A distributed optimization method for unit commitment P2P energy transaction

Abstract

The application provides a kind of distributed optimization method for unit combination P2P energy transaction, it is related to electrical engineering field, the method comprises: establishing the P2P energy transaction model of producer and consumer based on the maximization of social welfare;The matrix form P2P energy transaction model is reshaped;For the problem of non-convexity caused by integer variable contained in the reshaped P2P energy transaction model, the algorithm of the application is designed, and the mixed integer nonlinear P2P energy transaction problem is solved.The application can effectively solve the maximization of social welfare in mixed integer nonlinear programming (MINLP) in a distributed manner, realize parallel computing, better protect the privacy of both parties in transaction, while the application effectively solves the non-convexity problem caused by integer variable in unit combination, so as to obtain the optimal solution of mixed integer nonlinear programming programming problem with the minimum cost as the target, so as to better ensure the optimality of transaction result.

Description

Technical Field

[0001] This invention relates to the field of electrical engineering, specifically to a distributed algorithm for peer-to-peer (P2P) energy trading of unit combinations, and more particularly to a distributed optimization method for P2P energy trading of unit combinations. Background Technology

[0002] Advances in modern technology have driven the rise of peer-to-peer (P2P) networks. This fully decentralized network architecture allows market participants to easily share their resources. P2P networks allow users with distributed energy resources to trade directly, where each user is both a producer and a consumer, and all users in the network are equal. As an emerging energy supply model, P2P energy trading can guarantee the equal status of both parties and user privacy, increasing the enthusiasm of producers and consumers to participate in the energy market; it can also reduce electricity demand, minimize storage requirements, and improve the resilience of the power system.

[0003] Traditional energy trading problems mostly employ centralized optimization methods, using a central control system for comprehensive calculations and scheduling to directly obtain the optimal trading solution. However, with the rapid development of renewable energy and the increasing integration of distributed generation equipment into the grid, centralized optimization methods are consuming significantly more time and computational resources. These issues make centralized methods ineffective in handling large-scale grid-connected distributed energy resources, forcing energy trading systems to evolve towards greater openness and decentralization to improve system flexibility, reliability, and real-time responsiveness.

[0004] To address these issues, some research has begun to employ distributed methods to optimize trading schemes. Distributed optimization (DO) methods not only enable parallel computation but also maintain the optimality of trading results, making them an effective means of designing new energy trading mechanisms. However, current research primarily focuses on solving linear or convex global optimization problems. When considering unit commitment (UC) problems in the system, the introduced binary variables cause the problem to become non-convex, and traditional DO algorithms typically cannot guarantee convergence to the optimal result. Summary of the Invention

[0005] To address the non-convexity problem caused by integer variables in unit combination, this invention provides an SDM-CADMM distributed method for P2P energy trading in unit combination, constructing a social welfare maximization problem involving both the user and generation sides; designing the SDM-CADMM algorithm to solve the MINLP problem in a distributed manner; during the iteration process, this method transforms the non-convex set into an approximate set within the convex hull, and obtains the optimal solution to the MINLP problem upon algorithm convergence, thereby ensuring the optimality of the trading results. The specific steps of this method are as follows:

[0006] S1: Establish a P2P energy trading model for producers and consumers based on maximizing social welfare;

[0007] S2: Reshaping the matrix-form P2P energy trading model;

[0008] S3: To address the nonconvexity problem caused by integer variables in the reshaped P2P energy trading model, the SDM-CADMM algorithm was designed to solve the mixed-integer nonlinear P2P energy trading problem.

[0009] A storage device that stores instructions and data for implementing an SDM-CADMM distributed method for P2P energy trading of unit combinations.

[0010] An SDM-CADMM distributed device for unit-based P2P energy trading includes: a processor and a storage device; the processor loads and executes instructions and data in the storage device to implement an SDM-CADMM distributed method for unit-based P2P energy trading.

[0011] Compared to existing technologies, the technical solution provided by this invention offers the following advantages: Traditional P2P energy trading does not consider the unit combination problem, while this invention incorporates the uncertainties of unit combination and wind power generation, establishing a day-ahead P2P energy trading model based on maximizing social welfare. Traditional methods for solving P2P energy trading problems typically assume the problem is linear or convex. However, non-convex problems remain unresolved. Therefore, this invention proposes an SDM-CADMM algorithm that does not rely on a central coordinator. This algorithm can effectively solve the social welfare maximization problem in mixed integer nonlinear programming (MINLP) in a distributed manner, achieving parallel computation rather than centralized scheduling for unified management. This better protects the privacy of both trading parties. This invention effectively solves the non-convexity problem caused by integer variables in unit combination, thereby obtaining the optimal solution to the mixed integer nonlinear programming (MINLP) problem with cost minimization as the objective, thus better guaranteeing the optimality of the trading results. Attached Figure Description

[0012] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0013] Figure 1 This is a flowchart of an SDM-CADMM distributed method for P2P energy trading of unit combinations, as described in an embodiment of the present invention.

[0014] Figure 2 This is a schematic diagram of the hardware device working in an embodiment of the present invention. Detailed Implementation

[0015] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0016] Example 1

[0017] Please refer to Figure 1 , Figure 1 This is a flowchart of an SDM-CADMM distributed method for P2P energy trading of unit combinations, as described in an embodiment of the present invention, specifically including:

[0018] S1: Establish a P2P energy trading model for producers and consumers based on maximizing social welfare; the steps are as follows:

[0019] S1.1: Establish a bilateral transaction model between producers and consumers:

[0020] (1)

[0021] (2)

[0022] (3)

[0023] (4)

[0024] (5)

[0025] (6)

[0026] Wherein, formula (1) is the node n At any moment t The net injected power is calculated using formula (2), which is the formula for renewable energy in the scenario. s Under the constraint of active power output range, formula (3) represents the node n At any moment t The net injection power is equal to the sum of its transactions with all neighboring individuals. Formula (4) represents the producer-consumer ratio. n to adjacent producers and consumers m Electricity is sold, at which point the producers and consumers...n For producers and prosumers m For consumers; Formula (5) represents prosumers. n to adjacent producers and consumers m Purchase electricity, at this time, producer and consumer n For consumers, prosumers m For producers; Formula (6) represents prosumers. n and consumers m In energy trading, the total amount of electricity bought and sold between the two parties is 0.

[0027] in, Representing each producer and consumer n At any moment t Net injection power, , Represents an individual n For producers, Represents an individual n For consumers, Indicates in the scene s Downwind farm n At any moment t The predicted output value, Indicates in the scene s Downwind farm n At any moment t The amount of wind curtailment, Indicates in the scene s Lower thermal power unit n At any moment t of efforts, Indicates in the scene s Next user n electrical load, collection Represents all producers and consumers, collectively. Represents all producers, set Represents all consumers, set It is a set of transaction times, a set It is a scene set, a collection Prosumer n The set of neighbors.

[0028] S1.2: Taking generator sets as producers, establish their safe operation constraints in the energy trading process, i.e., the generator set combination problem; the safe operation constraints are:

[0029] (7)

[0030] (8)

[0031] (9)

[0032] (10)

[0033] (11)

[0034] (12)

[0035] (13)

[0036] Among them, formula (7) is the thermal power unit n At any moment t The upper and lower limits of output are constrained by the time, and formula (8) represents the scenario. s Lower thermal power unit n The upper limit constraint for the ramp is given by formula (9) in the scenario. s Lower thermal power unit n The lower limit constraint for ramping, formula (10) is the thermal power unit n exist t At any given time, the mutual exclusion constraint between start-up and shutdown is related to the thermal power unit in formula (11). n At any moment t and t The operating status and operation of -1, formula (12) is the thermal power unit n exist t and t The upper limit constraint on the continuous start-stop duration at time -1 is given by formula (13) for thermal power units. n exist t and t -1 is the lower limit constraint on the duration of continuous start-stop at time -1.

[0037] in, They represent thermal power units n The upper limit of output, Indicates thermal power unit n The lower limit of output, Indicates thermal power unit n exist t The minimum effort required at any given moment. Indicates thermal power unit n exist t The maximum output at any given moment; Indicates in the scene s Lower thermal power unit n exist t Constant effort Indicates in the scene s Lower thermal power unit n exist t The output at time -1; Indicates thermal power unit n exist t The start / stop status at any given time. Indicates thermal power unit n exist t It is always powered off. Indicates thermal power unit n exist t The device must be powered on at all times. They represent thermal power units n exist t Start-up / shutdown operations at any time Indicates thermal power unit n exist t Start-up operations are performed at all times. Indicates thermal power unit n exist t No startup operation was performed at any time. Indicates thermal power unit n exist t The machine must be stopped at any time. Indicates thermal power unit n exist t No shutdown operation was performed at any time. They represent thermal power units n Maximum climbing speed They represent thermal power units n Lower limit of climbing speed Indicates thermal power unit n exist t Start-stop status at time -1 Indicates thermal power unit n exist The start / stop status at any given time. T This is the last time period. This indicates the minimum duration for continuous startup of a thermal power unit. These represent the minimum duration of continuous shutdown of thermal power units.

[0038] S1.3: Taking electricity users as consumers, establish constraints on their electricity load range during the energy trading process; the electricity load range constraints are as follows:

[0039] (14)

[0040] (15)

[0041] Among them, formula (14) is the user n At any moment t The upper and lower limits of the power load are constrained, and formula (15) represents the user's... n Total energy demand at all times.

[0042] in, Indicates user n At any momentt The electrical load, Representing users respectively n At any moment t The upper limit of the power load below, Representing users respectively n At any moment t The lower limit of the power load, User n The sum of electrical loads at all times.

[0043] S1.4: Construct a P2P energy trading model that incorporates the problem of maximizing social welfare on both the user and power generation sides, with cost minimization as the objective function. The P2P energy trading model is as follows:

[0044] (16)

[0045] (17)

[0046] (18)

[0047] (19)

[0048] (20)

[0049] In this context, formula (16) is the objective function for minimizing costs, formula (17) includes various constraints from both the user and power generation sides, and formula (18) represents the expected social welfare of all prosumers. Representing the minimum expected social welfare of all producers and consumers, formula (19) is the producer's... n The cost-utility function, i.e. Indicates producer n Cost-utility function, formula (20) for consumers n The utility function, i.e. Consumers n Utility function; , , It is a positive number. , , It is a positive constant of the cost-utility function. , , It is a positive constant of the utility function. It is a thermal power unit n The startup cost coefficient.

[0050] The P2P energy trading model for minimizing costs can be written in matrix form as follows:

[0051] (twenty one)

[0052] (twenty two)

[0053] (twenty three)

[0054] (twenty four)

[0055] Wherein, formula (21) is the matrix form of the cost minimization objective function, and formula (16) is... Specific manifestations; yes The feasible set, It is the first n The feasible solutions of each producer and consumer constitute a nonconvex set formed by the mixed integer nonlinear constraints (17). . Represents the set of all decision variables. Indicates thermal power unit n The start and stop state vector, Indicates thermal power unit n The startup operation vector, Indicates thermal power unit n The shutdown operation vector, Indicates in the scene s Lower thermal power unit n The output vector, Represents each producer and consumer n The net injected power vector; Indicates the first n The constant matrix of each producer and consumer, Z This indicates the coupling relationship between producers and consumers. This indicates the introduced auxiliary variable. , Indicates the first n Auxiliary variables introduced by individual producers and consumers.

[0056] S2: Reshape the matrix-form P2P energy trading model. The specific steps are as follows.

[0057] Step 2.1: Transform the non-convex set into a convex set. The specific model is as follows:

[0058] (25)

[0059] (26)

[0060] (27)

[0061] (28)

[0062] Where, conv( X n ) represents a nonconvex set X n The convex hull is fixed at a point.

[0063] S2.2: The Lagrange dual function of the primal problem

[0064] (29)

[0065] (30)

[0066] (31)

[0067] (32)

[0068] Wherein, formula (29) is the convex relaxation dual function of the original problem, i.e. Describe the convex relaxation dual function of the primal problem. It is the set of dual multiplier vectors; Indicates the first n Equation constraints corresponding to individual producers and consumers The transpose of the dual multiplier vector. Indicates producer-consumer n The constant matrix, Represents the set of all decision variables. Indicates producer-consumer n Auxiliary variables, This represents the transpose of the auxiliary variable set. n Indicates the first n Individual consumers, Let represent the set of dual multiplier vectors. Indicated Transpose.

[0069] S2.3: Augmented Lagrangian function of the original problem

[0070] (33)

[0071] (34)

[0072] Formula (33) is the augmented Lagrangian function form of the original problem. Represent the augmented Lagrangian function of the original problem. ρ Indicates the penalty parameter. Indicates producer-consumer n With neighboring producers and consumers m Trading volume between Indicates producer-consumerm With neighboring producers and consumers n Trading volume between Indicates producer-consumer n and consumers m The average transaction volume between them.

[0073] S2.4: Approximation of the augmented Lagrangian function of the original problem

[0074] (35)

[0075] (36)

[0076] (37)

[0077] (38)

[0078] Among them, formula (35) is the mirror image of all augmented Lagrange functions, i.e. Describes the mirror function of all augmented Lagrange functions. Indicates the first k Next iteration of the consumer n The mirror function of the augmented Lagrange function, formula (36) is the objective function. f n In the k In the nth iteration n Feasible solutions for individual consumers The first-order Taylor expansion at is given by formula (37), which is the tangent plane approximation function of all augmented Lagrangian functions, i.e. Denotes the tangent plane approximation function of all augmented Lagrange functions. Indicates the first k Next iteration of the consumer n The tangent plane approximation function of the augmented Lagrange function, formula (38) is based on The approximate function of the tangent plane centered at the center is, i.e. Let represent the approximate function of the tangent plane, and let this approximate function satisfy . , Describe the objective function f n At point The first derivative at that point, Indicates the first k Feasible solutions in the next iteration. Indicates the first k Next iteration of the consumer n Auxiliary variables introduced.

[0079] S3: To address the nonconvexity issue caused by integer variables in the restructured P2P energy trading model, an SDM-CADMM algorithm independent of a central coordinator is designed to solve the mixed-integer nonlinear P2P energy trading problem. The SDM-CADMM algorithm is as follows:

[0080] S3.1: Update the first n Feasible solutions for individual consumers

[0081] (39)

[0082] in, Represents the set of all decision variables. Indicates the first k -1 iteration produces and consumes n The introduced auxiliary variables, Indicates the first n The transpose of the dual multiplier vector of each producer and consumer. It is the feasible region of the convex hull (conv( X n ) n The inner approximate set of each producer-consumer is composed of finite convex sets. composition, Indicates the number of data within the convex set. . α i The weights representing convex sets, It is a convex set The first in i One vertex, It represents the set of all decision variables.

[0083] S3.2: Update the first n Auxiliary variables of individual prosumers

[0084] (40)

[0085] in, This indicates that the nth prosumer is in the nth... k Auxiliary variables in the next iteration.

[0086] S3.3: Update the first n The feasible domain of the convex hull of an individual consumer

[0087] (41)

[0088] (42) (43)

[0089] Wherein, formula (41) represents minimizing At point gradient at, It is conv( X n A vertex of ). Indicates the first k The feasible region of the convex hull at the next iteration is conv( X n The inner approximate set of the nth producer-consumer. Indicates the first k The feasible region of the convex hull at -1 iterations is conv( X n ) n The inner approximate set of individual producers and consumers Representation function At point gradient at, It is the convex hull (conv( X n A vertex of ).

[0090] S3.4: Perform convergence testing through steps S3.1-S3.4.

[0091] (44)

[0092] in, The threshold for determining whether the SDM-CADMM algorithm has terminated. ,when When the algorithm converges. Indicates the first k During the nth iteration n The difference between the approximate function value of the tangent plane of an augmented Lagrange function and its mirror image value.

[0093] S3.5: Update the first n The dual multipliers of individual producers and mirror function

[0094] (45)

[0095] (46)

[0096] when Update and

[0097] (47)

[0098] (48)

[0099] in, This indicates a given strict step size condition; Indicates the first k The iteration and the k The difference between the mirror function values ​​of the augmented Lagrange function at the -1st iteration and the 1st iteration k The approximate function value of the tangent plane of the augmented Lagrange function in the nth iteration is similar to that in the nth iteration. k The ratio of the differences between the mirror image values ​​of the augmented Lagrange function at the -1st iteration; This represents the updated dual multiplier vector.

[0100] otherwise, and remain unchanged.

[0101] (49)

[0102] (50)

[0103] in, Indicates the first k In the next iteration, the consumer... n With neighboring producers and consumers m Trading volume between Indicates the first k In the next iteration, the consumer... m With neighboring producers and consumers n Trading volume between them; Indicates the first k The mirror function of the augmented Lagrange function at +1 iterations Indicates the first k At the +1st iteration, the... n The dual multiplier vector of each producer-consumer.

[0104] Example 2

[0105] An SDM-CADMM distributed device 201 for P2P energy trading of unit combinations, such as Figure 2 As shown, it includes: a processor 202 and a storage device 203; the processor 202 loads and executes the instructions and data in the storage device 203 to implement an SDM-CADMM distributed method for P2P energy trading of unit combinations.

[0106] Example 3

[0107] A storage device that stores instructions and data for implementing an SDM-CADMM distributed method for P2P energy trading of unit combinations.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A distributed optimization method for P2P energy trading of unit combinations, characterized in that: include: S1: Establish a P2P energy trading model for producers and consumers based on maximizing social welfare; S2: Reshaping the matrix-form P2P energy trading model; S3: To address the nonconvexity problem caused by integer variables in the reshaped P2P energy trading model, design an algorithm to solve the mixed-integer nonlinear P2P energy trading problem. The designed algorithm is as follows: S3.1: Update the first n Decision variables of individual consumers : in, This represents the augmented Lagrange function. Represents the set of all decision variables. Indicates the first k -1 iteration produces and consumes n The introduced auxiliary variables, Indicates the first k During the nth iteration n The dual multiplier vector of each producer-consumer, It is the feasible region of the convex hull (conv( X n ) n The inner approximate set of each producer-consumer is composed of finite convex sets. composition, Describing a convex set The amount of data within, ; α i The weights representing convex sets, It is the first in the convex set i One vertex; S3.2: Update the first n Auxiliary variables of individual prosumers : in, Indicates the first n Individual consumer in the first k Auxiliary variables in the next iteration Indicates the first n The constant matrix of each producer and consumer, Indicates the first n Individual consumer in the first k Decision variables in the next iteration Indicates the first n Auxiliary variables introduced by individual producers and consumers; S3.3: Update the first n The feasible domain of the convex hull of individual consumers : in, Represents the augmented Lagrange function At point gradient at, Indicates the first k Next iteration of the consumer n The introduced auxiliary variables, It is the feasible region of the convex hull (conv( X n A vertex of ) Describe the objective function f n At point The first derivative at that point, Indicates the first k Next iteration of the consumer n The set of dual multiplier vectors, ρ Indicates the penalty parameter. Indicates the first k Feasible solutions in the next iteration. Indicates the first k Next iteration of the consumer n With neighboring producers and consumers m Trading volume between Indicates the first k Next iteration of the consumer m With neighboring producers and consumers n Trading volume between Indicates the first k The feasible region of the convex hull at the next iteration is conv( X n ) n The inner approximate set of individual producers and consumers Indicates the first k The feasible region of the convex hull at -1 iterations is conv( X n ) n The inner approximate set of each producer-consumer; S3.4: Perform convergence checks through steps S3.1-S3.4: in, Indicates the first k During the nth iteration n The difference between the approximate function value of the augmented Lagrange function on the tangent plane and the mirror function value, when When the algorithm converges, The threshold for determining whether the designed algorithm has terminated, and ; S3.5: Update the first n The dual multipliers of individual producers and mirror function : in, This represents the updated dual multiplier vector; Indicates the first k The iteration and the k The difference between the mirror function values ​​of the augmented Lagrange function at the -1st iteration and the 1st iteration k The approximate function value of the tangent plane of the augmented Lagrange function in the nth iteration is similar to that in the nth iteration. k The ratio of the differences between the mirror image values ​​of the augmented Lagrange function at the -1st iteration; Indicates the first n Approximate function of the tangent plane of a producer-consumer, Indicates the first k During the nth iteration n The mirror function of the augmented Lagrange function of individual producers and consumers. Indicates the first n The mirror function of all augmented Lagrange functions of a single producer-consumer; when Update and : in, This indicates a given strict step size condition; otherwise, and Remain unchanged, that is: in, Indicates the first k The mirror function of the augmented Lagrange function at +1 iterations Indicates the first k At the +1st iteration, the... n The dual multiplier vector of each producer-consumer.

2. The distributed optimization method for P2P energy trading of unit combinations as described in claim 1, characterized in that: The specific steps of step S1 are as follows: S1.1: Establish a bilateral transaction model between producers and consumers; S1.2: Taking generator sets as producers, establish their safe operation constraints in the energy trading process, i.e., the generator set combination problem; S1.3: Establish constraints on the range of electricity load for electricity users in the energy trading process, taking electricity users as consumers; S1.4: Construct a P2P energy trading model with the objective functions of minimizing costs and maximizing social welfare.

3. The distributed optimization method for P2P energy trading of unit combinations as described in claim 2, characterized in that: In step S1.1, the bilateral transaction model is as follows: (1) (2) (3) (4) (5) (6) Wherein, formula (1) is the node n At any moment t The net injected power is calculated; Formula (2) is the renewable energy in the scenario. s The active power output range constraint is given by formula (3), which represents the node. n At any moment t The net injection power is equal to the sum of its transactions with all neighboring individuals; Formula (4) represents the producer-consumer relationship. n to adjacent producers and consumers m Electricity is sold, at which point the producers and consumers... n For producers and prosumers m For consumers; Formula (5) represents prosumers. n to adjacent producers and consumers m When purchasing electricity, the producer and consumer... n For consumers, prosumers m For producers; Formula (6) represents prosumers. n and consumers m In energy trading, the total amount of electricity bought and sold between the two parties is 0. in, Representing each producer and consumer n At any moment t Net injection power, , Represents an individual n For producers, Represents an individual n For consumers, Indicates in the scene s Downwind farm n At any moment t The predicted output value, Indicates in the scene s Downwind farm n At any moment t The amount of wind curtailment, Indicates in the scene s Lower thermal power unit n At any moment t of efforts, Indicates in the scene s Next user n electrical load, collection Represents all producers and consumers, collectively. Represents all producers, set Represents all consumers, set It is a set of transaction times, a set It is a scene set, a collection Prosumer n The set of neighbors.

4. The distributed optimization method for P2P energy trading of unit combinations as described in claim 3, characterized in that: In step S1.2, the safe operation constraint is: (7) (8) (9) (10) (11) (12) (13) Formula (7) is the thermal power unit n At any moment t The upper and lower limits of output are constrained by the time, and formula (8) represents the scenario. s Lower thermal power unit n The upper limit constraint for the ramp is given by formula (9) in the scenario. s Lower thermal power unit n The lower limit constraint for ramping, formula (10) is the thermal power unit n exist t At any given time, the mutual exclusion constraint between start-up and shutdown is related to the thermal power unit in formula (11). n At any moment t and t The operating status and operation of -1, formula (12) is the thermal power unit n exist t and t The upper limit constraint on the continuous start-stop duration at time -1 is given by formula (13) for thermal power units. n exist t and t -1 is the lower limit constraint on the duration of continuous start-stop at time -1; in, Indicates thermal power unit n The upper limit of output, Indicates thermal power unit n exist t The minimum effort required at any given moment. Indicates thermal power unit n exist t The maximum output at any given moment; Indicates in the scene s Lower thermal power unit n exist t Constant effort Indicates in the scene s Lower thermal power unit n exist t The output at time -1; Indicates thermal power unit n exist t The start / stop status at any given time. Indicates thermal power unit n exist t It is always powered off. Indicates thermal power unit n exist t The device must be powered on at all times. They represent thermal power units n exist t Start-up / shutdown operations at any time Indicates thermal power unit n exist t Start-up operations are performed at all times. Indicates thermal power unit n exist t No startup operation was performed at any time. Indicates thermal power unit n exist t The machine must be stopped at any time. Indicates thermal power unit n exist t No shutdown operation was performed at any time. Indicates thermal power unit n Maximum climbing speed Indicates thermal power unit n Lower limit of climbing speed Indicates thermal power unit n exist t Start-stop status at time -1 Indicates thermal power unit n exist The start / stop status at any given time. T This is the last time period. This indicates the minimum duration for continuous startup of a thermal power unit. These represent the minimum duration of continuous shutdown of thermal power units.

5. A distributed optimization method for P2P energy trading of unit combinations as described in claim 4, characterized in that: In step S1.3, the constraint on the electrical load range is as follows: (14) (15) Among them, formula (14) is the user n exist t The upper and lower limits of the power load at any given time are constrained, and formula (15) represents the user's... n Total energy demand at all times; in, Indicates user n At any moment t The electrical load, Representing users respectively n At any moment t The upper limit of the power load below, Representing users respectively n At any moment t The lower limit of the power load, User n The sum of electrical loads at all times.

6. A distributed optimization method for P2P energy trading of unit combinations as described in claim 5, characterized in that: In step S1.4, the P2P energy trading model is as follows: (16) (17) (18) (19) (20) In this context, formula (16) is the objective function for minimizing costs, formula (17) includes various constraints from both the user and power generation sides, and formula (18) represents the expected social welfare of all prosumers. Representing the minimum expected social welfare of all producers and consumers, formula (19) is the producer's... n The cost-utility function, i.e. Indicates producer n The cost-utility function, formula (20) is the consumer's n The utility function, i.e. Consumers n Utility function; , , It is a positive constant of the cost-utility function. , , It is a positive constant of the utility function. It is a thermal power unit n Startup cost coefficient; Transforming the P2P energy trading model into matrix form yields: (21) (22) (23) (24) Wherein, formula (21) is the matrix form of the cost minimization objective function, and formula (16) is... Specific manifestations; yes The feasible set, It is the first n The feasible solutions of each producer and consumer constitute a nonconvex set formed by the mixed integer nonlinear constraints (17). , Represents the set of all decision variables. Indicates thermal power unit n The start / stop state vector, Indicates thermal power unit n The startup operation vector, Indicates thermal power unit n The shutdown operation vector, Indicates in the scene s Lower thermal power unit n The output vector, Represents each producer and consumer n The net injected power vector; Indicates the first n The constant matrix of each producer and consumer, Z This indicates the coupling relationship between producers and consumers. This indicates the introduced auxiliary variable. , Indicates the first n Auxiliary variables introduced by individual producers and consumers.

7. A distributed optimization method for P2P energy trading of unit combinations as described in claim 1, characterized in that: In step S2, the specific implementation process of reshaping is as follows: S2.1.: Transform the non-convex set in the P2P energy trading model into a convex set, resulting in: (25) (26) (27) (28) Where, conv( X n ) represents a nonconvex set The vertices of the convex hull; S2.2: The Lagrange dual function of the primal problem for: (29) (30) (31) (32) Wherein, formula (29) is the convex relaxation dual function of the original problem, i.e. Describe the convex relaxation dual function of the original problem. It is the set of dual multiplier vectors; Indicates the first n Equation constraints corresponding to individual producers and consumers The transpose of the dual multiplier vector. Indicates producer-consumer n The constant matrix, Indicates producer-consumer n Auxiliary variables, This represents the transpose of the auxiliary variable set. n Indicates the first n Individual consumers, Let represent the set of dual multiplier vectors. Indicated Transpose; S2.3: The augmented Lagrangian function of the original problem is: (33) (34) Among them, formula (33) is the augmented Lagrangian function form of the original problem. Represent the augmented Lagrangian function of the original problem. ρ Indicates the penalty parameter. Indicates producer-consumer n With neighboring producers and consumers m Trading volume between Indicates producer-consumer m With neighboring producers and consumers n Trading volume between Indicates producer-consumer n and consumers m The average of the transaction volume between them; S2.4: The approximation of the augmented Lagrangian function of the original problem is: (35) (36) (37) (38) Among them, formula (35) is the mirror image of all augmented Lagrange functions, i.e. Describes the mirror function of all augmented Lagrange functions. Indicates the first k Next iteration of the consumer n The mirror function of the augmented Lagrange function, formula (36) is the objective function. f n In the k In the nth iteration n Feasible solutions for individual consumers The first-order Taylor expansion at is given by formula (37), which is the tangent plane approximation function of all augmented Lagrangian functions, i.e. Denotes the tangent plane approximation function of all augmented Lagrange functions. Indicates the first k Next iteration of the consumer n The tangent plane approximation function of the augmented Lagrange function, formula (38) is based on The approximate function of the tangent plane centered at the center is, i.e. Let represent the approximate function of the tangent plane, and let this approximate function of the tangent plane satisfy . , Describe the objective function f n At point The first derivative at that point, Indicates the first k Feasible solutions in the next iteration. Indicates the first k Next iteration of the consumer n Auxiliary variables introduced.

8. A storage device storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

9. A distributed optimization device for P2P energy trading of unit combinations, characterized in that: include: A processor and a storage device; the processor loads and executes instructions and data in the storage device to implement the distributed optimization method for P2P energy trading of unit combinations as described in any one of claims 1 to 7.

Images