Distribution network-microgrid electric energy interactive transaction method considering electric power market risk

Abstract

The invention belongs to the technical field of electric energy transaction, and particularly relates to a distribution network-microgrid electric energy interactive transaction method considering electric power market risks, which comprises the following steps: constructing a distribution network-microgrid collaborative architecture comprising an upper layer based on a distribution network and a lower layer based on multiple microgrids; with the maximum daily profit of a distribution network agent as a target, constructing a target function and constraint conditions of an upper-layer model; taking the minimum daily comprehensive operation cost of the multi-microgrid system as a target, and constructing a target function and constraint conditions of a lower-layer model; and converting the lower-layer model into an additional constraint condition of the upper-layer model, obtaining a converted upper-layer model objective function, and solving the converted upper-layer model objective function to obtain an optimized distribution network-microgrid collaborative interaction electric energy transaction scheme. Through a dynamic pricing mechanism, an energy storage transaction mode division mechanism and a load interruption mechanism, benefit balance and risk sharing of the distribution network and the microgrid under power market fluctuation are realized, the new energy consumption rate can be improved, and the system economy and flexibility are enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric energy trading, and particularly relates to a distribution network - microgrid electric energy interactive trading method considering power market risks. Background Art

[0002] Currently, the coordinated scheduling and operation of the distribution network and the microgrid are mainly divided into centralized optimal operation and distributed optimal operation. However, the traditional centralized optimal operation is difficult to meet the demand for coping with the changes in external power market trading prices under the background of high penetration of renewable energy. Therefore, it is urgent to study the distribution network - microgrid distributed optimal scheduling control theory and method.

[0003] The distribution network can interact with the microgrid by setting up energy storage. When the power generation of distributed power sources in the microgrid exceeds the load demand, the surplus power is sold to the distribution network to avoid the cost of curtailment of wind and light. At the same time, the distribution network stores the purchased electricity in its own energy storage. When the purchase price in the external power market is higher than the purchase price from the microgrid, the energy storage of the distribution network discharges and sells electricity to the power market to complete the electricity price transfer. The distribution network - microgrid collaborative interaction balances the interests of both parties.

[0004] However, the existing research does not consider the impact of real - time price fluctuations in the power market on the distribution network - microgrid transaction, and lacks a dynamic pricing and external risk coupling mechanism; the existing research encourages emerging market players to directly participate in the power market transaction, focusing on the P2P flat trading mode. However, when there is a deviation between the contract electricity quantity and the actual electricity consumption, there is a risk of loss of interest for both trading parties, ignoring the buffering role of the distribution network under the hierarchical market structure, resulting in limited risk - spreading ability; the existing research ignores the impact of energy storage on system economy under the fluctuation of external market trading prices, and does not deeply explore the potential of energy storage to avoid the risks of external power market price changes and balance the interests of both trading parties; the existing research implements fixed pricing or quantity - based pricing for the interruption load compensation price, ignoring the changes in external market conditions and the impact of price variables other than the compensation price on the demand side, resulting in insufficient flexibility for the demand side to participate in electric energy trading and not deeply exploring the potential of user willingness and two - way interaction of electric energy resources. Summary of the Invention

[0005] In view of the above deficiencies in the prior art, the purpose of the present invention is to provide a distribution network - microgrid electric energy interactive trading method considering power market risks. Through a dynamic pricing mechanism, dividing the energy storage trading mode and the interruption load mechanism, it realizes the interest balance and risk sharing between the distribution network and the microgrid under the power market fluctuation, can improve the new energy consumption rate, and enhance the economy and flexibility of the system.

[0006] To achieve the above - mentioned purpose, the present invention provides a distribution network - microgrid electric energy interactive trading method considering power market risks, including the following steps: S1. Build a coordinated distribution network - microgrid architecture, including an upper layer based on the distribution network and a lower layer based on multiple microgrids. The upper layer architecture includes a distribution network agent, a distribution network control center, and an energy storage system composed of distribution network energy storage and shared energy storage. Each microgrid in the lower layer contains wind power, users, photovoltaic power, and a microgrid control center. Each microgrid forms a multi - microgrid system; S2. With the goal of maximizing the daily profit of the distribution network agent, construct the objective function and its constraints of the upper - layer model; S3. With the goal of minimizing the daily comprehensive operating cost of the multi - microgrid system, construct the objective function and its constraints of the lower - layer model; S4. Transform the lower - layer model into additional constraints of the upper - layer model, obtain the objective function of the transformed upper - layer model and solve it to obtain an optimized coordinated distribution network - microgrid interactive power trading scheme. In this scheme: In the day - ahead stage, the microgrid reports the maximum amount of its interruptible load to the microgrid control center. The microgrid control center reports its new - energy output and load information to the distribution network control center. The distribution network formulates a day - ahead contract power purchase strategy based on the information reported by the microgrid; In the real - time stage, the distribution network formulates distribution network pricing information for real - time power purchase and sale to the microgrid according to real - time power market price information, as well as power trading strategies and distribution network energy storage capacity division strategies, and transmits the power market price information and distribution network pricing information to the lower - layer microgrid. The microgrid adjusts its interruptible load pricing strategy and power consumption strategy according to the information transmitted by the distribution network and feeds back to the distribution network in real - time; According to the feedback from the microgrid, on the basis of the energy storage capacity division strategy, the distribution network makes a secondary adjustment to the power purchase and sale pricing of the microgrid's new - energy output.

[0007] As a preferred solution of the present invention, in S2, the upper - layer interest subject is the distribution network agent, considering the operating income of the distribution network shared energy storage 、the income from the distribution network's real - time power sale to the microgrid 、the income from the distribution network agent's real - time power sale to the power market 、the cost of the distribution network agent's real - time power purchase from the power market 、the cost of the distribution network agent's day - ahead power purchase from the power market 、the cost of the distribution network agent's deviation declaration for the next day's transaction with the power market , the objective function of the upper - layer model is: (1); The upper-layer model constraints include the interactive power balance constraint between the distribution network and the microgrid, the interactive power balance constraint of the shared energy storage, the charge and discharge constraints of the shared energy storage, the charge and discharge constraints of the distribution network energy storage, the maximum charge and discharge power distribution constraint of the energy storage, the state of charge constraint of the shared energy storage, the state of charge constraint of the distribution network energy storage, the energy storage capacity distribution constraint, the electricity selling price constraint from the distribution network to the microgrid, the electricity selling and buying price constraints of the shared energy storage to the microgrid, the trading constraint between the distribution network agent and the upper-level power market, and the interest constraint of the distribution network agent.

[0008] As a preferred embodiment of the present invention, The calculation method of is: In the formula, t is the index of the scheduling period, and T is the total number of scheduling periods; i is the index of the microgrid, and N is the total number of microgrids; and are respectively the electricity selling price and the electricity buying price of the shared energy storage to the microgrid i in the t-th scheduling period; and are respectively the electricity selling power and the electricity buying power of the shared energy storage to the microgrid i in the t-th scheduling period; is the time interval; The calculation method of is: In the formula, is the electricity selling price from the distribution network to the microgrid i in the t-th scheduling period; is the electricity selling power from the distribution network to the microgrid i in the t-th scheduling period; and The calculation method of is: is: In the formula, and are respectively the real-time electricity selling price and the real-time electricity buying price of the distribution network agent to the power market in the t-th scheduling period; and are respectively the real-time electricity selling power and the real-time electricity buying power of the distribution network to the power market in the t-th scheduling period; is the real-time electricity selling power of the distribution network shared energy storage to the power market in the t-th scheduling period; The calculation method of is: In the formula, is the day-ahead contract price in the t-th scheduling period; is the day-ahead benchmark contract power in the t-th scheduling period; and are the positive and negative deviations of the day-ahead contract declaration in the t scheduling period, respectively; The power market recovers the declared deviation part according to the corresponding settlement income. The calculation method is: (7); (8); In the formula, , are the positive and negative deviation prices of the next-day declaration, respectively; , are the positive and negative deviation settlement income coefficients of the declaration, respectively.

[0009] As an optimal solution of the present invention, each constraint condition of the upper-layer model is specifically: Power balance constraint for the interaction between the distribution network and the microgrid: (9); In the formula, , are the charging and discharging powers of the distribution network energy storage in the t scheduling period, respectively; Shared energy storage interaction power balance constraint. During the period when the microgrid purchases and sells electricity from the shared energy storage, the charging and discharging power of the shared energy storage is determined by the total energy demand after the energy exchange at each microgrid bus. Its constraint is: (10); (11); (12); In the formula, , are the charging and discharging powers of the shared energy storage in the t scheduling period, respectively; , is the power that the shared energy storage purchases from the microgrid i and sells to the power market and the power-deficient microgrid during the t scheduling period; , is the charging power that the shared energy storage uses to sell to the power market and the power-deficient microgrid during the t scheduling period; , is the discharging power that the shared energy storage uses to sell to the power market and the power-deficient microgrid during the t scheduling period; Shared energy storage charging and discharging constraint: (13); In the formula, , are the charging and discharging flag bits of the shared energy storage in the t scheduling period, respectively, with values of 0 or 1. A value of 1 indicates that charging and discharging are in progress. The meanings of the other flag bits are the same; is the maximum charging and discharging power of the shared energy storage; Charging and discharging constraints of distribution network energy storage: (14); In the formula, and are the charging and discharging powers of the distribution network energy storage in the t - scheduling period respectively; and are the charging and discharging flag bits of the distribution network energy storage in the t - scheduling period respectively; is the maximum charging and discharging power of the distribution network energy storage; Maximum charging and discharging power distribution constraints of energy storage: (15); In the formula, is the total maximum charging and discharging power of the energy storage; Charging state constraints of shared energy storage: (16); (17); In the formula, and are the charging states of the shared energy storage in the t - and (t - 1)-scheduling periods respectively; is the charging and discharging efficiency of the energy storage; and are the upper and lower limit coefficients of the charging state of the energy storage respectively; is the maximum stored energy of the shared energy storage; Charging state constraints of distribution network energy storage: (18); (19); In the formula, and are the charging states of the distribution network energy storage in the t - and (t - 1)-scheduling periods respectively; is the maximum stored energy of the distribution network energy storage; Stored energy distribution constraints: (20); In the formula, is the total maximum stored energy; Selling price constraints for the distribution network to sell electricity to the microgrid: (21); (22); (23); In the formula, and are the upper and lower limits of the selling price for the distribution network to sell electricity to the microgrid in the t - scheduling period respectively; The flag bit of power sales from the distribution network to microgrid i during the dispatching period t; is the maximum number of electricity sales from the distribution network to microgrid i; The price constraints for shared energy storage to sell electricity to micro-grids are as follows: (twenty four); (25); In the formula, , They are the upper and lower limits of the electricity price sold by shared energy storage to the microgrid during the dispatch period t; , They are the upper and lower limits of the electricity purchase price from the shared energy storage to the microgrid during the dispatch period t; At the same time, the electricity purchase and sales price of shared energy storage is coupled with the electricity purchase price of the external power market to protect the interests of both the distribution network and the microgrid: (26); (27); (28); (29); In the formula, , They are the flags of electricity sold and purchased by the shared energy storage to microgrid i during the dispatching period t; It is the real-time electricity purchase coefficient of the power market corresponding to the full consumption of new energy in the microgrid when no energy storage is set; is the real-time electricity purchase coefficient in the current electricity market; , are the price protection coefficients for selling and purchasing electricity respectively; , are the maximum number of times the shared energy storage sells and purchases electricity from microgrid i; Distribution network agents and upper-level power market transaction constraints: (30); (31); (32); (33); (34); (35); (36); (37); In the formula, The maximum power purchased by the distribution network in the day-ahead contract with the power market; , are the purchase and sale electric power flag bits for the t scheduling period respectively; , are the maximum real-time purchase and sale electric powers of the distribution network and the power market respectively; represents or ; represents or , which are the maximum positive and negative deviation powers declared for the next day of the distribution network and the power market transaction respectively; , are the positive and negative deviation status flag bits for the t scheduling period respectively, represents or ; represents or , which are the positive and negative deviation coefficients respectively; Interest constraint for the distribution network agent. When the real-time power purchase coefficient in the external power market is not equal to the reference value, there is a risk of loss of interest in the distribution network's consumption of new energy from the microgrid. To ensure that the interests of the distribution network are not damaged and to prevent the operation cost of the microgrid from rising due to unreasonable secondary pricing of the distribution network, the following constraints are set: (38); In the formula, , are the upper and lower limits of the incremental revenue after secondary pricing respectively; is the power purchase price of the distribution network from microgrid i during the t scheduling period.

[0010] As a preferred solution of the present invention, in S3, the lower-level interest subject is the multi-microgrid system, considering the cost of the microgrid using shared energy storage services , the power purchase cost of the microgrid from the distribution network , the compensation cost paid by the microgrid's interruptible load to users , the cost of wind and light abandonment , the objective function of the lower-level model is: (39); The constraint conditions of the lower-level model include the microgrid power balance constraint, the microgrid power purchase power constraint from the distribution network, the microgrid purchase and sale power constraints with the shared energy storage, the microgrid internal interruptible load constraint, and the microgrid wind and light abandonment constraint.

[0011] As a preferred solution of the present invention, The calculation method of (40); The calculation method of (41); The calculation method of is: In the formula, is the compensation price per unit of electricity paid by Microgrid i to users during the t scheduling period; is the interrupted load power of Microgrid i during the t scheduling period; The calculation method of is: In the formula, is the unit cost of wind and light curtailment; is the wind and light curtailment power of Microgrid i during the t scheduling period.

[0012] As a preferred solution of the present invention, each constraint condition of the lower layer model is specifically: Microgrid power balance constraint: (44); In the formula, , , are the wind power, photovoltaic power, and load output of Microgrid i during the t scheduling period, respectively; is the Lagrange multiplier of the equality constraint; In the optimization problem, when the objective function reaches an extreme value under the condition of satisfying the equality constraint, the Lagrange multiplier of the equality constraint reflects the sensitivity of the objective function to the constraint condition, helps to incorporate the constraint condition into the optimization process, and guides the adjustment direction of the optimization variable to find the optimal solution that satisfies the constraint.

[0013] Microgrid power purchase from the distribution network constraint: (45); In the formula, is the maximum power purchase from the distribution network by the microgrid; , are the Lagrange multipliers corresponding to the inequality constraints; is the power selling flag of Microgrid i to the distribution network during the t scheduling period; The role of the colon is to indicate that , are the Lagrange multipliers related to this inequality constraint, and , are non - negative.

[0014] Corresponding to the lower - bound constraint , when is close to or equal to 0, this constraint condition is active, and the corresponding Lagrange multiplier will be non-zero if is non-zero, indicating that during the optimization process, is restricted by the lower bound constraint.

[0015] Corresponding to the upper bound constraint when is close to or equal to , this constraint is active, and the corresponding Lagrange multiplier will be non-zero if is non-zero, indicating that during the optimization process, is restricted by the upper bound constraint.

[0016] The same applies to the remaining Lagrange multipliers in the following formula.

[0017] Power purchase and sale power constraints between the microgrid and the shared energy storage: (46); (47); (48); (49); In the formula, , are the maximum power purchase and sale between the microgrid and the shared energy storage respectively; , are the power purchase and sale flag bits of the microgrid i to the shared energy storage during the t scheduling period respectively; , , , , , are the Lagrange multipliers corresponding to the inequality constraints; Interruptible load constraints within the microgrid: (50); (51); (52); (53); In the formula, is the maximum power of the interruptible load of the microgrid i during the t scheduling period; , are the upper and lower limits of the interruptible load compensation price of the microgrid during the t scheduling period respectively; is the interruptible load flag bit of the microgrid i during the t scheduling period; is the maximum number of interruptions of the interruptible load of the microgrid i; is the interrupt pricing protection coefficient; , , , , , are the Lagrange multipliers corresponding to the inequality constraints; Wind and PV curtailment constraints in the microgrid: (54); In the formula, is the maximum value of wind and PV curtailment in the microgrid; , are the Lagrange multipliers corresponding to the inequality constraints.

[0018] As a preferred solution of the present invention, in the said S4, converting the lower-layer model into an additional constraint condition of the upper-layer model to obtain the objective function of the converted upper-layer model, the process includes: S4.1. Conversion of non-convex terms in the lower-layer model: Since , are non-convex, use the exponential method to convert them into convex function forms. Let , , , be the logarithmic conversion forms corresponding to , , then the lower-layer model is converted to: (55); (56); (57); (58); In the formula, , ; represents the lower limit; ln is the natural logarithm with base e; Introduce an auxiliary variable , and convert the non-convex constraint to: (59); (60); (61); In the formula, , , are the Lagrange multipliers corresponding to the inequality constraints; M is a constant in the big M method (a very large positive value); S4.2. Standardization of the lower-layer model: Standardize the inequality constraints and equality constraints of the lower-layer model: (62); (63); Wherein, represents an inequality constraint; represents an equality constraint; , , respectively represent the power, price, and flag bit variables in the constraint; Construct the Lagrangian function: (64); , are the Lagrange multipliers of the equality constraint and the inequality constraint respectively; x and y are the numbers of the equality constraint and the inequality constraint respectively; represents the constructed Lagrangian function; S4.3, Lower-layer model transformation: Using the KKT method, take the partial derivative of the decision variables of the lower-layer model based on Equation (64), and transform the lower-layer model into an additional constraint condition of the upper-layer model: (65); Wherein, l represents the complete set of all decision variables in the equality and inequality constraints; The non-linear constraint in Equation (65) is transformed into the following linear form: (66); Wherein, is a Boolean variable; In the upper-layer model, the constraints of Equations (13)-(14), Equation (22), and Equations (26)-(27) are non-convex. For Equations (13)-(14), they are transformed into the following form: (67); (68); Wherein, is a maximum value; j represents abs or relea, and the corresponding are respectively , , Similarly; k represents ch or dis, and the corresponding are respectively , , Similarly; For Equations (22) and (26)-(27), let: (69); Wherein, corresponding to is taken as , or , represents , or , respectively representing , 's product, , 's product, and , 's product; represents , or , similarly; then there is: (70); At this time, equations (22), (26), and (27) are transformed into the following convex constraints: (71); (72); (73); S4.4. The objective function of the transformed upper-layer model is: (74).

[0019] As a preferred solution of the present invention, in S4, the solution process is as follows: Step 1. Define , as the distribution network revenue and the multi-microgrid operation cost when respectively. C is the current revenue of the distribution network, and F is the current operation cost of the multi-microgrid; is the step size; , , The initial values are taken as 0; Step 2. Input the load and new energy data of each microgrid, generate the day-ahead contract volume of the distribution network under the real-time electricity purchase and sale benchmark price in the electricity market, and calculate the real-time electricity purchase coefficient of the distribution network corresponding to the full consumption of the new energy of the microgrid when the distribution network does not lose its own revenue and the energy storage does not participate in the transaction, that is ; Step 3. Input the current real-time electricity purchase coefficient , and calculate , based on the conventional energy storage scenario (the distribution network sets up energy storage, but does not divide the energy storage transaction with the microgrid, and the transaction price adopts a dynamic pricing mechanism); Step 4. According to and The size relationship, in , , on the basis of , is corrected as follows: Judge whether the current real-time power purchase coefficient in the power market is greater than or equal to : When , judge whether the current revenue C of the distribution network is less than . If so, according to , update , , until ; if not, retain the current , , and judge whether the current operation cost F of the multi-microgrid is greater than . If so, according to update , until ; if not, retain the current ; When , judge whether holds. If so, according to , update , , until ; if not, judge whether and hold. If so, retain the current , , if not, according to , update , ; Step Five: According to the corrected , , recalculate , , , obtain the optimized dispatching result, and execute the distribution network-microgrid collaborative interactive power trading plan based on this dispatching result.

[0020] As a preferred solution of the present invention, in the day-ahead stage of the distribution network-microgrid collaborative interactive power trading plan, the maximum amount of its own interruptible load refers to , its own new energy output refers to , , the load information refers to , and the day-ahead contract power purchase strategy refers to ; In the real-time stage, the electricity market price information refers to , , and the distribution network pricing information refers to , , , and the electricity trading strategy refers to , , , , , , , , and the distribution network energy storage capacity division strategy refers to , , , , and the information transmitted by the distribution network refers to , , , , , and the self-interruptible load pricing strategy refers to , and the electricity consumption strategy refers to , , , , and the purchase and sale pricing of new energy output in the microgrid refers to , .

[0021] The algorithm involved in the present invention can be executed by an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. The above algorithm calculation is realized by the processor executing software.

[0022] The beneficial effects of the present invention are as follows: The present invention introduces a dynamic pricing mechanism based on the principal-agent game principle, enabling the distribution network to flexibly adjust the selling and purchasing prices of electricity for the microgrid according to the real-time price fluctuations in the external electricity market. This dynamic pricing strategy not only significantly improves the revenue capacity of the distribution network but also provides clear price signals for the microgrid during electricity market price fluctuations, guiding it to optimize electricity consumption and power generation decisions. By coupling the purchase and sale prices of shared energy storage with the real-time external market prices, the economic loss risk that the distribution network may suffer due to accommodating new energy in the microgrid is effectively hedged, while safeguarding the interests of the microgrid, achieving an interest balance between the two parties under risk fluctuations.

[0023] The present invention proposes a method for dividing the energy storage trading mode, which subdivides the energy storage system into conventional energy storage for day-ahead arbitrage and shared energy storage for new energy accommodation. This refined energy storage management method gives full play to the role of energy storage in transferring the time-space value. On the one hand, the conventional energy storage purchases electricity at a low price during the valley price period and sells electricity at a high price during the peak price period, maximizing the arbitrage income; on the other hand, the shared energy storage purchases and stores electricity at a low price when new energy generation is excessive, and sells it at a high price when needed in the power market or microgrid. This not only avoids the phenomenon of abandoning wind and light, ensures the full accommodation of new energy, but also further improves the overall economic efficiency of the system.

[0024] The interruption load mechanism introduced by the present invention on the microgrid side endows the microgrid with the ability to flexibly adjust the electricity consumption strategy according to real-time price signals. By dynamically associating the interruption load compensation price with the real-time price of the power market, the microgrid can independently decide whether to interrupt some non-critical loads according to the economic principle, thereby reducing its own operating costs. This mechanism not only enhances the flexibility of the microgrid to cope with power market price fluctuations, but also significantly improves the willingness and flexibility of users to participate in electricity trading, transforming the microgrid from a passive price acceptor to an active participant in optimizing electricity consumption decisions, and further tapping the adjustment potential of the demand side. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is the flow schematic diagram of the present invention; Figure 2 is the schematic diagram of the solution process of the present invention; Figure 3 is the schematic diagram of the distribution network trading price in Scenario 3 during the verification process of the present invention; Figure 4 is the schematic diagram of the distribution network trading electricity volume in Scenario 3 during the verification process of the present invention; Figure 5 is the schematic diagram of the trading situation between MG1 and the distribution network and shared energy storage during the verification process of the present invention; Figure 6 is the schematic diagram of the trading situation between MG2 and the distribution network and shared energy storage during the verification process of the present invention; Figure 7 is the schematic diagram of the trading situation between MG3 and the distribution network and shared energy storage during the verification process of the present invention; Figure 8 is the schematic diagram of the trading volume and energy storage situation during the verification process of the present invention; Figure 9 is the schematic diagram of the clearing strategy for the distribution network to trade with the power market during the verification process of the present invention; Figure 10 is the schematic diagram of the distribution network revenue and energy storage allocation situation during the verification process of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0026] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings: As Figure 1 shown, the distribution network - microgrid electricity interactive trading method considering power market risks includes the following steps: S1. Construct a distribution network - microgrid collaborative architecture, including an upper layer based on the distribution network and a lower layer based on multiple microgrids. The upper layer architecture includes a distribution network agent, a distribution network control center, and an energy storage system composed of distribution network energy storage and shared energy storage. Each microgrid in the lower layer contains wind power, users, photovoltaic power, and a microgrid control center, and each microgrid forms a multi - microgrid system; S2. With the goal of maximizing the daily profit of the distribution network agent, construct the objective function and its constraint conditions of the upper layer model; S3. With the goal of minimizing the daily comprehensive operating cost of the multi - microgrid system, construct the objective function and its constraint conditions of the lower layer model; S4. Convert the lower layer model into an additional constraint condition of the upper layer model, obtain the objective function of the converted upper layer model and solve it to obtain an optimized distribution network - microgrid collaborative interactive electricity trading scheme.

[0027] The distribution network control center completes the electricity quantity interaction by signing a day - ahead power purchase contract with the power market according to the load, photovoltaic power, wind power prediction information provided by the microgrid and the maximum interruptible load declared by the microgrid users. When the electricity quantity of the day - ahead contract signed during the real - time trading process cannot meet the user's demand, the distribution network agent conducts real - time trading with the power market to make up for the shortage. When there is surplus electricity generated from the interaction between the distribution network, shared energy storage and the microgrid, the distribution network agent can sell the electricity to the power market in real - time to complete the electricity price transfer. At the same time, the upper - layer distribution network agent formulates a real - time dynamic pricing strategy according to the energy consumption situation and the real - time power purchase price in the external power market, including the distribution network power selling pricing strategy, the shared energy storage power purchase and selling pricing strategy, and the energy storage system allocation capacity strategy to ensure the maximization of its own benefits. The microgrid can interact with the distribution network and shared energy storage, and adjust its own electricity consumption ratio in real - time according to the real - time power purchase price in the external power market, the distribution network power selling price, and the shared energy storage power purchase and selling price to decide on a favorable power purchase and selling strategy and interruption strategy for itself, and minimize its own operating cost.

[0028] The revenue of the distribution network agent comes from the distribution network itself, shared energy storage, the lower-level microgrid, and arbitrage in the electricity market. Therefore, when there is surplus new energy in the microgrid, the microgrid can sell it to the distribution network, and then the distribution network energy storage device can sell it back to the microgrid and the electricity market to improve the economic efficiency of system operation. However, during the trading process, the real-time electricity purchase price in the electricity market will change due to market fluctuations. When the real-time electricity selling price of the distribution network to the electricity market is lower than the electricity purchase price of the distribution network from the microgrid, the interests of the distribution network will suffer losses. If, to make up for this loss, the electricity selling price of the distribution network is increased, then at this time, the secondary electricity selling price of new energy and the electricity selling price of the distribution network's day-ahead electricity purchase resold to the microgrid will also be increased. However, the loss of the distribution network's interests is caused by the distribution network's absorption of the excess new energy in the microgrid to avoid the microgrid bearing the costs of curtailed wind and solar power. Increasing all electricity selling prices will damage the interests of the microgrid. Therefore, to ensure the interests of both the distribution network and the microgrid, the distribution network allocates the capacity of the energy storage system. A part of the energy storage is used for charging and discharging the electricity purchased from the electricity market ahead of schedule, and the selling price of this part remains unchanged. Another part of the energy storage (shared energy storage) directly trades with the microgrid for charging and discharging new energy electricity, and the purchase and sale pricing of this part of the electricity is coupled with the real-time electricity purchase price in the external electricity market to make up for the losses borne by the distribution network for absorbing new energy in the microgrid.

[0029] In summary, the coordinated interactive electricity trading scheme between the distribution network and the microgrid is as follows: In the day-ahead stage, the microgrid reports the maximum amount of its interruptible load ( ) to the microgrid control center. The microgrid control center reports its new energy output ( , ) and load information ( ) to the distribution network control center. The distribution network formulates a day-ahead contract electricity purchase strategy ( ) based on the information reported by the microgrid; In the real-time stage, since the real-time electricity purchase and sale prices in the electricity market fluctuate due to market impacts, the distribution network formulates the distribution network pricing information ( , ) for real-time electricity purchase and sale with the microgrid, as well as an electricity trading strategy ( , , ), a distribution network energy storage capacity division strategy ( , , , , , , , ) based on the real-time electricity market price information ( , , , ), and transmit electricity market price information and distribution network pricing information to the lower-level microgrid. The microgrid adjusts its interruptible load pricing strategy according to the information transmitted by the distribution network ( ), electricity consumption strategy ( , , , ), and provide real-time feedback to the distribution network; Based on the feedback from the microgrid, the distribution network adjusts the purchase and sale pricing of new energy output from the microgrid on the basis of the energy storage capacity allocation strategy ( , ) to avoid losses to its own interests caused by the absorption of new energy from the microgrid by the distribution network.

[0030] Through real-time dynamic optimization of the information flow and energy flow, both parties obtain the optimal electricity price and the optimal trading electricity volume, achieve a win-win situation for both parties under the master-slave game, and jointly address the problem of system economic decline caused by real-time price changes in the electricity market.

[0031] In S2, the upper-level interest subject is the distribution network agent, considering the operating income of the distribution network's shared energy storage , the real-time electricity sales income from the distribution network to the microgrid , the real-time electricity sales income from the distribution network agent to the electricity market , the real-time electricity purchase cost of the distribution network agent from the electricity market , the day-ahead electricity purchase cost of the distribution network agent from the electricity market , the deviation cost of the distribution network agent's transaction declaration with the superior electricity market on the next day . The objective function of the upper-level model is: (1); The upper-level model constraint conditions include the interactive power balance constraint between the distribution network and the microgrid, the interactive power balance constraint of the shared energy storage, the charge and discharge constraint of the shared energy storage, the charge and discharge constraint of the distribution network energy storage, the maximum charge and discharge power distribution constraint of the energy storage, the state of charge constraint of the shared energy storage, the state of charge constraint of the distribution network energy storage, the energy storage capacity distribution constraint, the electricity sales price constraint from the distribution network to the microgrid, the purchase and sale electricity price constraint of the shared energy storage to the microgrid, the transaction constraint between the distribution network agent and the superior electricity market, and the interest constraint of the distribution network agent.

[0032] In S3, the lower-level interest subject is the multi-microgrid system, considering the cost of the microgrid using the shared energy storage service , the electricity purchase cost of the microgrid from the distribution network , the compensation cost paid by the microgrid's interruptible load to users , the cost of abandoned wind and light . The objective function of the lower-level model is: (39); The constraint conditions of the lower-layer model include microgrid power balance constraint, power purchase constraint of the microgrid from the distribution network, power purchase and sale constraint of the microgrid with the shared energy storage, interruptible load constraint within the microgrid, and wind and light curtailment constraint of the microgrid.

[0033] After obtaining the objective function of the transformed upper-layer model, Gurobi is used for solving, and the solving process is as Figure 2 shown.

[0034] The verification process is as follows: Taking a certain actual distribution network as an example, the effectiveness of the proposed method is analyzed and verified. This distribution network contains a total of one energy storage system and three microgrids (denoted as MG1, MG2, and MG3 respectively).

[0035] Each microgrid is directly connected to the upper-layer distribution network, and the microgrids are not connected to each other. Take as , as , is 0.4 yuan / kW. Taking 24 hours a day as the scheduling period and 1 hour as a scheduling time slot.

[0036] Take and set five case scenarios for comparative analysis. The five case scenarios are set as follows: Scenario 1, the trading price adopts the time-of-use electricity price mechanism.

[0037] Scenario 2, the trading price adopts the dynamic pricing mechanism.

[0038] Scenario 3, the distribution network sets up energy storage, but does not divide the energy storage trading with the microgrid, and the trading price adopts the dynamic pricing mechanism.

[0039] Scenario 4, the distribution network divides the energy storage trading with the microgrid, and the trading price adopts the dynamic pricing mechanism.

[0040] Scenario 5, the distribution network divides the energy storage trading with the microgrid, the microgrid sets up an interruptible load compensation mechanism, and the trading price adopts the dynamic pricing mechanism.

[0041] The comparison of the system operation results under 5 scenarios is shown in Table 1. , , , are the distribution network revenue, the operation cost of the multi-microgrid system, the new energy consumption rate, the ratio of the cost saved by the interruptible load to the loss revenue of the distribution network respectively.

[0042] Table 1 Comparison of System Operation Results under 5 Scenarios

[0043] Comparing Scenario 2 with Scenario 1, the total revenue of power distribution increases by 102.78%, the comprehensive operating cost of the multi-microgrid system increases by 1.52%, and the new energy consumption rate increases by 6.48%. The dynamic pricing strategy formulated by the power distribution network can significantly improve the revenue of the power distribution network and the new energy consumption rate of the microgrid, and the increase in the operating cost of the multi-microgrid system is much smaller than the increase in the profit of the power distribution network. Therefore, the dynamic pricing strategy has certain advantages in the power distribution network and microgrid trading models.

[0044] Compared with Scenarios 1 and 2, in Scenario 3, due to the existence of energy storage, the power distribution network purchases a large amount of electricity during low-price periods in advance and discharges it to the microgrid and the power market during high-price periods, improving the profit of the power distribution network; in Scenarios 1 and 2, the power purchase price of the power distribution network from the microgrid is higher than its real-time selling price to the power market in some periods, resulting in the inability to consume the excess new energy of the microgrid, thus generating wind and light curtailment. In Scenario 3, the power distribution network can store the purchased new energy during this period and sell it in the high-price areas of the real-time power market and the microgrid, avoiding the loss of benefits caused by the power distribution network consuming new energy, increasing the new energy consumption rate of the microgrid, and reducing the operating cost of the microgrid.

[0045] As Figure 3 and Figure 4 shown, limited by the energy storage capacity, some of the surplus new energy output cannot be fully stored in the energy storage, and there are still some periods when the power purchase price of the power distribution network is higher than the real-time selling price to the power market, such as 11:00 - 13:00 and 16:00. At this time, to help the microgrid save costs, the power distribution network still sells to the power market in real time to reduce the cost of the microgrid. The power distribution network loses some revenue, and as decreases, the revenue loss of the power distribution network gradually increases. The total revenue of the power distribution network in Scenario 4 is 7.55% higher than that in Scenario 3, the operating cost of the multi-microgrid is 3.52% higher than that in Scenario 3, but compared with the costs caused by power curtailment in Scenarios 1 and 2, its costs have decreased by 5.96% and 7.37%, and the new energy consumption rate is still 100%.

[0046] Compared with Scenario 4, in Scenario 5 (the method of this embodiment), under the interruptible load mechanism, the revenue of the power distribution network decreases by 312.5 yuan, the operating cost of the multi-microgrid decreases by 638.2 yuan, and the cost reduction of the multi-microgrid system is 2.04 times the profit given up by the power distribution network. Compared with Scenario 3, the revenue of the power distribution network increases by 5.47% and the operating cost increases by 1.56%. The interruptible load compensation mechanism can reduce the cost of the lower-layer multi-microgrid system by multiple times the profit given up by the power distribution network and further increase the flexibility of microgrid users' electricity consumption.

[0047] The system trading and energy storage situation are as Figures 5 - 8 shown. By Figures 5 - 7It can be seen that, based on the principle that the periods with more electricity sales under the condition of conventional energy storage (Scenario 3) in the distribution network are set at the price ceiling or close to the price ceiling, and the periods with less electricity sales are set at the price floor or close to the price floor, the shared energy storage re - prices the new - energy trading price.

[0048] Compared with the electricity - selling price of conventional energy storage (Scenario 3), when the electricity sales volume of shared energy storage to the micro - grid is small, it reduces the electricity - selling price, thereby increasing the electricity - selling price during the periods that are not at the price ceiling and have the highest trading volume, and reducing the electricity - purchasing price at the moment that is not at the electricity - purchasing floor and has the largest electricity - purchasing volume to protect its own interests. For example, during the transaction with MG1, the shared energy storage reduces the electricity - selling price at 20:00 and 22:00, increases the electricity - selling price at 01:00, 09:00, and 17:00, and reduces the electricity - purchasing price at 13:00. In addition, the distribution network does not completely reduce the electricity - purchasing price when the electricity - purchasing volume is large. When the number of electricity - purchasing times is large, in order to protect the interests of the micro - grid, it will allocate part of the price reduction when the electricity - purchasing volume is small, such as at 05:00 in MG3.

[0049] The allocation results of the energy - storage capacity and the maximum charge - and - discharge power are shown in Table 2. Combining Figure 8 It can be seen that, on the one hand, the distribution network purchases a large amount of electricity at the lowest period of the day - ahead electricity - selling price, from 03:00 to 04:00. At the same time, it avoids the discharge period of the shared energy storage and stores the electricity in the distribution - network energy storage, and discharges the electricity during the period with a higher electricity - selling price, from 17:00 to 19:00, to sell electricity to the micro - grid to obtain benefits. On the other hand, on the premise of avoiding damage to its own interests, the distribution network uses the shared energy storage to directly trade with the micro - grid to help the micro - grid absorb the new - energy output. Under the protection of the number of transactions and the average price, it realizes the balance of the interests of itself and the micro - grid. From 02:00 to 04:00, 10:00, and 14:00 - 15:00, the shared energy storage purchases electricity and sells it to the power - short micro - grid and the power market, stores the remaining electricity, and discharges the electricity during the periods with a higher electricity - selling price, from 08:00 to 09:00, 13:00, and 16:00 - 18:00. Thus, the distribution network realizes the transfer of new energy in time and space through the shared energy storage, realizes the full absorption of new energy, realizes the secondary pricing of the purchase and sale of new energy by the micro - grid through the allocation of energy - storage capacity, avoids the secondary pricing of all electric energy, protects the interests of the micro - grid, and at the same time alleviates the loss of interests caused by the distribution network's absorption of the micro - grid's new energy due to the price changes in the external power market.

[0050] Table 2 Allocation of Energy - Storage Capacity and Maximum Charge - and - Discharge Power

[0051] Figure 9 It represents the clearing strategy for the distribution network to trade with the power market. Figure 10 It represents the distribution of the distribution - network revenue and the energy - storage amount. Since this embodiment is based on As the benchmark purchase price in the real-time electricity market, the day-ahead contract volume under this coefficient is the day-ahead benchmark contract volume, and the day-ahead contract deviation rate is 0. As decreases, the way for the distribution network to absorb the new energy of the microgrid gradually changes from selling to the electricity market to selling to the microgrid secondary. The original electricity sold to the microgrid comes from the day-ahead contract volume in the electricity market. Therefore, the distribution network reduces the day-ahead contract volume, that is, the day-ahead contract deviation rate gradually decreases from 0, and the allocated capacity of the shared energy storage gradually increases while the allocated capacity of the distribution network energy storage gradually decreases until , at this time, the real-time electricity purchase price in the electricity market is too low, and the surplus new energy of the microgrid stored in the shared energy storage no longer sells electricity to the electricity market but all sells electricity to the microgrid secondary, corresponding to the unchanged allocated capacity of the two energy storages; when rises, at this time, there is no loss of interest for the distribution network to sell electricity to the electricity market. Therefore, the day-ahead contract deviation rate starts to rise from 0, and the electricity sales volume of the shared energy storage and the distribution network to the electricity market increases. To avoid excessive profit concession and loss of the distribution network's own interests, the allocated capacity of the shared energy storage increases slightly so that most of the electricity is sold during the high-price period in the electricity market.

[0052] Under the conventional energy storage (Scenario 3), the revenue of the distribution network is greatly affected by the real-time pricing in the electricity market. As decreases, the revenue of the distribution network gradually decreases, and the peak-valley difference is large; in Scenario 5, the distribution network allocates the capacity of the distribution network energy storage and the shared energy storage under the perception condition. After the allocation, the revenue of the distribution network is less affected by the price fluctuation in the electricity market, the peak-valley difference of the distribution network revenue decreases, the fluctuation is more stable, and as continues to decrease, the revenue difference between the two gradually increases, which means that the distribution network has stronger risk resistance ability after allocating the energy storage capacity.

[0053] To reduce the operating cost of the microgrid, the microgrid formulates an interruption load strategy under the price decision made by the distribution network, and provides compensation to the microgrid users at an interruption load pricing lower than the selling price of the distribution network and the selling price of the shared energy storage, reducing the operating cost of the microgrid.

[0054] To minimize its own operating cost, each microgrid sets the interruption pricing at the lower limit during the periods with more interruption loads, and at the same time, to improve the enthusiasm of microgrid users to participate in the electricity consumption decision-making and increase the user compensation income, the microgrid sets the interruption pricing at the upper limit during the periods with fewer interruption loads. For example, MG1 sets the pricing at the lower limit at 01:00, 08:00 - 12:00, 14:00, 18:00 - 21:00, 23:00, and sets the pricing at the upper limit at 02:00 - 07:00, 13:00, 17:00, 22:00, 24:00.

[0055] When the electricity purchase price of shared energy storage is lower than the interruption price, the microgrid cannot obtain benefits. Considering its own interests, the microgrid encourages users to consume electricity, and the interrupted load is 0, such as 16:00 in MG2, 01:00 - 03:00, 19:00, and 23:00 in MG3.

[0056] For the interruption strategies of different under each microgrid, when At this time, the real-time electricity purchase price in the power market is relatively low. To avoid losses caused by helping the microgrid absorb excess new energy, the distribution network raises the selling price and reduces the purchase price of this part of new energy. Similarly, the microgrid makes a feedback on the above behavior of the distribution network, reduces the average value of the interruption load compensation pricing to reduce the price increment loss generated by the surplus new energy transaction, and the proportion of each microgrid interruption pricing placed at the pricing lower limit increases, reducing the microgrid's interest loss. In addition, due to the decrease in the average electricity purchase of shared energy storage and the increase in the proportion of the period when the interruption load pricing is higher than the microgrid selling price, the microgrid cannot make a profit, so the interruption load ratio decreases.

[0057] Same compared with When the real-time electricity purchase price in the power market is high, the distribution network can also make a profit by directly selling electricity to the power market in real time. Therefore, the shared energy storage reduces the selling price and increases the purchase price of this part of new energy to give up some benefits and reduce the microgrid cost. Under this strategy, the microgrid raises the interruption pricing to improve the interests of microgrid users. The proportion of each microgrid interruption pricing placed at the pricing upper limit increases, improving the microgrid user income on the premise of ensuring its own interests. In addition, due to the increase in the average electricity purchase of shared energy storage and the decrease in the proportion of the period when the interruption load pricing is higher than the microgrid selling price, the interruption load ratio increases.

[0058] In addition, under the two circumstances, when the interruption pricing is equal to the microgrid selling price, the distribution network and the microgrid make different feedbacks on this behavior. When At this time, the two pricing are equal at 11:00 - 14:00 in MG1, 13:00 - 16:00 in MG2, and at 10:00 in MG3. The microgrid does not choose to interrupt the load at this time; When When it is relatively large, the selling price of shared energy storage is lower than that of the distribution network. The distribution network can also make a profit by directly selling electricity to the power market, and the interests of the microgrid will not be damaged. Therefore, the microgrid interrupts the load to increase the revenue of the distribution network and reduce its own operating costs.

[0059] As becomes smaller and smaller, the purchase price of shared energy storage becomes lower and lower. That is, within the entire trading period, the proportion of the purchase price of shared energy storage being less than the interruption pricing is getting higher and higher, and the selling price of shared energy storage is also getting higher and higher. For its own interests, the proportion of the interrupted load of the microgrid shows a downward trend overall. At the same time, in order to cope with the increase in the operating cost of the microgrid caused by the re - pricing of shared energy storage, the microgrid reduces the interruption pricing to protect its own interests. When is less than 0.8, the interrupted cost after coupling is always less than the interrupted cost without coupling . When is greater than 0.8, the distribution network can also make a profit by directly selling electricity to the power market. At this time, the distribution network will reduce the selling price of shared energy storage and increase the purchase price of shared energy storage. The transaction cost between the microgrid and shared energy storage decreases. Therefore, the microgrid will give up some interests to microgrid users. At this time, the interrupted cost after coupling is greater than the interrupted cost without coupling .

[0060] From the above behaviors, it can be seen that the microgrid is no longer a simple passive leader accepting pricing, but can make strategies beneficial to itself according to external market changes and the response of the distribution network to market changes, and respond more flexibly to external market risks.

[0061] When fluctuates, within the entire fluctuation range, the dynamic pricing strategy in Scenario 2 effectively increases the revenue of the distribution network and improves the new - energy consumption rate of the microgrid compared with the day - ahead time - of - use pricing strategy in Scenario 1, but at the same time, it will increase the operating cost of the microgrid; for the distribution network's capacity allocation of distribution network energy storage and shared energy storage under the dynamic pricing strategy in Scenario 4 compared with Scenario 3, as decreases, the improvement of the distribution network revenue is more obvious; the operating cost of the microgrid in Scenario 5 further decreases compared with Scenario 4, is higher than Scenario 3, but is much less than Scenario 2 and Scenario 1. This is because in Scenario 3, the distribution network helps the microgrid avoid the cost of curtailment of wind and light at the expense of its own interests, which does not conform to the reality; compared with Scenario 4 in Scenario 5, the ratio of the reduced operating cost of the microgrid to the reduced interests of the distribution network is always greater than 1.95 times. When , the distribution network not only increases the revenue but also reduces the operating cost of the microgrid.

Claims

1. A distribution network-microgrid electric energy interactive trading method considering power market risks, characterized by The following steps are involved: S1. Construct a distribution network-microgrid collaborative architecture, including an upper layer based on the distribution network and a lower layer based on multiple microgrids. The upper layer architecture includes distribution network agents, distribution network control centers, and energy storage systems composed of distribution network energy storage and shared energy storage. Each microgrid in the lower layer includes wind power, users, photovoltaics, and microgrid control centers. Each microgrid constitutes a multi-microgrid system. S2. Taking the maximum daily profit of distribution network agents as the goal, construct the objective function and constraints of the upper model; S3, taking the minimization of the daily comprehensive operation cost of the multi-microgrid system as the goal, construct the objective function and constraint conditions of the lower model; S4. Convert the lower model into additional constraints of the upper model, obtain the converted upper model objective function and solve it, and obtain the optimized distribution network-microgrid collaborative interactive power trading scheme, in which: In the day-ahead stage, the microgrid reports its maximum interruptible load to the microgrid control center, and the microgrid control center reports its own renewable energy output and load information to the distribution network control center. The distribution network formulates a day-ahead contractual power purchase strategy based on the information reported by the microgrid. In the real-time stage, the distribution network formulates the distribution network pricing information for real-time purchase and sale of electricity to the microgrid based on the real-time electricity market price information, as well as the power trading strategy and the distribution network energy storage capacity division strategy, and transmits the electricity market price information and distribution network pricing information to the lower-level microgrid. The microgrid adjusts its own interruption load pricing strategy and power consumption strategy based on the information transmitted by the distribution network, and provides real-time feedback to the distribution network; Based on the feedback from the microgrid and the strategy for allocating energy storage capacity, the distribution network makes a secondary adjustment to the pricing of the microgrid’s new energy output for purchasing and selling electricity.

2. The distribution network-microgrid electric energy interactive trading method considering the power market risk according to claim 1 is characterized in that: In S2, the upper-level interest entity is the distribution network agent, considering the distribution network shared energy storage operation benefits , Real-time electricity sales revenue from distribution network to microgrid , distribution network agents’ real-time electricity sales revenue to the power market , the real-time cost of electricity purchased by distribution network agents from the power market , the cost of electricity purchased by distribution network agents from the power market on the day before , the distribution network agent and the power market transaction the next day reporting deviation costs , the objective function of the upper model is: （1）； The constraints of the upper-level model include the interactive power balance constraints between distribution network and microgrid, the interactive power balance constraints of shared energy storage, the shared energy storage charging and discharging constraints, the distribution network energy storage charging and discharging constraints, the energy storage maximum charging and discharging power allocation constraints, the shared energy storage charge state constraints, the distribution network energy storage charge state constraints, the energy storage energy allocation constraints, the distribution network to microgrid power sales price constraints, the shared energy storage to microgrid power sales price constraints, the distribution network agent and the upper-level power market transaction constraints, the distribution network agent interest constraints.

3. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 2 is characterized in that: The calculation method is: （2）； Where t is the index of the scheduling period, T is the total number of scheduling periods; i is the index of the microgrid, and N is the total number of microgrids; , are the electricity sales and purchase prices of shared energy storage to microgrid i during dispatch period t, respectively; , They are the power sold and purchased by the shared energy storage to microgrid i during the dispatch period t; is the time interval; The calculation method is: （3）； In the formula, The price of electricity sold by the distribution network to microgrid i during the dispatching period t; The power sold by the distribution network to microgrid i during the dispatching period t; , The calculation method is: （4）； （5）； In the formula, , They are the real-time electricity selling and purchasing prices of the distribution network agents to the power market during the t dispatching period; , They are the real-time power sold and purchased by the distribution network to the power market during the t dispatch period; The power sold by the shared energy storage of the distribution network to the power market in real time during the t dispatch period; The calculation method is: （6）； In the formula, is the day-ahead contract price during the dispatch period t; is the benchmark contract power of the day-ahead dispatch period t; , They are the positive and negative deviations of the day-ahead contract declaration during the t dispatch period, respectively; The electricity market recovers the reported deviation according to the corresponding settlement income. The calculation method is: （7）； （8）； In the formula, , Report the positive and negative deviation prices respectively for the next day; , Report the positive and negative deviation settlement profit coefficients respectively.

4. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 3 is characterized in that: The constraints of the upper model are as follows: Interactive power balance constraints between distribution network and microgrid: （9）； In the formula, , They are the charging and discharging power of the distribution network energy storage during the dispatching period t respectively; Shared energy storage interactive power balance constraint: During the period when the microgrid purchases and sells electricity to the shared energy storage, the shared energy storage charging and discharging power is determined by the total energy demand after the energy exchange is completed at each microgrid bus. The constraint is: （10）； （11）； （12）； In the formula, , They are respectively the charging and discharging power of shared energy storage during the scheduling period t; , The power purchased by the shared energy storage from microgrid i during the dispatch period t and sold to the power market and the power-deficient microgrid; , The charging power that the shared energy storage uses to sell to the power market and power-deficient microgrids during the t scheduling period; , The discharge power of the shared energy storage sold to the power market and power-deficient microgrids during the t scheduling period; Shared energy storage charging and discharging constraints: （13）； In the formula, , They are the charge and discharge flags of the shared energy storage during the t scheduling period, with values ​​of 0 or 1. A value of 1 indicates that charging or discharging is in progress, and the other flag symbols are similar; The maximum charging and discharging power of shared energy storage; Distribution network energy storage charging and discharging constraints: （14）； In the formula, , are the charging and discharging power of the distribution network energy storage during the dispatching period t respectively; , They are respectively the charge and discharge flags of the distribution network energy storage during the t scheduling period; The maximum charging and discharging power of the distribution network energy storage; Energy storage maximum charging and discharging power allocation constraints: （15）； In the formula, is the total maximum charging and discharging power of energy storage; Shared energy storage state of charge constraints: （16）； （17）； In the formula, , are the charge states of the shared energy storage in the scheduling periods t and t-1 respectively; It is the energy storage charging and discharging efficiency; , are the upper and lower limit coefficients of energy storage charge state respectively; is the maximum storage capacity of shared energy storage; Distribution network energy storage charge state constraints: （18）； （19）； In the formula, , They are the charge states of the distribution network energy storage in the dispatching periods t and t-1 respectively; The maximum storage capacity of the distribution network energy storage; Energy storage allocation constraints: （20）； In the formula, is the maximum value of total energy storage; The price constraints for electricity sold by distribution network to microgrid: （21）； （22）； （23）； In the formula, , They are the upper and lower limits of the price of electricity sold by the distribution network to the microgrid during the dispatching period t; The flag bit of power sales from the distribution network to microgrid i during the dispatching period t; is the maximum number of electricity sales from the distribution network to microgrid i; The price constraints for shared energy storage to sell electricity to micro-grids are as follows: （24）； （25）； In the formula, , They are the upper and lower limits of the electricity price sold by shared energy storage to the microgrid during the dispatch period t; , They are the upper and lower limits of the electricity purchase price from the shared energy storage to the microgrid during the dispatch period t; At the same time, the electricity purchase and sales price of shared energy storage is coupled with the electricity purchase price of the external power market to protect the interests of both the distribution network and the microgrid: （26）； （27）； （28）； （29）； In the formula, , They are the flags of electricity sold and purchased by the shared energy storage to microgrid i during the dispatching period t; It is the real-time electricity purchase coefficient of the power market corresponding to the full consumption of new energy in the microgrid when no energy storage is set; is the real-time electricity purchase coefficient in the current electricity market; , are the price protection coefficients for selling and purchasing electricity respectively; , are the maximum number of times the shared energy storage sells and purchases electricity from microgrid i; Distribution network agents and upper-level power market transaction constraints: （30）； （31）； （32）； （33）； （34）； （35）； （36）； （37）； In the formula, The maximum power purchased by the distribution network in the day-ahead contract with the power market; , They are the power flags for purchase and sale of electricity in the dispatching period t respectively; , They are the real-time maximum purchase and sale power of the distribution network and the power market respectively; represent or ; represent or , respectively, are the maximum positive and negative deviation powers reported by the distribution network on the second day of trading in the electricity market; , They are respectively the positive and negative deviation status flags of the scheduling period t, represent or ; represent or , respectively, positive and negative deviation coefficients; The interests of distribution network agents are constrained. When the real-time power purchase coefficient of the external power market is not equal to the benchmark value, the distribution network will suffer the risk of losing profits when absorbing new energy from the microgrid. In order to ensure that the interests of the distribution network are not lost and to prevent the unreasonable secondary pricing of the distribution network from causing the increase of microgrid operation costs, the constraints are set as follows: （38）； In the formula, , They are the upper and lower limits of incremental revenue after secondary pricing respectively; is the price of electricity purchased by the distribution network from microgrid i during dispatch period t.

5. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 4 is characterized in that: In S3, the lower-level stakeholders are multi-microgrid systems, and the cost of using shared energy storage services by microgrids is considered.

2. The cost of microgrid purchasing electricity from distribution network , compensation paid to users for interrupted loads of microgrids , Cost of curtailing wind and solar power , the objective function of the lower model is: （39）； The constraints of the lower-level model include microgrid power balance constraints, microgrid power purchase constraints from distribution networks, microgrid power purchase and sales constraints with shared energy storage, microgrid internal interruptible load constraints, and microgrid wind and solar power abandonment constraints.

6. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 5 is characterized in that: The calculation method is: （40）； The calculation method is: （41）； The calculation method is: （42）； In the formula, is the unit electricity compensation price paid by microgrid i to users during dispatch period t; is the interrupted load power of microgrid i during dispatch period t; The calculation method is: （43）； In the formula, is the unit cost of wind and solar power abandonment; is the wind and solar power abandoned by microgrid i during scheduling period t.

7. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 6 is characterized in that: The constraints of the lower model are as follows: Microgrid power balance constraints: （44）； In the formula, , , are the wind power, photovoltaic, and load outputs of microgrid i during dispatch period t, respectively; is the Lagrange multiplier of the equality constraint; Constraints on power purchased by microgrid from distribution network: （45）； In the formula, The maximum power purchased by the microgrid from the distribution network; , is the Lagrange multiplier corresponding to the inequality constraint; The flag position of microgrid i selling electricity to the distribution network during the dispatching period t; Power constraints for microgrids to purchase and sell electricity with shared energy storage: （46）； （47）； （48）； （49）； In the formula, , They are the maximum purchase and sale power of the microgrid and shared energy storage transactions respectively; , They are respectively the flags of microgrid i purchasing and selling electricity to shared energy storage during dispatch period t; , , , , , is the Lagrange multiplier corresponding to the inequality constraint; Interruptible load constraints within the microgrid: （50）； （51）； （52）； （53）； In the formula, is the maximum power of the load that can be interrupted by microgrid i during the dispatch period t; , They are the upper and lower limits of the compensation price for microgrid interruption load during dispatch period t respectively; is the interruption load flag of microgrid i in scheduling period t; is the maximum number of load interruptions of microgrid i; pricing protection factors for disruptions; , , , , , is the Lagrange multiplier corresponding to the inequality constraint; Constraints on wind and solar power abandonment in microgrids: （54）； In the formula, The maximum value of wind and solar power abandonment in the microgrid; , is the Lagrange multiplier corresponding to the inequality constraint.

8. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 7 is characterized in that: In the above S4, the lower layer model is converted into additional constraints of the upper layer model to obtain the converted upper layer model objective function, and the process includes: S4.

1. Non-convex transformation of lower model: because , Non-convex, use the exponential method to transform it into a convex function form, let , , , For the corresponding , The logarithmic transformation form of , then the lower model is converted to: （55）； （56）； （57）； （58）； In the formula, , ; Indicates the lower limit; Introducing auxiliary variables , converting the non-convex constraint to: （59）； （60）； （61）； In the formula, , , is the Lagrange multiplier corresponding to the inequality constraint; M is the constant in the big-M method; S4.

2. Standardization of lower-level models: Standardize the inequality constraints and equality constraints of the underlying model: （62）； （63）； In the formula, represents an inequality constraint; represents an equality constraint; , , They represent the power, price, and flag variables in the constraints respectively; Construct the Lagrangian function: （64）； , are the Lagrange multipliers of equality constraints and inequality constraints respectively; x and y are the number of equality constraints and inequality constraints respectively; represents the constructed Lagrangian function; S4.

3. Lower layer model conversion: Using the KKT method, the partial differential of the decision variables of the lower model is calculated based on formula (64), and the lower model is transformed into additional constraints of the upper model: （65）； In the formula, l represents the complete set of decision variables in the equality and inequality constraints; Formula (65) contains nonlinear constraints, which can be transformed into the following linear form: （66）； In the formula, is a Boolean variable; In the upper model, the constraints of equations (13)-(14), (22), (26)-(27) are non-convex. Equations (13)-(14) are transformed into the following form: （67）； （68）； In the formula, is a maximum value; j represents abs or relea, corresponding They are , , Similarly; k represents ch or dis, corresponding They are , , Similarly; For equation (22), equation (26)-(27), let: （69）； In the formula, corresponding to Take , or , express , or , representing , The product of , The product of , The product of express , or , Similarly; Then we have: （70）； At this time, equations (22), (26), and (27) are transformed into the following convex constraints: （71）； （72）； （73）； S4.4, the objective function of the transformed upper model is: （74）。 9. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 8 is characterized in that: In the above S4, the solution process is: Step 1: Definition , They are The distribution network revenue and multi-microgrid operating cost at that time, C is the current distribution network revenue, and F is the current multi-microgrid operating cost; is the step length; , , The initial value is 0; Step 2: Input the load and new energy data of each microgrid, generate the day-ahead contract volume of the distribution network under the real-time power purchase and sale benchmark price of the power market, and calculate the real-time power purchase coefficient of the power market corresponding to the distribution network fully absorbing the new energy of the microgrid when the distribution network does not lose its own income and the energy storage does not participate in the transaction, that is, ; Step 3: Input the current real-time electricity purchase coefficient of the electricity market , and calculated based on conventional energy storage scenarios , ; Step 4: According to and The size relationship, in , On the basis of , Make corrections. The correction process is as follows: Determine the real-time power purchase coefficient of the current power market Is it greater than or equal to : when When the current income C of the distribution network is less than If so, then according to , renew , , until If not, keep the current , , and determine whether the current multi-microgrid operation cost F is greater than If so, then according to renew , until If not, keep the current ; when When judging Is it established? If so, according to , renew , , until If not, judge and Is it true? If so, keep the current , If not, according to , renew , ; Step 5: Based on the revised , , Recalculate , , , and obtain the optimized dispatching result. According to the dispatching result, the distribution network-microgrid collaborative interactive energy trading plan is implemented.

10. The distribution network-microgrid electric energy interactive trading method considering the electric power market risk according to claim 9 is characterized in that: In the day-ahead phase of the distribution network-microgrid collaborative interactive power trading solution, the maximum amount of self-interruptible load is , its own new energy output index , , load information refers to , the day before the contract power purchase strategy refers to ; In the real-time stage, the electricity market price information refers to , , distribution network pricing information refers to , , , power trading strategy refers to , , , , , , , , the distribution network energy storage capacity division strategy refers to , , , , the information transmitted by the distribution network refers to , , , , , the self-interruption load pricing strategy refers to , electricity strategy refers to , , , , Microgrid New Energy Power Purchase and Sales Pricing Index , .

Images