A fully decentralized distributed power transaction method and system based on P2P transaction

Abstract

The application relates to a fully-decentralized distributed power transaction method and system based on P2P transaction, which comprises the following steps: constructing a basic model of P2P transaction in a power distribution system, including an optimization target of taking the minimum total power cost of the power distribution system where all distributed transaction subjects are located as a target function, and distributed transaction subject output constraints, power transaction constraints and multi-period alternating current flow constraints based on second-order cone programming; based on a standard alternating direction multiplier method, the basic model is reconstructed into a standard consistency optimization problem, and non-convex constraints based on the CCP principle are introduced; based on the standard alternating direction multiplier method, a fully-decentralized communication mechanism is introduced, and an optimal transaction strategy is obtained by iterative solving under the condition of protecting individual data privacy and with minimum communication cost. The application can realize fully-decentralized distributed power transaction based on P2P transaction under the conditions of meeting the requirements of power distribution system operation physical feasibility and market participant privacy protection.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power system optimization scheduling, and in particular to a completely decentralized distributed power transaction method based on P2P transaction. BACKGROUND

[0002] In recent years, the rapid development of distributed energy such as photovoltaic power generation (PV) and combined heat and power (CHP) has brought great changes to the production, transmission and consumption modes of electric energy, and the proportion of producers and consumers with power generation and consumption capabilities in the distribution system has gradually increased. Under the situation of the continuous advancement of the new round of power system reform, these independent decision-making electric energy producers and consumers will actively participate in the competition in the electricity market, so that the use of a more flexible and effective electric energy transaction mechanism in the distribution system can effectively optimize resource allocation and improve the income of distributed power generation. This means that producers and consumers can not only reduce their electricity costs through their own distributed energy generation, but also can obtain income by selling excess electricity.

[0003] P2P transaction is an important mode of distribution network electric energy transaction. Treating producers and consumers containing distributed energy as distributed transaction subjects, P2P transaction allows each subject to trade with each other for the remaining electric energy, which can not only improve the utilization rate of renewable energy in the distribution system, but also bring certain benefits to the distribution system operator (DSO), such as reducing the maximum demand of the distribution system, reducing the standby demand, improving the system reliability, etc. In the distribution system, the existing traditional centralized transaction mode faces two main problems: first, a large number of distributed transaction subjects participating in the market will lead to high operating costs and low efficiency of the transaction center; second, the centralized transaction mode requires all market participants to communicate with the transaction center in both directions, which cannot meet the demand for privacy protection. SUMMARY

[0004] The purpose of the present application is to provide a completely decentralized distributed electric energy transaction method and system based on P2P transaction, which can realize completely decentralized distributed electric energy transaction while meeting the needs of distribution system operation physical feasibility and market participant privacy protection.

[0005] To achieve the above purpose, the present application adopts the following technical solutions.

[0006] A completely decentralized distributed electric energy transaction method based on P2P transaction, comprising:

[0007] Obtaining the required system basic technical data, market participant operating parameters, and price data;

[0008] The basic model for P2P transactions in the power distribution system is constructed, including an optimization objective with the goal of minimizing the total power cost of the power distribution system in which all distributed trading entities are located, as well as output constraints of distributed trading entities, power trading constraints, and multi-period AC power flow constraints based on second-order cone programming.

[0009] Based on the standard alternating direction multiplier method, the basic model is reconstructed into a standard consistency optimization problem, and a non-convex constraint based on the CCP principle is introduced;

[0010] Based on the standard alternating direction multiplier method, a fully decentralized communication mechanism is introduced to facilitate information exchange among distributed entities without the need for a transaction center. This results in the fully decentralized ADMM algorithm, which iteratively solves the problem with minimal communication cost while protecting individual data privacy, thereby obtaining the optimal transaction strategies for each distributed entity, the power grid, and P2P transactions.

[0011] As a further improvement to the present invention, the objective function is:

[0012] The objective function is to minimize the total energy cost of the distribution system in which the distributed entity resides:

[0013] (1)

[0014] In the formula: It is the set of all nodes where distributed transaction entities reside; It is the main body The cost of purchasing electricity from the grid; It is the main body Electricity costs paid through P2P transactions; It is the main body The cost of natural gas consumed by the combined heat and power (CHP) units, including:

[0015] (2)

[0016] (3)

[0017] (4)

[0018] In equation (2): the cost of purchasing electricity from the power grid consists of two parts: electricity consumption fee and demand fee. for Time-of-use electricity pricing at any given moment; as the main body Electricity purchased from the power grid; This refers to the unit cost of electricity based on demand. For the maximum demand; in equation (3): as the main body from P2P transaction price of electricity purchase; For the main body With Transaction electricity purchase; For the network fee; formula (4) is: For the unit cost of natural gas; For natural gas consumption.

[0019] As a further improvement of the present application, the basic model of constructing P2P transaction in power distribution system also constructs constraint conditions, including:

[0020] Distributed transaction subject output constraint, including: technical constraint of combined heat and power unit, upper and lower limit constraint of photovoltaic panel output, reactive power balance and heat load supply sufficiency constraint;

[0021] Electricity transaction constraint of each subject, including: net load operation demand and its electricity transaction constraint, maximum demand constraint of transaction with power grid, bilateral constraint of P2P transaction;

[0022] Multi-period alternating current flow constraint based on second-order cone programming, including: node power balance constraint, node voltage equation, Branch Flow line flow constraint after relaxation, node voltage amplitude square constraint, line current amplitude square constraint;

[0023] The constraint condition specifically includes:

[0024] The distributed transaction subject output constraint is:

[0025] (5a)

[0026] (5b)

[0027] (5c)

[0028] (6a)

[0029] (6b)

[0030] (6c)

[0031] (7a)

[0032] (7b)

[0033] Wherein, formula (5) represents the reactive power balance and heat load supply sufficiency of the main body , is the output reactive power of SVG, is the output reactive power of the photovoltaic, is the reactive load, is the reactive capacity of the SVG, is the thermal energy output of the CHP unit, is the thermal load; equation (6) is the technical constraint of the CHP unit, which limits its electric energy output range, natural gas consumption characteristics, and the relationship between the output electric energy and thermal energy, is the electric energy output of the CHP unit, and are the upper and lower limit values of the output electric energy, is the low calorific value of the natural gas, is the efficiency factor of the CHP unit, is the loss coefficient; equation (7) is the photovoltaic output constraint, is the electric energy output of the photovoltaic, is the predicted value of the photovoltaic output, and are the apparent capacity and active capacity of the photovoltaic;

[0034] The electric energy transaction constraint of each subject is:

[0035] (8)

[0036] (9)

[0037] (10)

[0038] wherein equation (8) indicates that the subject satisfies its net load demand through the transaction amount with the power grid and the P2P transaction amount equation (9) is the maximum demand limit, is the maximum demand; equation (10) indicates the bilateral constraint of the P2P transaction between users, is the purchase amount of electric energy of the subject from ;

[0039] The multi-period alternating current flow constraint based on the second-order cone programming is:

[0040] (11a)

[0041] (11b)

[0042] (11c)

[0043] (12)

[0044] (13)

[0045] (14a)

[0046] (14b)

[0047] (14c)

[0048] (15)

[0049] (16)

[0050] Equation (11) represents the node injection power constraint and node reactive power balance constraint for end-user nodes and intermediate nodes. For lines with the same sending node The set of receiving nodes, and For the line The trend of meritorious and ineffective actions and For line resistance and reactance, Let be the square of the line current; Equation (12) represents the node voltage equation without considering the phase angle. For nodes The square of the voltage, The system is a set of lines; Equation (13) represents the Branch Flow line power flow constraint. The relaxed second-order cone constraint; Equation (14) is the active / reactive power balance constraint of the root node connected to the power grid. and It is the active / reactive power provided by the power grid. Equation (15) represents the power factor limiting factor of the power grid; Equation (15) represents the square constraint of the node voltage amplitude. and Equation (16) represents the upper and lower limits of the node voltage amplitude; Equation (16) represents the square constraint of the line current amplitude. This represents the maximum current in the line.

[0051] As a further improvement of the present invention, the basic model is reconstructed into a standard consistency optimization problem based on the standard alternating direction multiplier method, and a non-convex constraint based on the CCP principle is introduced, including the following steps:

[0052] The power distribution system is divided into several regions according to an appropriate zoning strategy. Each distributed trading entity belongs to a region and is responsible for performing local calculations based on the local variables assigned to its region.

[0053] Based on the variable replication rule, the coupling constraints in the basic model are separated by region, including:

[0054] The sending boundary node and the receiving boundary node are divided, the local variables at the boundary nodes in the adjacent region are determined, and the corresponding replicated variables are determined, the coupling constraints of each region are re-expressed as the constraints of the local variables and the replicated variables in the region, so that all the constraints can be separated by region; and the global variables are introduced to connect the replicated variables and the original local variables in the adjacent region, so as to ensure that the replicated variables and the corresponding original local variables in the adjacent region have the same value;

[0055] Based on the standard alternating direction multiplier method and the variable replication rule, the basic model is reconstructed into a standard consistency optimization problem in a compact format, and , and represent all local variables, global variables and corresponding Lagrange multipliers of the region , based on the ADMM solving algorithm, the reconstructed model will be iteratively solved according to the following three steps: -Update, -Update, -Update;

[0056] Based on the CCP principle, the non-convex constraint corresponding to the convex relaxed line flow constraint is introduced in the -Update model, and the non-convex constraint is first-order Taylor expanded at the optimal solution, and the non-convex constraint is linearized, and the subsequent model solving process will be iteratively solved.

[0057] As a further improvement of the application, the coupling constraints in the basic model are separated by region, and the variable replication rule used is:

[0058] The line connecting two adjacent regions is defined as a tie line, the sending boundary node is defined as the node where the tie line flows out, and the receiving boundary node is defined as the other node on the tie line; the coupling variable at the boundary node is defined as a local variable, and the local variable replicated to the adjacent region is defined as a replicated variable;

[0059] Rule 1: for the sending boundary node, the local variable in the region where the receiving boundary node is located is replicated to the region where the sending boundary node is located, as the replicated variable of the sending boundary node;

[0060] Rule 2: for the receiving boundary node, the local variable in the region where the sending boundary node is located is replicated to the region where the receiving boundary node is located, as the replicated variable of the receiving boundary node;

[0061] Rule 3: Introduce a global variable to connect the replicated variable and the original local variable, ensuring that the replicated variable and the corresponding local variable value of the adjacent region are the same.

[0062] As a further improvement of the present application, based on the standard alternating direction multiplier method, the standard consistency optimization problem is obtained by reconstructing the basic model as:

[0063]

[0064]

[0065]

[0066] (17)

[0067] (18a)

[0068]

[0069] (18b)

[0070]

[0071] (18c)

[0072]

[0073] (19a)

[0074]

[0075] (19b)

[0076] (20a)

[0077] (20b)

[0078] (21a)

[0079] (21b)

[0080] (22a)

[0081] (22b)

[0082] (23a)

[0083] (23b)

[0084] (24a)

[0085] (24b)

[0086] wherein, formula (17) is the reconstructed P2P transaction bilateral constraint, is the area of all nodes, is the area wherein the distributed agent is located, is the area corresponding to the area of the P2P transaction volume of the area where the agent is located; formula (18)-(20) is the reconstructed alternating current flow constraint, is the copy variable transmitted to the area ; formula (21)-(24) is the constraint for corresponding the copy variable and the original local variable through the global variable, is the global variable;

[0087] The consistency optimization problem of the above standard is written in a compact form:

[0088] First, define three vectors,

[0089]

[0090]

[0091]

[0092] wherein , and respectively represent all local variables, global variables and Lagrange multipliers of the area ; the reconstructed model will be based on the traditional AMDD algorithm, and is iteratively solved according to the three steps of -Update, -Update, -Update;

[0093] -Update

[0094] (25a)

[0095] (25b)

[0096] -Update

[0097] (26a)

[0098] (26b)

[0099] (26c)

[0100] (26d)

[0101] -Update

[0102] (27)

[0103] wherein, formula (25) can be calculated in parallel in each area; formula (26) represents updating After that, the global variable is updated to the average value of the copy variable and the original local variable; formula (27) updates all Lagrange multipliers.

[0104] As a further improvement of the application, based on the CCP principle, a non-convex constraint corresponding to the convex relaxed line flow constraint is introduced, and the -Update is:

[0105] (28)

[0106]

[0107] (29)

[0108] (30)

[0109] wherein, is the current iteration number of CCP, is a penalty factor, and the growth coefficient is , is a relaxation variable; formula (29) is a non-convex constraint After first-order Taylor expansion at the optimal solution of each iteration, the linearized constraint is obtained.

[0110] As a further improvement of the present application, on the basis of the standard alternating direction multiplier method, a fully decentralized communication mechanism is introduced, and the information exchange between the distributed subjects is carried out without the need for a transaction center, so that a fully decentralized ADMM algorithm is obtained. In the case of protecting individual data privacy, the optimal transaction strategy of each distributed subject and the power grid and P2P transaction is obtained by iterative solution with the minimum communication cost, including:

[0111] The related basic parameters of the iterative algorithm are set, and the iteration initial value of the related variables is set and The iteration parameters and the basic data are input into the reconstructed optimization model based on the ADMM algorithm for iterative solution. The global variables generated during the solution process and the corresponding Lagrange multiplier data are transmitted between the distributed transaction subjects by using a fully decentralized communication strategy until the system operation and transaction results that meet the convergence accuracy and relaxation accuracy are obtained.

[0112] The solution results include: the electric energy transaction amount of each subject and the power grid, the P2P electric energy transaction amount of each subject, the target function value, and the distributed power generation amount of each subject.

[0113] As a further improvement of the present application, the fully decentralized communication mechanism is:

[0114] The original local variable corresponding to the global variable is defined as the leading variable, and the region where the leading variable is located is the leading region. The fully decentralized communication mechanism follows the following:

[0115] The value of the replicated variable solved by -Update is sent to the leading region.

[0116] The leading region updates the global variable by -Update.

[0117] The updated global variable in the leading region is sent back to the corresponding adjacent region.

[0118] The fully decentralized ADMM iterative algorithm solving process is:

[0119] a Input data:

[0120] The basic technical data of the power distribution system obtained in step 1, the operating parameters of the market participants, and the price data are read, such as , and , and the price data , , , and .

[0121] b Initialization:

[0122] Set the iteration number of outer loop as ; Set the initial value of global variable and Lagrange multiplier

[0123] as 0; Set the penalty parameter as ; Set the convergence precision of primal residual and dual residual of iteration stop condition as and respectively; ; Set the iteration number of inner loop as ; Set the initial value of iteration as

[0124] ; Set the initial penalty coefficient as ; Set the penalty coefficient growth factor as ; Set the maximum penalty coefficient as ; Set the maximum allowable relaxation gap of iteration stop condition as ;

[0125] c Solving outer loop:

[0126] After receiving global variable and Lagrange multiplier , the system iteratively solves according to the three steps of -Update, -Update, -Update;

[0127] d Solving inner loop:

[0128] Each region solves -Update in parallel, and obtains ;

[0129] e Inner loop convergence:

[0130] If the stop condition (30) is met, it means that the inner loop result can meet the accuracy of relaxation, and the inner loop stops iteration; otherwise, let , and update the penalty coefficient according to equation (31), and continue to execute -Update in each region;

[0131] (30)

[0132] (31)

[0133] ​f Fully decentralized data communication:

[0134] will be sent to the master region, which will send the values of the replicated variables to the neighboring regions through -Update ;

[0135] The updated global variables of the master region will be sent back to the corresponding neighboring regions, and the -Update ;

[0136] g Outer loop convergence judgment:

[0137] If the stop conditions (32) and (33) are met, it means that the outer loop result can meet the convergence condition of the ADMM algorithm, and the outer loop stops iteration; otherwise, let , and continue to execute the outer loop;

[0138] (32)

[0139] (33)

[0140] h Output result:

[0141] The output result includes system operation and transaction results, including: the electricity trading volume of each subject and the power grid , the P2P electricity trading volume of each subject , the objective function value, the distributed power generation of each subject, and the system power flow result.

[0142] A fully decentralized distributed electricity trading system based on P2P transaction, comprising:

[0143] An acquisition module for obtaining the required system basic technical data, market participant operation parameters, and price data:

[0144] A construction module for constructing a basic model of P2P transaction in a power distribution system, including an optimization objective with the minimum total electricity cost of all distributed transaction subjects in the power distribution system as the objective function, and distributed transaction subject output constraints, electricity transaction constraints, and multi-period alternating current power flow constraints based on second-order cone programming;

[0145] A reconstruction module for reconstructing the basic model into a standard consistency optimization problem based on the standard alternating direction multiplier method, and introducing non-convex constraints based on the CCP principle;

[0146] The solving module is used for introducing a fully decentralized communication mechanism on the basis of the standard alternating direction multiplier method, performing information exchange between the distributed subjects without a transaction center, obtaining a fully decentralized ADMM algorithm, and solving iteratively with minimum communication cost while protecting individual data privacy to obtain optimal transaction strategies of the distributed subjects, the power grid and P2P transaction.

[0147] The application has the advantages that: based on the alternating current power flow constraint of the distribution network, the application establishes a fully decentralized distributed transaction mode on the basis of meeting the physical feasibility of P2P transaction, and each subject performing P2P transaction in the distribution network makes self-decision, transaction and settlement. The application can effectively solve the problems of low transaction efficiency, high transaction cost, inability to protect the privacy data of market participants, insufficient utilization rate of renewable energy, low economic benefit and non-optimal solution of the traditional centralized electric energy transaction method in the distribution system with increasing proportion of producers and consumers, and provides a more economical and private transaction method. Based on the alternating current power flow constraint of the distribution network, the application introduces the CCP principle to ensure the accuracy of the second-order cone relaxation, establishes a distributed transaction model on the basis of meeting the physical feasibility of P2P transaction, reconstructs and solves the model by using the fully decentralized ADMM algorithm, so that each subject performing P2P transaction can make self-decision, transaction and settlement, and the application realizes effective optimization of resource allocation, improves the income of distributed power generation and the utilization rate of renewable energy, and brings certain benefits to the distribution system operator (DSO) by reducing the maximum demand and reserve demand. By using the method provided by the application, the optimal transaction strategies of each distributed transaction subject in the distribution network, the power grid and P2P transaction can be obtained, and the fully decentralized distributed electric energy transaction based on P2P is realized. BRIEF DESCRIPTION OF DRAWINGS

[0148] Figure 1 A solving flowchart of the fully decentralized distributed electric energy transaction method based on P2P transaction is shown in the application.

[0149] Figure 2 A variable replication rule diagram taking a four-node system as an example is shown in the application.

[0150] Figure 3 A fully decentralized communication mechanism diagram taking a four-node system as an example is shown in the application.

[0151] Figure 4 An iteration solving flowchart of the fully decentralized ADMM algorithm is shown in the application.

[0152] Figure 5 A system diagram of the fully decentralized distributed electric energy transaction system based on P2P transaction is shown in the application. DETAILED DESCRIPTION

[0153] In order to make the person skilled in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by the person skilled in the art without creative labor should belong to the protection scope of the present application.

[0154] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0155] The present application relates to a fully decentralized distributed power transaction method based on peer-to-peer (P2P) transaction, which can optimize the power transaction strategy of producers and consumers with power generation and consumption capacity in the distribution system in a distributed form without the need for a transaction center, while effectively protecting the data privacy of each transaction subject. It includes the following steps: obtaining the required basic data from the relevant departments; constructing a basic model of P2P transaction in the distribution system, including the objective function and constraint conditions of the model; based on the standard alternating direction multiplier method (ADMM), a distributed optimization algorithm, the basic model is reconstructed into a consistency optimization problem in a compact form, and non-convex constraints based on the Convex-concave Procedure (CCP) principle are introduced to ensure the accuracy of the second-order cone relaxation; based on the traditional ADMM algorithm, a fully decentralized communication mechanism is introduced to realize the exchange of information between each transaction subject, and the optimal transaction strategy of each subject and the power grid and P2P is obtained by iterative solution with the minimum communication cost under the protection of individual data privacy, realizing the fully decentralized distributed power transaction based on P2P.

[0156] As shown in Figure 1 The present application provides a fully decentralized distributed power transaction method based on P2P transaction, which includes:

[0157] Obtaining the required system basic technical data, market participant operating parameters, price data:

[0158] Constructing a basic model of P2P transaction in the power distribution system, including an optimization objective of taking the minimum total electric energy cost of the power distribution system where all distributed transaction subjects are located as an objective function, and constraint conditions such as distributed transaction subject output constraint, electric energy transaction constraint and multi-period alternating current flow constraint based on second-order cone programming;

[0159] Based on the standard alternating direction multiplier method, the basic model is reconstructed into a standard consistency optimization problem, and non-convex constraints based on the CCP principle are introduced to ensure the feasibility of the flow result;

[0160] On the basis of the standard alternating direction multiplier method, a fully decentralized communication mechanism is introduced, information exchange between the distributed subjects is carried out without the need for a transaction center, a fully decentralized ADMM algorithm is obtained, the optimal transaction strategy of the distributed subjects and the power grid and P2P transaction is obtained by iterative solution with the minimum communication cost while protecting individual data privacy.

[0161] The present application not only effectively optimizes the resource allocation of the power distribution system, improves the distributed power generation income and the utilization rate of renewable energy, but also meets the demand of privacy protection, so that each subject participating in P2P transaction can make self-decision, transaction and settlement.

[0162] The method of the present application will be described in detail in combination with specific embodiments:

[0163] The present application provides a fully decentralized distributed electric energy transaction method based on P2P transaction, comprising the following steps:

[0164] Step 1: obtaining the required system basic technical data, market participant operating parameters and price data from relevant departments:

[0165] System basic technical data: number of nodes in the power distribution system, number of lines, node where the distributed subject is located, node load, network topology, line impedance and maximum carrying capacity, etc.

[0166] Market participant operating parameters: system operating period data, load demand of the distributed transaction subject, combined heat and power unit parameters, upper and lower limit values of photovoltaic output, etc.

[0167] Price data: two-part price data of electric energy transaction between the power distribution system and the power grid, price data of P2P transaction between the subjects, cost data of the distributed generator, etc.

[0168] Step 2: constructing a basic model of P2P transaction in the power distribution system, comprising the following steps:

[0169] Step 2.1: Constructing optimization objective: taking the total cost of electrical energy of all distributed transaction subjects in the power distribution system as the objective function, including the cost of purchasing electrical energy from the power grid, the cost of electrical energy paid through P2P transaction, and the cost of power generation of the distributed generator set;

[0170] Step 2.2: Constructing constraint conditions, including:

[0171] Output constraint of the distributed transaction subject, including: technical constraint of the combined heat and power unit, upper and lower limit constraint of the output of the photovoltaic panel, reactive power balance and supply sufficiency constraint of the heat load;

[0172] Electrical energy transaction constraint of each subject, including: net load operation demand and its electrical energy transaction constraint, maximum demand constraint of transaction with the power grid, bilateral constraint of P2P transaction;

[0173] Multi-period alternating current flow constraint based on second-order cone programming, including: node power balance constraint, node voltage equation, Branch Flow line flow constraint after relaxation, node voltage amplitude square constraint, line current amplitude square constraint.

[0174] Step 3: based on the distributed optimization algorithm of standard alternating direction multiplier method (ADMM), the basic model constructed in step 2 is reconstructed into a standard consistency optimization problem in a compact form; and considering that P2P transaction may cause reverse power flow, causing inaccuracy of second-order cone relaxation, a non-convex constraint based on the principle of Convex-concave Procedure (CCP) is introduced to ensure the feasibility of the flow result, including the following steps:

[0175] Step 3.1: dividing the power distribution system into several regions according to a proper zoning strategy, each distributed transaction subject belonging to a region and being responsible for local calculation according to the specified local variable of the region.

[0176] Step 3.2: based on the variable replication rule proposed in the application, separating the coupling constraints in the basic model according to regions, including:

[0177] dividing the sending boundary node and the receiving boundary node, determining the local variable at the boundary node in the adjacent region and the corresponding replicated variable, re-describing the coupling constraints of each region as the constraints of the local variable and the replicated variable in the region, so that all the constraints are regionally separable; and introducing a global variable to connect the replicated variable and the original local variable in the adjacent region, ensuring that the replicated variable and the corresponding original local variable in the adjacent region have the same value.

[0178] Step 3.3: based on the standard alternating direction multiplier method (ADMM) and the variable replication rule, the basic model in step 2 is reconstructed into a standard consistency optimization problem in a compact form, and the following definitions are made , and represent all local variables, global variables and corresponding Lagrange multipliers of the region , based on the ADMM solving algorithm, the reconstructed model will follow the three steps of -Update, -Update, -Update to solve iteratively.

[0179] Step 3.4: Based on the CCP principle, introduce non-convex constraints corresponding to the convex relaxed line flow constraints in the -Update model, and perform first-order Taylor expansion of the non-convex constraint at the optimal solution, linearize the non-convex constraint, which will be solved iteratively in the subsequent model solving process.

[0180] Step 4: Based on the traditional ADMM algorithm, introduce a fully decentralized communication mechanism to exchange information between distributed subjects without the need for a transaction center, obtain a fully decentralized ADMM algorithm, solve iteratively with minimal communication cost while protecting individual data privacy, and obtain the optimal trading strategy of each distributed subject and the power grid and P2P trading, realize fully decentralized distributed power trading based on P2P, including:

[0181] Set the relevant basic parameters of the iterative algorithm, such as the iteration index, model convergence accuracy, etc., set the initial values of related variables such as and ; input the iteration parameters and basic data obtained in step 1 into the optimization model reconstructed in step 3 based on the ADMM algorithm for iterative solving; the global variables and corresponding Lagrange multipliers generated during the solving process are transmitted between distributed trading subjects using a fully decentralized communication strategy until the system operation and trading results that meet the convergence accuracy and relaxation accuracy are obtained. The solving results include: the amount of power trading between each subject and the power grid, the amount of P2P power trading of each subject, the value of the objective function, the amount of distributed power generation of each subject, etc.

[0182] The application will be further described below with reference to the accompanying drawings, but the content of the application is not limited to this.

[0183] Referring to Figure 1 , a fully decentralized distributed power trading method based on P2P trading, the method comprising the following steps:

[0184] Step 1: When applying the model proposed in the application, the required system basic technical data, market participant operating parameters, and price data need to be obtained from relevant departments.

[0185] System basic technical data: number of nodes, number of lines, nodes where distributed subjects are located, node load, network topology, line impedance and maximum carrying capacity, etc.

[0186] Market participant operating parameters: system operating period data, load demand of distributed subjects, combined heat and power unit parameters, upper and lower limit values of photovoltaic output, etc.

[0187] Price data: two-part price data for power transactions between the distribution system and the grid, price data for P2P transactions between subjects, cost data of distributed generators, etc.

[0188] Step 2: Construct the basic model of P2P transactions in the distribution system, including building the optimization objective function and the constraint conditions.

[0189] The objective function is to minimize the total power cost of the distribution system where the distributed subjects are located:

[0190] (1)

[0191] In the formula: is the set of nodes where all distributed transaction subjects are located; is the cost of purchasing power from the grid; is the cost of power paid by the subject through P2P transactions; is the natural gas consumption cost of the combined heat and power unit owned by the subject, where: (2)

[0192] (3)

[0193] (4)

[0194] (5)

[0195] In formula (2), the cost of purchasing power from the grid consists of two parts, electricity charge and demand charge, is the time-of-use electricity price at the moment; is the electricity purchase amount from the grid by the subject; is the unit cost of demand charge; is the maximum demand. In formula (3): is the P2P transaction price for the subject to buy electricity from is the electricity purchase amount of the subject from ​​​​​​​​is the cost of passing through the grid. In equation (4), is the unit cost of natural gas; is the consumption of natural gas.

[0196] The distributed transaction subject output constraint is:

[0197] (5a)

[0198] (5b)

[0199] (5c)

[0200] (6a)

[0201] (6b)

[0202] (6c)

[0203] (7a)

[0204] (7b)

[0205] wherein equation (5) represents the reactive power balance and the supply sufficiency of the thermal load of the subject , is the output reactive power of the SVG, is the output reactive power of the photovoltaic, is the reactive load, is the reactive capacity of the SVG, is the thermal energy output by the CHP unit, is the thermal load; equation (6) is the technical constraint of the CHP unit, which limits the range of its electric energy output, the natural gas consumption characteristics, and the relationship between the output electric energy and thermal energy, is the electric energy output by the CHP unit, and are the upper and lower limit values of the output electric energy, is the low calorific value of the natural gas, is the efficiency factor of the CHP unit, is the loss coefficient; equation (7) is the photovoltaic output constraint, is the electric energy output by the photovoltaic, is the predicted value of the photovoltaic output, and are the apparent capacity and active capacity of the photovoltaic.

[0206] The electric energy transaction constraint of each subject is:

[0207] (8)

[0208] (9)

[0209] (10)

[0210] wherein, equation (8) represents the body of the transaction with the grid and the P2P transaction volume satisfies its net load demand; equation (9) is the maximum demand limit, is the maximum demand; equation (10) represents the bilateral constraints of the P2P transaction between users, is the body of the electricity purchase from .

[0211] The multi-period AC power flow constraints based on the second-order cone programming are as follows:

[0212] (11a)

[0213] (11b)

[0214] (11c)

[0215] (12)

[0216] (13)

[0217] (14a)

[0218] (14b)

[0219] (14c)

[0220] (15)

[0221] (16)

[0222] wherein, equation (11) represents the node injection power constraints and the node reactive power balance constraints of the end user node and the intermediate node, is the set of receiving nodes having the same sending node of the line, and are the active and reactive power flows of the line , and and are the line resistance and reactance, is the square of line current; equation (12) represents the node voltage equation without considering phase angle, is the voltage square of node , is the set of system lines; equation (13) is the Branch Flow line flow constraint after second-order cone relaxation; equation (14) is the active / reactive power balance constraint of the root node connected to the power grid, and is the active / reactive power provided by the power grid, is the power grid limited power factor; equation (15) is the node voltage amplitude square constraint, and is the upper and lower limit of the node voltage amplitude; equation (16) is the line current amplitude square constraint, is the maximum line current.

[0223] Finally, a P2P-based basic transaction model in the power distribution system is constructed:

[0224]

[0225]

[0226] Step 3: The above basic model is suitable for centralized solution, and to realize distributed solution, the basic model constructed in step 2 needs to be reconstructed into a standard consensus optimization problem in a compact form based on the distributed optimization algorithm of standard Alternating Direction Multiplier Method (ADMM); and a non-convex constraint based on the principle of Convex-concave Procedure (CCP) is introduced to ensure the accuracy of the second-order cone relaxation and the feasibility of the power flow result:

[0227] First, the system is divided into several regions according to a proper partitioning strategy, and each transaction subject belongs to a region and is responsible for distributed calculation according to the specified local variables in the region.

[0228] In the basic model constructed in step 2, the bilateral constraints (10) of P2P transaction and the alternating current flow constraints (11)-(13) are coupled constraints that cannot be naturally divided by region, and need to be copied to the adjacent region based on the two variable replication rules proposed in the application, so that the coupled constraints can be separated by region. Taking a four-node system of Figure 3 as an example, the variable replication rule is specifically explained:

[0229] The four-node system is divided into two regions, region 1 has nodes , , and region 2 has nodes ,​ , define the line connecting two areas as tie line , when the power flow direction is from to , is the sending boundary node, is the receiving boundary node.

[0230] Rule 1: for sending boundary node , copy the local variable in area 2 to area 1, denoted as respectively. .

[0231] Rule 2: for receiving boundary node , copy the local variable in area 1 to area 2, denoted as respectively. .

[0232] Rule 3: introduce global variable to connect the copied variable and the original local variable. For example, connect and with global variable .

[0233] After adding the copied variables in each area, the coupling constraints can be re-expressed as the constraints of local variables and copied variables in each area, so that all constraints are region-wise separable. After introducing global variables, it can be ensured that the copied variables and the corresponding original local variables in adjacent areas have the same values, which satisfies the feasibility of the reconstructed model.

[0234] Based on the distributed optimization algorithm of standard alternating direction method of multipliers (ADMM) and the variable copying rules, the basic model in step 2 is reconstructed into a standard consensus optimization problem:

[0235]

[0236]

[0237]

[0238] (17) (18a)

[0239]

[0240] (18b)

[0241]

[0242] (18c)

[0243]

[0244] (19a)

[0245]

[0246] (19b)

[0247] (20a)

[0248] (20b)

[0249] (21a)

[0250] (21b)

[0251] (22a)

[0252] (22b)

[0253] (23a)

[0254] (23b)

[0255] (24a)

[0256] (24b)

[0257] wherein, formula (17) is a reconstructed P2P transaction bilateral constraint, is a region a set of all nodes, is a region where the distributed agent is located, is the corresponding region of the P2P transaction volume of the region in which the replicated variable is located; formula (18)-(20) is a reconstructed alternating current flow constraint, is a replicated variable transmitted to the region ; formula (21)-(24) is a constraint for corresponding the replicated variable and the original local variable through a global variable, is a global variable.

[0258] The consistency optimization problem of the above standard can be written in a compact form:

[0259] First, define three vectors,

[0260]

[0261]

[0262]

[0263] where , and represent all local variables, global variables and Lagrange multipliers of region respectively. The reconstructed model will be based on the traditional AMDD algorithm, following the three steps of -Update, -Update, -Update to solve iteratively.

[0264] -Update

[0265] (25a)

[0266] (25b)

[0267] -Update

[0268] (26a)

[0269] (26b)

[0270] (26c)

[0271] (26d)

[0272] -Update

[0273] (27)

[0274] where formula (25) can be calculated in parallel in each region; formula (26) represents updating after, the global variable is updated to the average value of the copy variable and the original local variable; formula (27) updates all Lagrange multipliers.

[0275] Considering that P2P transaction may lead to reverse power flow, which causes the inaccuracy of second-order cone relaxation, non-convex constraints based on CCP principle are introduced to ensure the accuracy of second-order cone relaxation and the feasibility of power flow results, and -Update expression is expressed in the form of CCP solution:

[0276] (28)

[0277]

[0278] (29)

[0279] (30)

[0280] wherein, is the current iteration number of CCP, is the penalty factor, and the growth coefficient is , is the relaxation variable; equation (29) is the non-convex constraint At each iteration, the first-order Taylor expansion is performed at the optimal solution , and the linearized constraint is obtained.

[0281] Step 4: On the basis of the traditional ADMM algorithm, a fully decentralized communication mechanism is introduced to exchange information between distributed subjects without the need for a transaction center, and a fully decentralized ADMM algorithm is obtained. In the case of protecting individual data privacy, the optimal transaction strategy of each distributed subject and the power grid and P2P transaction is obtained by iterative solution with the minimum communication cost, and a fully decentralized distributed power transaction based on P2P is realized.

[0282] Take Figure 3 for example to illustrate the fully decentralized communication mechanism:

[0283] First, define the original local variable corresponding to the global variable as the leading variable, and the region where the leading variable is located as the leading region. Corresponding to the four-node system in Figure 2 , it can be seen that for global variables and , the leading variable is and , and the leading region is region 1. For global variables , and , the leading variable is , , , and the leading region is region 2. Then the fully decentralized communication mechanism follows:

[0284] will The values of the replicated variables solved by the -Update are sent to the leading region, for example Figure 3 The updated values are sent to region 1 as shown by the medium red arrows.

[0285] The leading region updates the global variables by the -Update as shown by the yellow and gray boxes, and are updated in region 1 to ,

[0286] The updated global variables in the leading region are sent back to the corresponding neighboring regions, for example as shown by the blue arrows, and

[0287] The above communication mechanism can realize data exchange between two regions without the need for a transaction center, and the data exchanged between them is not sensitive, which can effectively protect the information privacy of each region itself.

[0288] Finally, the model is solved by using the fully decentralized ADMM algorithm based on the iteration solving flowchart shown in Figure 4 , to obtain the optimal transaction strategy and realize fully decentralized distributed power transaction based on P2P. The specific process is as follows:

[0289] The power distribution system grid data obtained in step 1, the operating parameters of market participants, price data, such as , and , and price data , , , and , etc. are read in.

[0290] The fully decentralized ADMM iteration process is set as the outer loop, and the CCP iteration process required for solving the -Update is set as the inner loop.

[0291] The number of iterations of the outer loop is set to ; the initial values of the global variables and the Lagrange multiplier are set to 0; the penalty parameter is set to ; and the convergence accuracies of the two items of the iteration stopping condition, the primal residual and the dual residual, are set to and , respectively.

[0292] The number of iterations of the inner loop is set to​​​​​ ; set the initial iteration value as ; set the initial penalty coefficient as ; set the penalty coefficient growth factor as ; set the maximum penalty coefficient as ; set the maximum allowable relaxation gap for the iteration stop condition as .

[0293] After receiving the global variable and the Lagrange multiplier , the system iteratively solves according to the three steps of -Update, -Update, -Update.

[0294] Each region solves -Update in parallel to obtain .

[0295] If the stop condition (30) is satisfied, it means that the inner loop result can satisfy the accuracy of relaxation, and the inner loop stops iteration; otherwise, let , and update the penalty coefficient according to equation (31), and continue to execute -Update in each region.

[0296] (30)

[0297] (31)

[0298] The values of the regional copy variables solved by -Update are sent to the master region, and the master region updates the global variable through -Update to obtain .

[0299] Subsequently, the global variable updated by the master region is sent back to the corresponding adjacent region, and -Update is executed in each region to obtain .

[0300] If the stop conditions (32) and (33) are satisfied, it means that the outer loop result can satisfy the convergence condition of the ADMM algorithm, and the outer loop stops iteration; otherwise, let , and continue to execute the outer loop.

[0301] (32)

[0302] (33)

[0303] The output results include system operation and transaction results, such as the electricity transaction amount of each subject and the power grid , the P2P electricity transaction amount of each subject , the objective function value, the distributed power generation amount of each subject, and the system power flow results.

[0304] As shown in Figure 5 , the present application also provides a fully decentralized distributed power transaction system based on P2P transaction, comprising:

[0305] An acquisition module is used to obtain the required system basic technical data, market participant operation parameters, and price data:

[0306] A construction module is used to construct a basic model of P2P transaction in the power distribution system, including taking the minimum total power cost of the power distribution system where all distributed transaction subjects are located as the objective function, including the cost of purchasing power from the power grid, the cost of paying for power through P2P transaction, and the power generation cost of the distributed generator set;

[0307] A reconstruction module is used to reconstruct the basic model into a standard consistency optimization problem based on the standard alternating direction multiplier method, and based on the non-convex constraint of the CCP principle;

[0308] A solution module is used to introduce a fully decentralized communication mechanism on the basis of the standard alternating direction multiplier method, to exchange information between each distributed subject without the need for a transaction center, to obtain a fully decentralized ADMM algorithm, to iteratively solve with the minimum communication cost in the case of protecting individual data privacy, and to obtain the optimal transaction strategy of each distributed subject and the power grid and P2P transaction.

[0309] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application and not to limit it, although the present application has been described in detail with reference to the above examples, those skilled in the art should understand that: the specific embodiments of the present application can still be modified or replaced, without departing from the spirit and scope of the present application, any modification or equivalent replacement, which should be covered within the protection scope of the claims of the present application.

Claims

1. A fully decentralized distributed electricity trading method based on P2P transactions, characterized in that, include: Obtain the necessary basic technical data of the power distribution system, operating parameters of market participants, and price data; A basic model for P2P transactions in a power distribution system is constructed, including an optimization objective that minimizes the total energy cost of the power distribution system in which all distributed trading entities reside, as well as constraints on the output of distributed trading entities, energy trading, and multi-period AC power flow based on second-order cone programming. The objective function is: The objective function is to minimize the total energy cost of the distribution system in which the distributed entity resides: （1） In the formula: It is the set of all nodes where distributed transaction entities reside; It is the main body The cost of purchasing electricity from the grid; It is the main body Electricity costs paid through P2P transactions; It is the main body The cost of natural gas consumed by the combined heat and power (CHP) units, including: （2） （3） （4） In equation (2): the cost of purchasing electricity from the power grid consists of two parts: electricity consumption fee and demand fee. for Time-of-use electricity pricing at any given moment; as the main body Electricity purchased from the power grid; This refers to the unit cost of electricity based on demand. For the maximum demand; in equation (3): as the main body from The price of electricity purchased through P2P transactions; as the main body and The amount of electricity purchased in the transaction; For internet access fees; in formula (4): Unit cost of natural gas; This refers to natural gas consumption. Based on the standard alternating direction multiplier method, the basic model is reconstructed into a standard consistency optimization problem, and non-convex constraints based on the CCP principle are introduced; specifically, the following steps are included: The power distribution system is divided into several regions according to an appropriate zoning strategy. Each distributed trading entity belongs to a region and is responsible for performing local calculations based on the local variables assigned to its region. Based on the variable replication rule, the coupling constraints in the base model are separated by region, including: The system divides the sending and receiving boundary nodes, determines the local variables and corresponding replication variables at the boundary nodes in adjacent regions, and reformulates the coupling constraints of each region as constraints of local variables and replication variables within the region, so that all constraints are separable by region. A global variable is introduced to connect the replication variables and the original local variables of adjacent regions, ensuring that the values ​​of the replication variables and the original local variables corresponding to the adjacent regions are the same. Based on the standard alternating direction multiplier method and the variable replication rule, the basic model is reconstructed into a standard consistency optimization problem in a compact format, and defined as follows: , and Representing regions All local and global variables, along with their corresponding Lagrange multipliers, are solved using the ADMM algorithm. The reconstructed model will then follow... -Update、 -Update、 The solution is obtained by iteratively solving the three steps: -Update; Based on the CCP principle, - The Update model introduces non-convex constraints corresponding to the line power flow constraints after convex relaxation, and performs a first-order Taylor expansion of the non-convex constraints at the optimal solution to linearize the non-convex constraints. Iterative solutions will be performed in the subsequent model solution process. Based on the standard alternating direction multiplier method, a fully decentralized communication mechanism is introduced to facilitate information exchange among distributed entities without the need for a transaction center. This results in a fully decentralized ADMM algorithm, which iteratively solves the problem with minimal communication cost while protecting individual data privacy. The algorithm yields the optimal transaction strategies for each distributed entity in relation to the power grid and P2P transactions, specifically including: Set the basic parameters of the iterative algorithm, set and Initial values ​​for the relevant variables are obtained during iteration; the iteration parameters and basic data are input into the reconstructed optimization model based on the ADMM algorithm for iterative solution; the global variables and corresponding Lagrange multiplier data generated during the solution process are transmitted between distributed trading entities using a fully decentralized communication strategy until the system operation and trading results that meet the convergence accuracy and relaxation accuracy are obtained. The solution results include: the amount of electricity traded between each entity and the power grid, the amount of P2P electricity traded between each entity, the objective function value, and the amount of distributed generation by each entity.

2. The fully decentralized distributed power trading method based on P2P transactions as described in claim 1, characterized in that: The basic model for P2P transactions in the power distribution system also establishes constraints, including: The output constraints of distributed trading entities include: technical constraints of cogeneration units, upper and lower limits of photovoltaic panel output, reactive power balance and sufficient supply of heat load; The electricity trading constraints for each entity include: net load operating demand and its electricity trading constraints, maximum demand constraints for grid transactions, and bilateral constraints for P2P transactions; Multi-period AC power flow constraints based on second-order cone programming include: node power balance constraints, node voltage equations, relaxed Branch Flow line power flow constraints, node voltage magnitude square constraints, and line current magnitude square constraints. The specific components of constructing constraints include: The output constraints of distributed transaction entities are: （5a） （5b） （5c） （6a） （6b） （6c） （7a） （7b） Among them, equation (5) represents the main body. The balance of reactive power and the adequacy of heat load supply, It is the output reactive power of the SVG. It is the reactive power output of photovoltaics. It is reactive load. It is the reactive power capacity of SVG. It is the heat energy output by the CHP unit. The heat load is given by equation (6), which represents the technical constraints of the CHP unit, limiting its power output range, natural gas consumption characteristics, and the relationship between output power and heat. The electrical energy output by the CHP unit. and These are the upper and lower limits of the output electrical energy. Due to the low calorific value of natural gas, The efficiency factor for the CHP unit. Here is the loss coefficient; Equation (7) is the photovoltaic output constraint. The electricity output by photovoltaics, This is the predicted value of photovoltaic output. and These are the apparent capacity and active capacity of photovoltaic power. The constraints on electricity trading for each entity are as follows: （8） （9） （10） Wherein, equation (8) represents the main body. Trading volume with the power grid and P2P transaction volume To meet its net load requirements; Equation (9) is the maximum demand limit. It is the maximum demand; Equation (10) represents the bilateral constraints of P2P transactions between users. as the main body from Electricity purchased; The multi-time-period AC power flow constraint based on second-order cone programming is: （11a） （11b） （11c） （12） （13） （14a） （14b） （14c） （15） （16） Equation (11) represents the node injection power constraint and node reactive power balance constraint for end-user nodes and intermediate nodes. For lines with the same sending node The set of receiving nodes, and For the line The trend of meritorious and ineffective actions and For line resistance and reactance, Let be the square of the line current; Equation (12) represents the node voltage equation without considering the phase angle. For nodes The square of the voltage, The system is a set of lines; Equation (13) represents the Branch Flow line power flow constraint. The relaxed second-order cone constraint; Equation (14) is the active / reactive power balance constraint of the root node connected to the power grid. and It is the active / reactive power provided by the power grid. Equation (15) represents the power factor limiting factor of the power grid; Equation (15) represents the square constraint of the node voltage amplitude. and Equation (16) represents the upper and lower limits of the node voltage amplitude; Equation (16) represents the square constraint of the line current amplitude. This represents the maximum current in the line.

3. The fully decentralized distributed power trading method based on P2P transactions as described in claim 1, characterized in that: The coupling constraints in the base model are separated by region, using the following variable copying rules: Define the line connecting two adjacent areas as a tie line, the power flow out node on the tie line as the sending boundary node, and the other node on the tie line as the receiving boundary node; define the coupling variable at the boundary node as a local variable, and the local variable copied to the adjacent area as the copied variable; Rule 1: For outgoing boundary nodes, copy the local variables in the region where the receiving boundary node is located to the region where the outgoing boundary node is located, and use them as copied variables for the outgoing boundary node. Rule 2: For the receiving boundary node, copy the local variables in the region where the sending boundary node is located to the region where the receiving boundary node is located, and use them as the copied variables of the receiving boundary node. Rule 3: Introduce a global variable to connect the copied variable and the original local variable, ensuring that the copied variable and the corresponding local variable in the adjacent region have the same value.

4. The fully decentralized distributed power trading method based on P2P transactions as described in claim 1, characterized in that: Based on the standard alternating direction multiplier method, the basic model is reconstructed to obtain the standard consistency optimization problem as follows: （17） （18a） （18b） （18c） （19a） （19b） （20a） （20b） （21a） （21b） （22a） （22b） （23a） （23b） （24a） （24b） Wherein, equation (17) represents the reconstructed bilateral constraints of P2P transactions. It is a region The set of all nodes It is a region The node where the distributed entity resides. yes The region corresponding to the P2P transaction volume in the area The replication variable in the equations; equations (18)-(20) are the reconstructed AC power flow constraints, To be transmitted to the area The copy variable in the equations (21)-(24) are constraints that map the copy variable to the original local variable through global variables. It is a global variable; The consistency optimization problem of the above standards can be written in a compact form: First, define three vectors. in , and Representing regions All local variables, global variables, and Lagrange multipliers; the reconstructed model will be based on the traditional AMDD algorithm, according to... -Update、 -Update、 The solution is obtained by iteratively solving the three steps: -Update; -Update （25a） （25b） -Update （26a） （26b） （26c） （26d） -Update （27） Equation (25) can be computed in parallel across regions; Equation (26) represents the update. Then, global variables Update to the average of the copied variable and the original local variable; Equation (27) updates all Lagrange multipliers.

5. The fully decentralized distributed power trading method based on P2P transactions as described in claim 1, characterized in that: Based on the CCP principle, The Update model introduces non-convex constraints corresponding to the line power flow constraints after convex relaxation, and performs a first-order Taylor expansion of the non-convex constraints at the optimal solution. -Update model is: （28） （29） （30） in, This represents the current iteration number of the CCP. As a penalty factor, its growth coefficient is , These are slack variables; Equation (29) represents a non-convex constraint. In each iteration of the optimal solution The linearization constraint is obtained after performing a first-order Taylor expansion at the given location.

6. The fully decentralized distributed power trading method based on P2P transactions as described in claim 5, characterized in that: A fully decentralized communication mechanism is as follows: Define the original local variable corresponding to the global variable as the dominant variable, and the region where the dominant variable resides as the dominant region. The fully decentralized communication mechanism follows these steps: Will -Update sends the value of the replicated variable obtained from the solution to the master region; Dominant region through -Update updates global variables; Send the updated global variables from the dominant region back to the corresponding adjacent regions; The solution process of the fully decentralized ADMM iterative algorithm is as follows: Input data: Read in the basic technical data of the power distribution system, the operating parameters of market participants, and price data obtained in step 1, such as... , and and price data , , , and ; b initialization: The fully decentralized ADMM iterative process is defined as the outer loop, in which the solution... -Update requires an inner loop for its CCP iteration process; Set the number of iterations for the outer loop to be... ; Setting global variables and Lagrange multipliers The initial value is 0; the penalty parameter is set to... ; The convergence accuracies of the original residual and dual residual, which are set as the iteration stopping conditions, are respectively... and ; Set the number of iterations for the inner loop to be... ; Set the initial value for iteration to The initial penalty coefficient is set to... The penalty coefficient growth factor is set to... ; Set the maximum penalty coefficient to ; Set the maximum allowable relaxation gap for the iteration stopping condition to be ; C solves the outer loop: Receive global variables and Lagrange multipliers Then, the system followed -Update、 -Update、 The solution is obtained by iteratively solving the three steps: -Update; d. Solve for the inner loop: Parallel solution in each region -Update, get ; e-loop convergence test: If the stopping condition (30) is met, it means that the result of the inner loop has met the accuracy of relaxation, and the inner loop stops iterating; otherwise, let The penalty coefficient is updated according to formula (31), and the partitioning is continued. -Update; （30） （31） f. Fully decentralized data communication: Will -Update sends the values ​​of the copy variables from each region to the dominant region, and the dominant region then... -Update updates global variables and obtains... ; Then update the global variables after the dominant region. Send it back to the corresponding adjacent area, and then divide it into zones. -Update, get ; Convergence of outer loop: If stopping conditions (32) and (33) are met, it means that the outer loop result has met the convergence condition of the ADMM algorithm, and the outer loop stops iterating; otherwise, let And continue executing the outer loop; （32） （33） h Output results: The output results include system operation and transaction results, including: the amount of electricity traded between each entity and the power grid. P2P electricity trading volume of each entity Objective function value, distributed generation of each entity, and system power flow results.

7. A fully decentralized distributed power trading system based on P2P transactions, used to execute the fully decentralized distributed power trading method based on P2P transactions as described in any one of claims 1 to 6, characterized in that, include: The acquisition module is used to obtain the necessary basic technical data of the power distribution system, operating parameters of market participants, and price data. The construction module is used to construct the basic model of P2P transactions in the power distribution system. It includes an optimization objective with the goal of minimizing the total power cost of the power distribution system in which all distributed trading entities are located, as well as output constraints of distributed trading entities, power trading constraints, and multi-period AC power flow constraints based on second-order cone programming. The reconstruction module is used to reconstruct the basic model into a standard consistency optimization problem based on the standard alternating direction multiplier method, and introduce non-convex constraints based on the CCP principle. The solution module introduces a fully decentralized communication mechanism based on the standard alternating direction multiplier method. It enables information exchange between distributed entities without the need for a transaction center, resulting in a fully decentralized ADMM algorithm. This algorithm iteratively solves the problem with minimal communication cost while protecting individual data privacy, thereby obtaining the optimal transaction strategy for each distributed entity in relation to the power grid and P2P transactions.

Images