extreme scenarios is represented as follows: (32);

[0082] (33);

[0083] In the formula, C, x, and p represent the optimization variables in the first stage; X is the feasible region in the first stage. Expressing expectations; Represents the objective function in a single scenario;

[0084] According to the strong duality principle, the expectation can be transformed into the following form:

[0085] (34);

[0086] In the formula, is the dual variable; l is the index of the sample; sup denotes the supremum;

[0087] Introducing intermediate variables , , ,make:

[0088] (35);

[0089] (36);

[0090] Then equation (32) is transformed into the following form: (37);

[0091] (38).

[0092] As a preferred embodiment of the present invention, in S5, the optimized distribution-microgrid collaborative interaction transaction scheme is as follows: the distribution network formulates a day-ahead contract power purchase strategy based on the source and load information provided by the microgrid, the maximum interruptible load declared by the microgrid users, and the day-ahead price signal of the power market, and completes the power interaction with the power market.

[0093] When the daily contracted electricity volume cannot meet the microgrid's demand during real-time trading, the distribution network and the electricity market conduct real-time trading to supplement the shortfall. When the distribution network interacts with the microgrid to generate surplus electricity, the distribution network sells electricity to the microgrid during periods of insufficient renewable energy output, periods of high real-time electricity purchase price in the electricity market, and periods of high real-time heat purchase price in the heat market through the intraday and interday energy transfer characteristics of hybrid energy storage, so as to complete the transfer of electricity price.

[0094] Meanwhile, the distribution network formulates a real-time, continuous, and dynamic pricing strategy based on the energy consumption of the microgrid to ensure its own maximum revenue; the microgrid adjusts its own electricity consumption ratio in real time based on the distribution network's pricing strategy to decide on the electricity purchase and sale strategy and interruption strategy that are beneficial to itself, and to minimize its own operating costs.

[0095] The beneficial effects of this invention are:

[0096] This invention introduces a hybrid long- and short-cycle energy storage system into the coordinated dispatch of distribution and microgrids, effectively solving the problem of mismatch between renewable energy output and load. Long-cycle energy storage (such as hydrogen energy storage) can perform energy shifting across days and weeks to cope with long-term source-load fluctuations; short-cycle energy storage (such as electrochemical energy storage and thermal energy storage) is responsible for rapid intraday power regulation. The synergy between the two significantly improves the system's source-load matching capability and operational flexibility, significantly increases the renewable energy absorption rate, and reduces the cost of wind and solar curtailment.

[0097] This invention proposes a master-slave trading model based on a continuous dynamic pricing mechanism. The distribution network formulates a real-time dynamic pricing strategy based on the energy consumption of the microgrid, and the microgrid optimizes its own electricity consumption strategy accordingly. This model fully leverages the effect of improving the average pricing margin over long time scales, significantly increasing the distribution network's revenue compared to traditional independent pricing methods, while reducing the microgrid's operating costs, achieving a win-win situation for both parties, enhancing the system's ability to cope with electricity market price fluctuations, and improving overall economic efficiency.

[0098] This invention addresses the uncertainty of electricity market price fluctuations by introducing a Wasserstein uncertainty set based on a joint price distribution. Compared to independent uncertainty sets, this method accurately characterizes the coupling relationship between electricity purchase and sales prices, avoiding both from simultaneously taking the worst-case scenario, thus making the optimization results closer to the actual market. At different confidence levels, the distribution network revenue based on the joint uncertainty set outperforms the independent method, effectively reducing model conservatism, improving system robustness, and facilitating the stable and efficient operation of distribution-microgrids in complex market environments. Attached Figure Description

[0099] Figure 1 This is a flowchart illustrating the principle of this invention;

[0100] Figure 2 This is a schematic diagram of the distribution-microgrid collaborative architecture during the verification process of this invention;

[0101] Figure 3 This is a schematic diagram illustrating distribution network dispatching strategies under different scenarios of market price fluctuations during the verification process of this invention. Figure 3 (a) in the diagram is a schematic diagram of the distribution network dispatching strategy in the first week of scenario 2; Figure 3 (b) in the diagram is a schematic diagram of the distribution network dispatching strategy for the first week of scenario 1;

[0102] Figure 4 This is a schematic diagram illustrating distribution network dispatching strategies under different scenarios of market price fluctuations during the verification process of this invention. Figure 4 (a) in the diagram is a schematic diagram of the distribution network dispatching strategy in the second week of scenario 2; Figure 4 (b) in the diagram is a schematic diagram of the distribution network dispatching strategy in the second week of scenario 1;

[0103] Figure 5 This is a schematic diagram illustrating distribution network dispatching strategies under different scenarios of market price fluctuations during the verification process of this invention. Figure 5 (a) in the diagram is a schematic diagram of the distribution network dispatching strategy in the third week of scenario 2; Figure 5 (b) in the diagram is a schematic diagram of the distribution network dispatching strategy in the third week of scenario 1;

[0104] Figure 6 This is a schematic diagram of the coordinated charging and discharging strategy of energy storage at all levels in scenario 2 during the verification process of this invention;

[0105] Figure 7 This is a schematic diagram of the distribution network pricing strategy on the second day of the first week within the scheduling cycle during the verification process of this invention;

[0106] Figure 8 This is a schematic diagram of the distribution network pricing strategy on the 5th day of the 2nd week within the scheduling cycle during the verification process of this invention;

[0107] Figure 9 This is a schematic diagram of the distribution network pricing strategy on the first day of the third week within the scheduling cycle during the verification process of this invention;

[0108] Figure 10 This is a schematic diagram of the microgrid interruption strategy on the second day of the first week during the verification process of this invention;

[0109] Figure 11 This is a schematic diagram of the microgrid interruption strategy on the 5th day of the 2nd week during the verification process of this invention;

[0110] Figure 12 This is a schematic diagram of the microgrid interruption strategy on the first day of the third week during the verification process of this invention. Detailed Implementation

[0111] The embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0112] Example 1: As Figure 1 As shown, the distribution-microgrid collaborative interaction trading method including long- and short-cycle hybrid energy storage includes the following steps:

[0113] S1. Construct a distribution-microgrid collaborative architecture that includes hybrid energy storage with both long and short cycles. This includes an upper layer where the distribution network is the main stakeholder and a lower layer where the microgrid is the main stakeholder. The upper layer, the distribution network, includes distribution network agents, heat recovery devices, short-cycle energy storage, and long-cycle energy storage. The lower layer, the microgrid, includes wind power, users, and photovoltaics.

[0114] S2. Considering the interaction between the distribution network and the electricity market, heat market, and microgrid, and taking the maximization of distribution network agent profits as the optimization objective, construct the objective function and its constraints for the upper-level model.

[0115] S3. Establish a joint price uncertainty set based on Wasserstein distance to characterize changes in electricity market prices;

[0116] S4. With the goal of minimizing the microgrid operating cost, construct the objective function and its constraints for the lower-level model.

[0117] S5. Perform non-convex term transformation and constraint standardization on the lower-level model, then transform the lower-level model into additional constraints of the upper-level model, linearize the objective function of the upper-level model, optimize the linearized objective function based on the joint price uncertainty set during the master-slave game between the upper and lower-level models, and solve (using a commercial solver) the optimized objective function to obtain the optimal distribution-microgrid collaborative interaction trading scheme.

[0118] In S1, short-cycle energy storage in the distribution network includes electrochemical energy storage and thermal energy storage, while long-cycle energy storage includes electrolyzers, hydrogen storage tanks (hydrogen energy storage), and fuel cells.

[0119] In S2, the objective function of the upper-level model is: (39);

[0120] In the formula, For revenue generated from transactions between the distribution network and the microgrid; To generate revenue from real-time electricity market transactions between the distribution network and the power market; For revenue generated from transactions between the power distribution network and the heat market; The day-ahead electricity purchase cost for the distribution network and the electricity market; For energy storage operating costs;

[0121] The constraints of the upper-level model include energy balance constraints, electricity purchase and sale price constraints between the distribution network and the microgrid, and electricity market transaction constraints between the distribution network and the power market. Short-cycle energy storage device models, long-cycle energy storage models, and heat recovery device models are also established.

[0122] Represented as: (40);

[0123] In the formula, , These represent the electricity sales and purchase prices from the distribution network to the microgrid during dispatch period t on day d, where D is the total number of days and T is the total number of dispatch periods. , These represent the power sold and purchased by the distribution network to the microgrid during the dispatching period t on day d. For time intervals;

[0124] Represented as: (41);

[0125] In the formula, , These represent the real-time electricity sales and purchase prices from the distribution network to the electricity market during the dispatching period on day d, respectively. , These represent the real-time power sold and purchased by the distribution network to the electricity market during the dispatching period t on day d.

[0126] Represented as: (42);

[0127] In the formula, The real-time heat price for the distribution network to sell heat to the heat market during the dispatch period on day d; This represents the real-time heat sales capacity of the distribution network to the heat market during the dispatching period on day d.

[0128] Represented as: (43);

[0129] In the formula, The day-ahead contract price for the dispatch period t on day d; The day-ahead contract power for the scheduling period t on day d;

[0130] Represented as: (44);

[0131] In the formula, , , , , The operating costs are as follows: electrochemical energy storage, hydrogen energy storage, thermal energy storage, fuel cells, and electrolyzers.

[0132] (45);

[0133] (46);

[0134] (47);

[0135] (48);

[0136] (49);

[0137] In the formula, , , , , These are the unit operating costs for electrochemical energy storage, hydrogen energy storage, thermal energy storage, fuel cells, and electrolyzers, respectively. , These represent the electrochemical energy storage charging and discharging power during the t-period scheduling on day d, respectively. , These represent the hydrogen storage charging and discharging amounts during the t-period scheduling time on day d. , The heat charge and release during the t-schedule period on day d are respectively; The power consumption of the electrolytic cell during the scheduling period t on day d; The fuel cell power generation during the t-th scheduling period on day d;

[0138] The energy balance constraint is:

[0139] In the scenario of power distribution network selling electricity to microgrid, the power purchased by the power distribution network from the electricity market on a day-ahead basis, the power purchased by the power distribution network from the electricity market in real time, the power purchased by the power distribution network from the microgrid, the power discharged by short-cycle energy storage, and the power generated by fuel cells together constitute the power input. The power sold by the power distribution network to the microgrid, the power charged by short-cycle energy storage, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal.

[0140] In the scenario of the distribution network selling electricity to the electricity market, the power purchased by the distribution network from the electricity market on the day-ahead, the power purchased by the distribution network from the electricity market in real time, the power purchased by the distribution network from the microgrid, the power of short-cycle energy storage discharge, and the power generated by fuel cells together constitute the power input. The power sold by the distribution network to the electricity market, the power of short-cycle energy storage charging, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal.

[0141] The sum of hydrogen production from the electrolyzer and hydrogen release from long-cycle hydrogen storage equals the sum of hydrogen charging from long-cycle hydrogen storage and hydrogen consumption from fuel cells.

[0142] The sum of the heat release power of the heat recovery device and the heat release power of the thermal energy storage is equal to the sum of the heat charging power of the thermal energy storage and the heat selling power to the heat market.

[0143] The price constraint for electricity purchase and sale from the distribution network to the microgrid is:

[0144] (50);

[0145] (51);

[0146] (52);

[0147] (53);

[0148] In the formula, , These represent the upper and lower limits of the electricity price sold from the distribution network to the microgrid during the dispatching period t on day d. , These represent the upper and lower limits of the electricity purchase price from the microgrid by the distribution network during the t-period dispatch on day d; , These are the power sales and purchase flags for the distribution network during the dispatching period t on day d. When the distribution network sells electricity to the microgrid during the dispatching period t on day d... Otherwise, it equals 0. Similarly;

[0149] The sum of the difference between the total revenue from the real-time sale of electricity from the distribution network to the electricity market and the total cost of purchasing electricity from the microgrid is greater than or equal to 0.

[0150] The constraints on power distribution network and electricity market transactions are:

[0151] (54);

[0152] (55);

[0153] (56);

[0154] (57);

[0155] In the formula, This refers to the maximum daily power purchase capacity of the distribution network under contract with the electricity market. , This refers to the maximum real-time power sales and purchase capacity of the distribution network and the electricity market. , These are the power sales and purchase flags for the dispatching period t on day d, respectively. When the distribution network sells electricity to the electricity market during the dispatching period t on day d... Otherwise, it equals 0. Similarly.

[0156] Short-cycle energy storage device models include electrochemical energy storage models and thermal energy storage models. The electrochemical energy storage model is represented as follows:

[0157] (58);

[0158] (59);

[0159] In the formula, , These represent the stored energy of electrochemical energy storage during scheduling periods t and t-1 on day d, respectively. The self-discharge rate of electrochemical energy storage; To improve the charge and discharge efficiency of electrochemical energy storage; , These represent the stored energy of electrochemical energy storage during scheduling periods 1 and 24 on day d, respectively. , These represent the electrochemical energy storage charging and discharging power during the first scheduling period on day d;

[0160] The constraints of the electrochemical energy storage model are:

[0161] (60);

[0162] (61);

[0163] (62);

[0164] (63);

[0165] In the formula, , These are the upper and lower limits of energy storage for electrochemical energy storage, respectively. This represents the upper limit of the charge and discharge power of electrochemical energy storage. This represents the upper limit of energy storage capacity for electrochemical energy storage. , These are the charging and discharging flags for the scheduling period t on day d of electrochemical energy storage, with values ​​of 0 or 1 (the other flags are similar).

[0166] Similarly, a thermal energy storage model is constructed (electrochemical energy storage is transformed into thermal energy storage).

[0167] Long-term energy storage models include: hydrogen energy storage device models:

[0168] (64);

[0169] (65);

[0170] (66);

[0171] In the formula, , These represent the hydrogen storage capacity during the hydrogen storage period on day d and t-1, respectively. To improve the efficiency of hydrogen charging and discharging for hydrogen energy storage; This represents the amount of hydrogen stored in the hydrogen storage system during the first scheduling period on day d+1. This represents the amount of hydrogen stored in the hydrogen storage system during the 24-hour scheduling period on day d. , Similarly; , These represent the hydrogen storage charging and discharging amounts for the first scheduling period on day d+1. , Similarly;

[0172] The constraints for the hydrogen energy storage device model are:

[0173] (67);

[0174] (68);

[0175] (69);

[0176] (70);

[0177] In the formula, , These are the upper and lower limits of hydrogen storage capacity for hydrogen energy storage, respectively. This represents the upper limit for the amount of hydrogen that can be charged or released in hydrogen energy storage. This represents the upper limit of hydrogen storage capacity for hydrogen energy storage; , These are the hydrogen storage charging and discharging flags for the first scheduling period on day d;

[0178] Electrolyzer-Fuel Cell Model: (71);

[0179] (72);

[0180] In the formula, The amount of hydrogen produced by the electrolyzer during the scheduling period t on day d; This represents the hydrogen consumption of the fuel cell during the t-day scheduling period on day d. , These are the efficiencies of the electrolyzer and the fuel cell, respectively.

[0181] The constraints for the electrolyzer-fuel cell model are:

[0182] (73);

[0183] (74);

[0184] (75);

[0185] In the formula, This is the upper limit of the power consumption of the electrolytic cell; This represents the upper limit of fuel cell power generation. , These are the flags for hydrogen production via electrolysis in an electrolyzer and hydrogen consumption for power generation in a fuel cell.

[0186] The heat recovery device model is represented as follows: (76);

[0187] In the formula, The heat recovered by the heat recovery device during the t-day scheduling period on day d is greater than or equal to zero and less than or equal to the upper limit of the heat recovery device. , These refer to the efficiency of the heat recovery device in recovering waste heat from the electrolyzer and hydrogen fuel cell, respectively.

[0188] Electricity market transaction prices fluctuate due to market supply and demand. In order to consider possible future price fluctuations during the optimization process and to avoid losing the coupling relationship between price variables by independently establishing uncertainty sets for multiple price variables in the electricity market, which would make the optimization results too conservative and seriously deviate from reality, this embodiment uses the Wasserstein distance method to construct a joint price distributed bar optimization scheduling model using historical data samples to improve the economic efficiency of system operation.

[0189] In S3, the process of establishing the joint price uncertainty set based on Wasserstein distance is as follows:

[0190] S3.1 To balance economy and robustness, and to characterize the coupling characteristics between market transaction prices, a system based on... , Joint price support set Historical price data support set :

[0191] (77);

[0192] (78);

[0193] In the formula, c represents a price pair containing , ; , These represent the upper and lower limits of the day-ahead contract power purchase price of the distribution network and the electricity market during the dispatching period on day d; , These represent the upper and lower limits of the real-time electricity sales price from the distribution network to the electricity market during the dispatching period on day d; This represents the l-th price sample of the distribution network (in the distribution-microgrid collaborative architecture, there is only one distribution network; here, it refers to samples in the price uncertainty set, where each sample corresponds to one distribution network and one price sample) during scheduling period t, including... , , This represents the l-th price sample of the day-ahead contract power purchase between the distribution network and the electricity market during the t-schedule period. The l-th price sample for the real-time electricity sale from the distribution network to the electricity market during the t-schedule period;

[0194] S3.2. Obtain the empirical probability distribution based on the N sets of historical data. To estimate the true distribution Q, where To concentrate on The Dirac measure; described using 1-Wasserstein distance based on the joint price distribution. Distance between Q and :

[0195] (79);

[0196] In the formula, inf represents the infimum; Represents the 1-norm; Indicates based on and Q distribution The joint distribution of c;

[0197] S3.3, Define the joint price uncertainty set as: (80);

[0198] In the formula, express All probability distributions on; Indicated by With the center of the ball, A Wasserstein sphere with radius .

[0199] The method for determining it is as follows: (81);

[0200] (82);

[0201] In the formula, G is a coefficient related to the data distribution; Confidence level; This is the adjustment coefficient (a positive coefficient greater than 0, which can be found using a binary search method). express The mean of the sample; e is the natural constant; n is the sample index.

[0202] In S4, the objective function of the lower-level model is: (83);

[0203] In the formula, Compensation payments made to users for load interruptions in microgrids; Costs associated with curtailing wind and solar power;

[0204] Represented as: (84);

[0205] In the formula, The unit electricity compensation price paid by the microgrid to users during the t-schedule period on day d; The interrupted user load power during the t-period scheduling on day d;

[0206] Represented as: (85);

[0207] In the formula, Cost per unit of wind and solar curtailment; The power of wind and solar power curtailed by the microgrid during the t-th dispatch period on day d;

[0208] The constraints of the lower-level model include microgrid power balance constraints:

[0209] (86);

[0210] In the formula, , , These represent the wind power, photovoltaic, and load output of the microgrid during the dispatch period t on day d. Lagrange multipliers for equality constraints;

[0211] Constraints on the power purchased and sold by microgrids from the distribution network:

[0212] (87);

[0213] (88);

[0214] (89);

[0215] In the formula, , These represent the maximum power purchase and sale capacity of the microgrid in transactions with the distribution network; , , , , These are the Lagrange multipliers corresponding to the inequality constraints;

[0216] Interruptible load constraints within the microgrid:

[0217] (90);

[0218] (91);

[0219] (92);

[0220] In the formula, The maximum power of the microgrid interruptible load during the t-day dispatch period on day d; , These represent the upper and lower limits of the microgrid interruption load compensation price for the dispatching period t on day d, respectively. , , , , These are the Lagrange multipliers corresponding to the inequality constraints; This is the interruption load flag bit for the microgrid during the t-th dispatch period on day d;

[0221] Constraints of wind and solar power curtailment in microgrids: (93);

[0222] In the formula, This represents the maximum value of wind and solar power curtailment in microgrids. , For the Lagrange multipliers corresponding to the inequality constraints.

[0223] The lower-level model undergoes non-convex term transformation and constraint standardization. Subsequently, the lower-level model is transformed into additional constraints for the upper-level model. The objective function of the upper-level model is linearized. Finally, the objective function is transformed into the following form:

[0224] (94);

[0225] In the formula, For the corresponding The logarithmic transformation form; For the corresponding The logarithmic transformation form;

[0226] In S5, the objective function for optimizing linearization based on the joint price uncertainty set is as follows: In the master-slave game process of the upper and lower level models (the process of seeking a win-win situation based on the interaction of different optimization objectives and constraints), after obtaining the first-stage optimization variables, the second stage obtains the price distribution under the extreme scenario by optimizing the expectation under the uncertainty scenario, and then optimizes the linearization objective function for the price distribution under the extreme scenario.

[0227] The price distribution in extreme scenarios is represented as follows: (95);

[0228] (96);

[0229] In the formula, C, x, and p represent the optimization variables in the first stage (including...). , , , , , , , , , , , , , , , , , X represents the feasible region in the first phase; Expressing expectations; This represents the objective function in a single scenario; the superscript T indicates transpose.

[0230] According to the strong duality principle, the expectation can be transformed into the following form:

[0231] (97);

[0232] In the formula, is the dual variable; l is the index of the sample, and each sample contains a microgrid; sup denotes the supremum;

[0233] Introducing intermediate variables , , ( It is an implicit definition, which indirectly limits the value logic of variables through constraints. Let:

[0234] (98);

[0235] (99);

[0236] Then equation (95) is transformed into the following form: (100);

[0237] (101).

[0238] In S5, the optimized distribution-microgrid collaborative interaction trading scheme is as follows: the distribution network, based on the source and load information provided by the microgrid ( , , ), the maximum interruptible load declared by microgrid users ( Electricity market day-ahead price signals ( ) Formulate a day-ahead power purchase strategy ( ), to complete the exchange of electricity volume with the electricity market;

[0239] When the day-ahead contracted electricity volume cannot meet the microgrid's demand during real-time trading, the distribution network and the electricity market engage in real-time trading ( , To supplement the deficit, when the distribution network interacts with the microgrid to generate surplus electricity ( During periods when the distribution network's output to microgrid renewable energy is insufficient, the network utilizes the intraday and interday energy transfer characteristics of hybrid energy storage. Electricity sales during periods of high real-time purchase price in the electricity market ( ), and the real-time purchase price of heat in the heat market during peak hours for heat sales ( ( ), to complete the electricity price transfer;

[0240] Meanwhile, the distribution network formulates a real-time, continuous, and dynamic pricing strategy based on the energy consumption of the microgrid. , To ensure its own maximum benefit; the microgrid adjusts its own electricity consumption ratio in real time based on the distribution network pricing strategy to determine the electricity purchase and sale strategy that is beneficial to itself. , Interrupt strategy () , (and minimize its own operating costs).

[0241] Both parties achieve a win-win situation under the master-slave game by dynamically optimizing information flow and energy flow in real time over a long period of time, and jointly address the problem of declining system economics caused by source-load mismatch and fluctuations in electricity market prices.

[0242] The verification process is as follows:

[0243] Taking a real distribution-microgrid coordinated system as an example, this paper analyzes and verifies the effectiveness of the long- and short-cycle hybrid energy storage coordinated dispatch system (distribution-microgrid coordinated system) in improving the economic efficiency of operation. Take 0.25 , Take 1.5 , A price of 0.4 yuan / kW is used. A schematic diagram of the distribution-microgrid collaborative architecture (distribution-microgrid collaborative system) is shown below. Figure 2 As shown.

[0244] Economic operation analysis of different types of energy storage: In order to verify the advantages of hybrid energy storage with long and short cycles in improving the economic efficiency of distribution-microgrid coordinated operation under market price fluctuations, two different energy storage scenarios were set up for comparative verification.

[0245] Scenario 1: In a distribution-microgrid collaborative system, only short-timescale electrochemical energy storage is used.

[0246] Scenario 2: In a distribution-microgrid coordinated system, considering the high combined heat and power efficiency of the electro-hydrogen coupling system, short-timescale electrical energy storage and thermal energy storage are adopted, in conjunction with long-timescale hydrogen energy storage devices.

[0247] Distribution network dispatching strategies under different scenarios of market price fluctuations, such as Figures 3-5 As shown. Analysis Figure 3 (a) Figure 4 (a) and Figure 5As shown in (a), from a week-to-week perspective, the first six days of the first week are within the dispatch cycle where the distribution network has a low day-ahead contract price for electricity in the power market, and the microgrid's renewable energy output is surplus. During this time, the distribution network purchases a large amount of electricity from the microgrid and the power market and converts it into hydrogen for storage through hydrogen energy storage. In the second and third weeks, when the renewable energy output of the microgrid is insufficient, the distribution network sells electricity to the microgrid. In the last three days of the third week, the distribution network sells electricity to the power market during the real-time high electricity purchase price period to obtain the price difference revenue. From an intraday perspective, the distribution network's electrochemical energy storage utilizes its high charge and discharge efficiency to repeatedly charge and discharge within 24 hours to balance the intraday source-load imbalance of the microgrid; thermal energy storage releases heat during the intraday high heat price period and sells heat to the heat market.

[0248] analyze Figure 3 (b) Figure 4 (b) and Figure 5 As shown in (b), due to the short-term charge-discharge balance cycle of electrochemical energy storage, in Scenario 1, electrochemical energy storage cannot absorb as much electricity as possible in the first week when the microgrid's renewable energy generation is high and the distribution network's day-ahead contract electricity purchase price is low. Therefore, it cannot mitigate the inter-weekly microgrid source-load imbalance through inter-weekly energy transfer and obtain the price difference revenue from electricity market transactions. Conversely, because the real-time electricity sales price of the distribution network to the electricity market is low in the first week, lower than its purchase price from the microgrid, the distribution network cannot absorb the surplus renewable energy of the microgrid, resulting in power curtailment in the microgrid. In the second and third weeks, the day-ahead contract electricity purchase price and the real-time electricity purchase price of the distribution network to the electricity market are high, and the renewable energy output of the microgrid cannot meet the load demand. Therefore, the distribution network can only purchase electricity from the electricity market at a high purchase price to meet the microgrid load demand, resulting in low price difference revenue.

[0249] Figure 6 For scenario 2, the coordinated charging and discharging strategy of energy storage at all levels, such as Figure 6 As shown, electrochemical energy storage utilizes its high charge and discharge efficiency to mitigate the imbalance of source and load output within the system during the day; limited by the lower charge and discharge efficiency of hydrogen energy storage systems, hydrogen energy storage only performs energy transfer on a long time scale, mitigating the imbalance of source and load output between days and weeks, and using its low self-discharge rate to trade with the electricity market across long time scales to obtain price difference profits; thermal energy storage releases heat during periods of high heat price each day, selling heat to the heat market to obtain profits.

[0250] Table 1 Comparison of system economics under different scenarios

[0251]

[0252] Table 1 compares the system's economic efficiency under different scenarios. As shown in Table 1, in Scenario 2, based on hybrid energy storage coordinated scheduling, the distribution network revenue is higher than in Scenario 1, the microgrid operating cost is lower than in Scenario 1, and the renewable energy absorption rate reaches 100%, avoiding the cost of wind and solar curtailment borne by the microgrid. Therefore, long- and short-cycle hybrid energy storage coordinated scheduling is beneficial to the economic operation of distribution-microgrid systems.

[0253] Advantage analysis of joint price uncertainty set for scenario 2 (the method in this embodiment):

[0254] Table 2 Comparison of Distribution Network Revenue under Two Uncertainty Sets

[0255]

[0256] Table 2 compares the revenue of the distribution network under two uncertainty sets. As shown in Table 2... Within the range of variation, DRO (Distributed Optimization of the Distribution Sphere) based on joint price distribution is more profitable and economical than independent DRO distribution networks. This is because the purchase price and sales price of electricity often have a coupling relationship on the same day or different days. Compared to splitting the real-time purchase price of electricity in the electricity market and the day-ahead contract purchase price of electricity in the distribution network into two independent uncertain sets, using a joint Wasserstein sphere to characterize the uncertainty of the two can maintain this coupling structure, avoid taking the worst-case scenario for both prices at the same time, and make the price combination in the worst-case scenario closer to the real market.

[0257] Analysis of Distribution-Microgrid Collaborative Pricing Strategy: Based on the daily and weekly energy transfer characteristics of hydrogen energy storage, one day is selected from each week within a 3-week dispatch cycle as a typical day. The advantages of continuous pricing within the distribution network dispatch cycle compared to independent pricing between days are analyzed. The distribution network pricing strategy within the dispatch cycle is as follows: Figures 7-9 As shown.

[0258] analyze Figures 7-9It can be seen that the pricing principles of the distribution network are as follows: During periods of high electricity purchase from microgrids, the distribution network sets the purchase price at or near the lower limit; during periods of low electricity purchase from microgrids, the distribution network sets the purchase price at or near the upper limit. Similarly, during periods of high electricity sales to microgrids, the distribution network sets the sales price at or near the upper limit; during periods of low electricity sales to microgrids, the distribution network sets the sales price at or near the lower limit. Compared with independent pricing, which uses daily transaction volume as the benchmark, continuous pricing uses transaction volume over the entire dispatch cycle as the benchmark, thus improving the average adjustment margin. Taking the distribution network purchase price as an example, the proportion of the price at the upper limit is higher on the first day of the third week (when purchase volume is lowest) than with independent pricing, thus reserving an average margin to increase the proportion of the price at the lower limit on the second day of the first week and the fifth day of the second week (when purchase volume is higher). During this period, the proportion of the lower limit for the continuous pricing method is higher than that for the independent pricing method. The same applies to the distribution network sales price. Independent pricing cannot adjust the upper and lower limits of pricing based on the average price margin of cross-day trading volume, lacking cross-day price adjustment capabilities and exhibiting poor flexibility. Table 3 shows a comparison of distribution network revenue under different pricing methods. As can be seen from Table 3, continuous pricing improves distribution network revenue.

[0259] Table 3 Comparison of Distribution Network Revenue under Different Pricing Methods

[0260]

[0261] Microgrid interruption strategies under continuous pricing: Figures 10-12 As shown. From an intraday perspective, microgrids set interruption prices at the lower limit during periods of high interruption load and at the upper limit during periods of low interruption load. From a cross-day perspective, microgrids adjust interruption pricing based on the average price margin under long-term interruption loads. The lower limit pricing ratio is higher on the second day of the first week and the first day of the third week when interruption loads are high than on the fifth day of the second week when interruption loads are low. In addition, influenced by the distribution network's pricing decisions, microgrids interrupt loads to reduce their operating costs during periods when their renewable energy output is not surplus and the distribution network's electricity sales price is higher than the interruption price, and during periods when their renewable energy output is surplus and the distribution network's electricity purchase price is higher than the interruption price. During periods when their renewable energy output is surplus and the distribution network's electricity purchase price is lower than their interruption price, the microgrid cannot profit, and the interruption load is 0.

[0262] Example 2: A distribution-microgrid collaborative trading device with hybrid long- and short-cycle energy storage, comprising:

[0263] One or more processors; a memory for storing one or more computer programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the method of embodiment 1.

[0264] Example 3: A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, implement the method in Example 1.

Claims

1. A method for coordinated and interactive trading between distribution and microgrids, incorporating hybrid energy storage with both long and short lifecycles, characterized in that... Includes the following steps: S1. Construct a distribution-microgrid collaborative architecture that includes hybrid energy storage with both long and short cycles. This includes an upper layer where the distribution network is the main stakeholder and a lower layer where the microgrid is the main stakeholder. The upper layer, the distribution network, includes distribution network agents, heat recovery devices, short-cycle energy storage, and long-cycle energy storage. The lower layer, the microgrid, includes wind power, users, and photovoltaics. Short-cycle energy storage in power distribution networks includes electrochemical energy storage and thermal energy storage, while long-cycle energy storage includes electrolyzers, hydrogen storage tanks, and fuel cells. S2. Considering the interaction between the distribution network and the electricity market, heat market, and microgrid, and taking the maximization of distribution network agent profits as the optimization objective, construct the objective function and its constraints for the upper-level model. The objective function of the upper-level model is: （1）； In the formula, For revenue generated from transactions between the distribution network and the microgrid; To generate revenue from real-time electricity market transactions between the distribution network and the power market; For revenue generated from transactions between the power distribution network and the heat market; The day-ahead electricity purchase cost for the distribution network and the electricity market; For energy storage operating costs; The constraints of the upper-level model include energy balance constraints, power purchase and sale price constraints between the distribution network and the microgrid, and power market transaction constraints. Short-cycle energy storage device models, long-cycle energy storage models, and heat recovery device models are also established. Represented as: （6）； In the formula, , , , , These are the operating costs of electrochemical energy storage, hydrogen energy storage, thermal energy storage, fuel cells, and electrolyzers, respectively. The energy balance constraint is: In the scenario of power distribution network selling electricity to microgrid, the power purchased by the power distribution network from the electricity market on a day-ahead basis, the power purchased by the power distribution network from the electricity market in real time, the power purchased by the power distribution network from the microgrid, the power discharged by short-cycle energy storage, and the power generated by fuel cells together constitute the power input. The power sold by the power distribution network to the microgrid, the power charged by short-cycle energy storage, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal. In the scenario of the distribution network selling electricity to the electricity market, the power purchased by the distribution network from the electricity market on the day-ahead, the power purchased by the distribution network from the electricity market in real time, the power purchased by the distribution network from the microgrid, the power of short-cycle energy storage discharge, and the power generated by fuel cells together constitute the power input. The power sold by the distribution network to the electricity market, the power of short-cycle energy storage charging, and the power consumed by the electrolyzer to produce hydrogen together constitute the power output. The power input and power output are equal. The sum of hydrogen production from the electrolyzer and hydrogen release from long-cycle hydrogen storage equals the sum of hydrogen charging from long-cycle hydrogen storage and hydrogen consumption from fuel cells. The sum of the heat release power of the heat recovery device and the heat release power of the thermal energy storage is equal to the sum of the heat charging power of the thermal energy storage and the heat selling power to the heat market. S3. Establish a joint price uncertainty set based on Wasserstein distance to characterize changes in electricity market prices. The establishment process is as follows: S3.1 Establish day-ahead contract electricity price based on the scheduling period t on day d. The real-time electricity price of the distribution network to the electricity market during the dispatching period on day d. Joint price support set Historical price data support set : （23）； （24）； In the formula, c represents a price pair containing , ; , These represent the upper and lower limits of the day-ahead contract power purchase price of the distribution network and the electricity market during the dispatching period on day d; , These represent the upper and lower limits of the real-time electricity sales price from the distribution network to the electricity market during the dispatching period on day d; This represents the l-th price sample of the distribution network during dispatch period t, including , , This represents the l-th price sample of the day-ahead contract power purchase between the distribution network and the electricity market during the t-schedule period. The l-th price sample for the real-time electricity sale from the distribution network to the electricity market during the t-schedule period; S3.

2. Obtain the empirical probability distribution based on the N sets of historical data. To estimate the true distribution Q, where To concentrate on The Dirac measure; described using 1-Wasserstein distance based on the joint price distribution. Distance between Q and : （25）； In the formula, inf represents the infimum; Represents the 1-norm; Indicates based on and Q distribution The joint distribution of c; S3.3, Define the joint price uncertainty set as: （26）； In the formula, express All probability distributions on; Indicated by With the center of the ball, A Wasserstein sphere with radius . The method for determining it is as follows: （27）； （28）； In the formula, G is a coefficient related to the data distribution; Confidence level; This is the adjustment coefficient; express The mean of the sample; e is the natural constant; n is the sample index; S4. With the goal of minimizing the microgrid operating cost, construct the objective function and its constraints for the lower-level model. The objective function of the lower-level model is: （29）； In the formula, Compensation payments made to users for load interruptions in microgrids; Costs associated with curtailing wind and solar power; S5. Perform non-convex term transformation and constraint standardization on the lower-level model, then transform the lower-level model into additional constraints of the upper-level model, linearize the objective function of the upper-level model, optimize the linearized objective function based on the joint price uncertainty set during the master-slave game between the upper and lower-level models, and solve the optimized objective function to obtain the optimal distribution-microgrid collaborative interaction trading scheme. The objective function for optimizing linearization based on the joint price uncertainty set is as follows: In the master-slave game process of the upper and lower level models, after obtaining the first-stage optimization variables, the second stage optimizes the expectation under the uncertainty scenario to obtain the price distribution under the extreme scenario, and then optimizes the linearization objective function for the price distribution under the extreme scenario. The price distribution in extreme scenarios is represented as follows: （32）； （33）； In the formula, C, x, and p represent the optimization variables in the first stage; X is the feasible region in the first stage. Expressing expectations; This represents the objective function in a single scenario; D is the total number of days, and T is the total number of scheduling periods. This represents the real-time power output of the distribution network to the electricity market during the dispatching period on day d. The day-ahead contract power for the scheduling period t on day d; According to the strong duality principle, the expectation can be transformed into the following form: （34）； In the formula, is the dual variable; l is the index of the sample, and each sample contains a microgrid; sup denotes the supremum; Introducing intermediate variables , , ,make: （35）； （36）； Then equation (32) is transformed into the following form: （37）； （38）。 2. The distribution-microgrid collaborative interaction trading method with hybrid long- and short-cycle energy storage according to claim 1, characterized in that, Represented as: （2）； In the formula, , These represent the electricity sales and purchase prices from the distribution network to the microgrid during the dispatching period t on day d. , These represent the power sold and purchased by the distribution network to the microgrid during the dispatching period t on day d. For time intervals; Represented as: （3）； In the formula, The real-time electricity purchase price from the power market by the distribution network during the dispatch period on day d; The power purchased by the distribution network from the electricity market in real time during the dispatching period on day d; Represented as: （4）； In the formula, The real-time heat price for the distribution network to sell heat to the heat market during the dispatch period on day d; This represents the real-time heat sales capacity of the distribution network to the heat market during the dispatching period on day d. Represented as: （5）。 3. The distribution-microgrid collaborative interaction trading method with hybrid long- and short-cycle energy storage according to claim 2, characterized in that, The price constraint for electricity purchase and sale from the distribution network to the microgrid is: （7）； （8）； （9）； （10）； In the formula, , These represent the upper and lower limits of the electricity price sold from the distribution network to the microgrid during the dispatching period t on day d. , These represent the upper and lower limits of the electricity purchase price from the microgrid by the distribution network during the t-period dispatch on day d; , These are the power sales and purchase flags for the distribution network during the dispatching period t on day d. When the distribution network sells electricity to the microgrid during the dispatching period t on day d... Otherwise, it equals 0. Similarly; The sum of the difference between the total revenue from the real-time sale of electricity from the distribution network to the electricity market and the total cost of purchasing electricity from the microgrid is greater than or equal to 0. The constraints on power distribution network and electricity market transactions are: （11）； （12）； （13）； （14）； In the formula, This refers to the maximum daily power purchase capacity of the distribution network under contract with the electricity market. , This refers to the maximum real-time power sales and purchase capacity of the distribution network and the electricity market. , These are the power sales and purchase flags for the dispatching period t on day d, respectively. When the distribution network sells electricity to the electricity market during the dispatching period t on day d... Otherwise, it equals 0. Similarly.

4. The distribution-microgrid collaborative interaction trading method with hybrid long- and short-cycle energy storage according to claim 2, characterized in that, Short-cycle energy storage device models include electrochemical energy storage models and thermal energy storage models. The electrochemical energy storage model is represented as follows: （15）； （16）； In the formula, , These represent the stored energy of electrochemical energy storage during scheduling periods t and t-1 on day d, respectively. The self-discharge rate of electrochemical energy storage; To improve the charge and discharge efficiency of electrochemical energy storage; , These represent the electrochemical energy storage charging and discharging power during the t-period scheduling on day d, respectively. , These represent the stored energy of electrochemical energy storage during scheduling periods 1 and 24 on day d, respectively. , These represent the electrochemical energy storage charging and discharging power during the first scheduling period on day d; Similarly, a thermal energy storage model is constructed; Long-term energy storage models include: Hydrogen energy storage device model: （17）； （18）； （19）； In the formula, , These represent the hydrogen storage capacity during the hydrogen storage period on day d and t-1, respectively. , These represent the hydrogen storage charging and discharging amounts during the t-period scheduling time on day d. To improve the efficiency of hydrogen charging and discharging for hydrogen energy storage; This represents the amount of hydrogen stored in the hydrogen storage system during the first scheduling period on day d+1. This represents the amount of hydrogen stored in the hydrogen storage system during the 24-hour scheduling period on day d. , Similarly; , These represent the hydrogen storage charging and discharging amounts for the first scheduling period on day d+1. , Similarly; Electrolyzer-Fuel Cell Model: （20）； （21）； In the formula, The amount of hydrogen produced by the electrolyzer during the scheduling period t on day d; The power consumption of the electrolytic cell during the scheduling period t on day d; The fuel cell power generation during the t-th scheduling period on day d; This represents the hydrogen consumption of the fuel cell during the t-day scheduling period on day d. , These are the efficiencies of the electrolyzer and the fuel cell, respectively. The heat recovery device model is represented as follows: （22）； In the formula, The heat recovered by the heat recovery device during the t-day scheduling period on day d; , These refer to the efficiency of the heat recovery device in recovering waste heat from the electrolyzer and hydrogen fuel cell, respectively.

5. The distribution-microgrid collaborative interaction trading method with hybrid long- and short-cycle energy storage according to claim 2, characterized in that, In S4, Represented as: （30）； In the formula, The unit electricity compensation price paid by the microgrid to users during the t-schedule period on day d; The interrupted user load power during the t-period scheduling on day d; Represented as: （31）； In the formula, Cost per unit of wind and solar curtailment; The power of wind and solar power curtailed by the microgrid during the t-th dispatch period on day d; The constraints of the lower-level model include microgrid power balance constraints, microgrid power purchase and sale to the distribution network constraints, microgrid interruptible load constraints, and microgrid wind and solar power curtailment constraints.

6. The distribution-microgrid collaborative interaction trading method with hybrid long- and short-cycle energy storage according to claim 1, characterized in that, In S5, the optimized distribution-microgrid collaborative interaction trading scheme is as follows: the distribution network formulates a day-ahead contract power purchase strategy based on the source and load information provided by the microgrid, the maximum interruptible load declared by the microgrid users, and the day-ahead price signal of the power market, and completes the power interaction with the power market. When the daily contracted electricity volume cannot meet the microgrid's demand during real-time trading, the distribution network and the electricity market conduct real-time trading to supplement the shortfall. When the distribution network interacts with the microgrid to generate surplus electricity, the distribution network sells electricity to the microgrid during periods of insufficient renewable energy output, periods of high real-time electricity purchase price in the electricity market, and periods of high real-time heat purchase price in the heat market through the intraday and interday energy transfer characteristics of hybrid energy storage, so as to complete the transfer of electricity price. Meanwhile, the distribution network formulates a real-time, continuous, and dynamic pricing strategy based on the energy consumption of the microgrid to ensure its own maximum revenue; the microgrid adjusts its own electricity consumption ratio in real time based on the distribution network's pricing strategy to decide on the electricity purchase and sale strategy and interruption strategy that are beneficial to itself, and to minimize its own operating costs.

Images