Point-to-point transaction method, green certificate transaction method and market clearing method

Abstract

The invention relates to the technical field of distributed energy market transaction, in particular to a point-to-point transaction method, a green certificate transaction method and a market clearing method, and the method comprises the steps: obtaining local optimization parameters and collaborative target constraint conditions of a plurality of participants; decomposing a global optimization problem into local sub-problems of participants according to local optimization parameters and collaborative target constraint conditions, and defining consensus variables associated with a global target; initializing a consensus variable and a corresponding dual variable, and establishing a consensus alternating direction multiplier method to iteratively update the consensus variable; the participants solve local sub-problems in parallel based on the initialized consensus variables, and updated local variables are generated; a global consensus variable is calculated according to a communication network exchange local variable; according to the deviation between the global consensus variable and the local variable, the dual variable is updated, and when the deviation is smaller than a preset threshold value, a global optimization result is output. According to the invention, the problem of transaction and power distribution network stability constraint balance in the energy market is solved.

Description

Technical Field

[0001] The present invention relates to the field of distributed energy market trading technology, and in particular to a point-to-point trading method, a green certificate trading method and a market clearing method. Background Art

[0002] With the rapid development of peer-to-peer energy trading and renewable energy certificate (green certificate) mechanisms, the distributed energy market needs to simultaneously meet the multiple goals of economic transactions, green rights certification and safe grid operation.

[0003] In the existing technology, centralized optimization methods face the risk of privacy leakage due to the lack of a flexible architecture that supports asynchronous communication and encrypted boundary transmission and the need to aggregate global sensitive data. In addition, they are difficult to adapt to the high-frequency trading needs of large-scale distributed resources under scenarios with high real-time requirements or fluctuating network conditions, and are prone to iterative divergence or data leakage problems. Although traditional distributed algorithms can achieve decentralized computing, when dealing with the coordinated optimization of green certificate trading and distribution network stability constraints, they often suffer from low non-convex constraint conversion efficiency and lack of dynamic parameter adjustment mechanism, resulting in a decrease in clearing speed. In addition, they cannot effectively balance the relationship between renewable energy absorption benefits and grid node voltage and frequency stability, resulting in insufficient economic efficiency of market clearing results or safety hazards.

[0004] The information disclosed in this background technology section is only intended to deepen the understanding of the overall background technology of the present disclosure and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art known to those skilled in the art. Summary of the Invention

[0005] The present invention provides a point-to-point trading method, a green certificate trading method and a market clearing method, which can effectively solve the problems in the background technology.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: A point-to-point transaction method, comprising: Obtain local optimization parameters and collaborative objective constraints of multiple participants; Decomposing the global optimization problem into local sub-problems of each participant according to the local optimization parameters and the collaborative objective constraints, and defining consensus variables associated with the global objective; Initialize the consensus variable and the corresponding dual variable, establish a consensus alternating direction multiplication method to iteratively update the consensus variable; Each of the participants solves the local subproblems in parallel based on the initialized consensus variables to generate updated local variables; Calculating a global consensus variable by exchanging the local variables over the communication network; According to the deviation between the global consensus variable and the local variable, the dual variable is updated, and when the deviation is less than a preset threshold, the global optimization result is output.

[0007] Furthermore, the consensus variables include: Energy transaction volume boundaries, used to constrain the executable scope of peer-to-peer energy transactions among the participants; Green certificate trading volume quota, used to balance the supply and demand of renewable energy certificates; Stability constraint boundaries are used to characterize the allowable deviation range of distribution network node voltage and distribution network node frequency.

[0008] Furthermore, the communication network uses a homomorphic encryption protocol to encrypt and transmit the local variables, generate an encrypted transaction volume boundary, and only transmit the encrypted transaction volume boundary so that the original data cannot be reversely parsed.

[0009] Furthermore, solving the local subproblems in parallel includes: Identifying a non-convex stability constraint in the collaborative objective constraint, the non-convex stability constraint comprising at least one of a voltage amplitude fluctuation limit, a frequency deviation limit, or a line power transmission capacity limit at a distribution network node; Performing a projection operation on the non-convex stability constraint to generate an equivalent second-order cone programming formal constraint; A local optimization problem is constructed based on the second-order cone programming formal constraints, and the local optimization problem is solved by a convex optimization algorithm to generate an update result of the local variable that meets the stability requirements of the distribution network.

[0010] Furthermore, the dual variable update adopts a dynamic penalty parameter, including: Calculate the original residual and the dual residual of the current iteration round, the original residual represents the deviation amplitude of the local variable and the global consensus variable, and the dual residual represents the cumulative adjustment amplitude of the dual variable; Comparing the ratio of the original residual to the dual residual with the preset threshold, and if the ratio exceeds the preset threshold, increasing the dynamic penalty parameter by a preset proportional coefficient; If the ratio is lower than or equal to the preset threshold, reducing the dynamic penalty parameter by a preset proportional coefficient; The dual variable is updated based on the adjusted dynamic penalty parameter to drive the local variable and the global consensus variable to converge.

[0011] Furthermore, exchanging the local variables via a communication network includes: Defining an asynchronous communication time window, the length of which is set based on network delay tolerance and market clearing real-time requirements; For the participants who fail to complete the local variable update within the current time window, the corresponding historical global consensus variable is retained as a temporary reference value and marked as a delayed state; The non-delayed participants continue to perform iterative calculations based on the latest global consensus variables until the delayed participants complete the update and synchronize to the current iteration round, or the deviation between the local variables of the non-delayed participants and the global consensus variables reaches the convergence standard; When the delayed participant reconnects, a compensatory update is performed based on the corresponding historical global consensus variable and the dual variable of the current iteration round.

[0012] Furthermore, the collaborative target constraint conditions include: Energy trading supply and demand balance constraints: the total energy supply of each participant is dynamically matched with the total energy demand in each time period. The charging and discharging power of the energy storage equipment in the process is limited by the rated capacity and state of charge; Green certificate trading quota matching constraints: the green certificate generation of renewable energy sellers is linked to actual power generation and is subject to upper-level market quota restrictions. The green certificate purchase volume of renewable energy buyers must meet the consumption responsibility weight, and the transaction certificate must be delivered within the validity period; Distribution network stability constraints, including voltage amplitude, frequency deviation, and line power transmission capacity limitations.

[0013] Furthermore, the deviation is less than a preset threshold, including: The magnitude of the raw residual is less than a preset first threshold, the raw residual representing a direct deviation of the local variable from the global consensus variable; The magnitude of the dual residual is less than a preset second threshold, the dual residual representing the accumulated deviation of the dual variable update; The degree of violation of the stability constraint does not exceed the safety margin value set by the distribution network operator, and the safety margin value includes at least one of the number of times the node voltage amplitude exceeds the safe operating range, the duration of the line power exceeding the limit, or the accumulated value of the frequency deviation.

[0014] A green certificate trading method, the method comprising: Obtaining the local optimization parameters of the renewable energy seller, the local optimization parameters including the seller's actual renewable energy power generation, and obtaining the synergistic target constraint conditions of the renewable energy buyer, the synergistic target constraint conditions including the consumption responsibility weight of the buyer's region; Decomposing the global optimization problem of the green certificate trading market into the local sub-problems of the seller and the buyer according to the local optimization parameters and the collaborative objective constraints, and defining a green certificate trading volume quota consensus variable associated with the global objective; Initialize the green certificate trading volume quota consensus variable and the corresponding dual variable, establish the consensus alternating direction multiplication method to iteratively update the green certificate trading volume quota consensus variable; The seller and the buyer solve their respective local sub-problems in parallel based on the green certificate trading volume quota consensus variable to generate updated local variables. The seller's local variable includes the number of green certificates that can be sold, and the buyer's local variable includes the number of green certificates that need to be purchased. Calculate the global green certificate trading volume quota consensus variable based on the exchange of the local variables over the communication network; According to the deviation between the global green certificate trading volume quota consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the green certificate trading matching plan is output. The green certificate trading matching plan includes the green certificate delivery quantity and the corresponding transaction price.

[0015] A market clearing method, comprising: Obtaining the local optimization parameters of market participants, including energy demand data, green certificate quota data, and distribution network stability parameters, and obtaining collaborative target constraint conditions; Decomposing the global optimization problem of market clearing into the local sub-problems of each participant according to the local optimization parameters and the collaborative objective constraints, and defining the consensus variables associated with the global objective; Initialize the consensus variable and the corresponding dual variable, establish the consensus alternating direction multiplier method to iteratively update the consensus variable; Each of the market participants solves the local sub-problems in parallel based on the consensus variables to generate updated local variables; Exchanging the local variables according to the communication network and calculating the global consensus variable; According to the deviation between the global consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the market clearing result is output. The market clearing result includes the energy trading price, the green certificate trading matching plan and the injection adjustment instruction of the distribution network node power.

[0016] The technical solution of the present invention can achieve the following technical effects: This invention integrates the consensus alternating direction multiplier method framework, homomorphic encryption transmission and dynamic constraint projection technology to efficiently and collaboratively optimize green certificate transactions and distribution network stability constraints while ensuring data privacy and security, dynamically balance economic benefits and grid security, and enhance the flexibility and scalability of the distributed energy market.

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A flowchart of a peer-to-peer transaction method; Figure 2 This is a flow chart of the green certificate and peer-to-peer energy market transaction scenario in the distribution network; Figure 3 A schematic diagram of green certificates and multi-energy transactions in the peer-to-peer market; Figure 4 Schematic diagram of decentralized peer-to-peer energy market and centralized retail market; Figure 5 Energy and cost composition for end users in mixed energy markets. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0022] Embodiment 1; like Figure 1 As shown, the present application provides a point-to-point transaction method, the method comprising: S10: Obtaining local optimization parameters and collaborative objective constraints of multiple participants; S20: Decompose the global optimization problem into local sub-problems of each participant based on the local optimization parameters and collaborative goal constraints, and define consensus variables associated with the global goal; S30: Initialize the consensus variables and the corresponding dual variables, establish the consensus alternating direction multiplication method and iteratively update the consensus variables; S40: Each participant solves the local sub-problem in parallel based on the initialized consensus variables and generates updated local variables; S50: Calculate global consensus variables by exchanging local variables based on the communication network; S60: Update the dual variable according to the deviation between the global consensus variable and the local variable. When the deviation is less than a preset threshold, output the global optimization result.

[0023] Specifically, each participant (which can be a different computing unit, device or user) collects and organizes its own local optimization parameters (such as cost function, resource usage, etc.) according to its own optimization goals and constraints. Local optimization parameters include local decision variables and constraints. Each participant will also share its collaborative goal constraints. The collaborative goal constraints may involve the status or behavior of other participants to ensure that all parties follow the common goal during the optimization process; use mathematical methods (such as Lagrange multiplier method, KKT condition, etc.) to decompose the global optimization problem into multiple local sub-problems, each sub-problem is solved independently by different participants, and the optimization problem of each participant is designed so that It includes a consensus variable, which is the goal that all participants are concerned about, and the local problems of all participants depend on this consensus variable; the consensus variable represents the global goal shared by all participants, and the consensus variable is usually associated with the local decision variable of each participant, which may be the allocation of some resources, coordination of information or other forms of cooperative decision-making. The consensus variable can be a scalar or a vector, depending on the global goal and the structure of the sub-problem; at the beginning of the algorithm, the consensus variable (usually set to an initial value, such as a zero vector) and the dual variable (associated with the constraint condition) are initialized. The dual variable reflects the influence of each constraint in the global optimization process, and the initial value of the dual variable is the value of the initial value of the dual variable. The optimization is usually based on a reasonable starting assumption, which may be inferred through a priori analysis of existing data; the consensus alternating direction multiplier method (ADMM) is used to solve distributed optimization problems. Each participant updates according to the local subproblem and consensus variables. In each round of iteration, the participant first solves in his own local subproblem to obtain local variables; then, based on the deviation between the global consensus variable and the local variable, the consensus variable is adjusted, and the dual variable is used for feedback update; each participant independently solves his own local optimization problem based on the shared initialization consensus variable to obtain the updated local decision variables (such as local resource allocation, decision strategy, etc.). The optimization problem of each participant may include The objective function may be linear or nonlinear, and the constraints may be related to the states of other participants. Participants exchange their calculated local variables through a communication network. After the exchange, all participants will share the updated local variable information and calculate the global consensus variable. Usually, some aggregation method such as weighted average or summation is used to update the global consensus based on the local variables of each participant. The deviation between the global consensus variable and the local variable is calculated. This deviation is used to evaluate the convergence of the global optimization problem. The dual variable update is adjusted according to the deviation to improve the coordination of the global consensus variable and the local variable. The update rule is based on the principle of the ADMM method and reduces the deviation by adjusting the dual variable.When the deviation between the local variables of all participants and the global consensus variables is less than a preset threshold, it is considered to have converged. At this point, the entire system has achieved the global optimization goal and outputs the final global optimization result, which is usually the optimal solution obtained after the collaboration of all participants.

[0024] Through the technical solution of the present invention, under the premise of ensuring data privacy and security, green certificate trading and distribution network stability constraints are efficiently and collaboratively optimized, dynamically balancing economic benefits and grid security, and enhancing the flexibility and scalability of the distributed energy market.

[0025] Specifically, consensus variables include: Energy transaction volume boundaries, used to constrain the scope of peer-to-peer energy transactions that can be executed by participants; Green certificate trading volume quota, used to balance the supply and demand of renewable energy certificates; Stability constraint boundaries are used to characterize the allowable deviation range of distribution network node voltage and distribution network node frequency.

[0026] As a preferred embodiment of the above, the energy transaction volume boundary is used to constrain the amount of energy that can be exchanged between each participant in the point-to-point energy transaction. It is mainly set based on the resource status of the participants (such as power generation capacity, storage capacity, etc.) and transaction goals (such as meeting energy demand, reducing costs, etc.). Each participant first determines the maximum transaction capacity based on its power generation facilities (such as wind power, solar energy, gas, etc.) and its energy consumption needs. The boundary can be adjusted according to different market rules or power system scheduling requirements. For example, the maximum transaction volume of each participant cannot exceed its production capacity or the capacity of its energy storage equipment, or the minimum transaction volume of each participant can be based on the balance between its demand and supply capacity. Participants The energy transaction volume will be dynamically adjusted according to the real-time demand and supply of the electricity market, and will be updated regularly to ensure a balance between the stability and economy of the power grid operation. The green certificate trading volume quota is used to balance the supply and demand of renewable energy certificates (green certificates). A green certificate is a certificate that proves that a certain amount of renewable energy has been produced and put into use. It is usually a market mechanism established by the government or regulatory agency to encourage the development of renewable energy. Each participant obtains a green certificate based on the amount of renewable energy used or produced, and participates in green certificate trading. The supply and demand of green certificates are balanced in the market according to the production capacity, usage needs and market rules of each participant. Participants must follow The predetermined green certificate quota ensures that the proportion of renewable energy meets the requirements of the government or the market. For example, the green certificate quota of each participant is determined by the ratio of its renewable energy power generation to total energy consumption. As market demand changes, policies are adjusted, or renewable energy technology develops, the green certificate trading quota will be dynamically updated. Through transactions with other participants, the green certificate quota of the participant will change with its supply and demand situation; the stability constraint boundary is used to characterize the allowable deviation range of the voltage and frequency of the distribution network node. The stability of the distribution network is very important because any fluctuation beyond the set boundary may cause grid failure or unstable operation. Each participant must consider the quota in its power exchange. The voltage range of power grid nodes. For example, the voltage of each node must be maintained within a predetermined safety range. Voltages outside this range may affect the power supply of other nodes or cause equipment damage. The frequency of the distribution network must also be maintained within a set range. If the frequency deviates from the specified range, it may cause damage to power equipment or power supply interruptions. Participants' transactions must ensure that their behavior does not cause excessive frequency deviations. The voltage and frequency constraint boundaries of the distribution network will be dynamically adjusted based on real-time data and the operating status of the power grid. For example, if the voltage deviation of certain nodes is too large or the frequency fluctuates too frequently, the transaction volume and exchange strategy will be automatically adjusted to avoid affecting the stability of the power grid.Throughout the peer-to-peer transaction process, energy trading volume boundaries, green certificate trading volume quotas, and stability constraint boundaries are coordinated and updated using the consensus alternating direction method of multipliers (ADMM) algorithm. Each participant adjusts their behavior based on their local constraints and optimization objectives, guided by the consensus variables. Each participant's local optimization problem comprehensively considers these three boundary conditions, ensuring that each participant's behavior does not cause the transaction to exceed the agreed feasible range. At the same time, local interests are maximized while meeting system stability requirements. Consensus variables are repeatedly optimized to balance the needs and constraints of all parties, resulting in a consistent coordination result among all participants.

[0027] Furthermore, the communication network uses a homomorphic encryption protocol to encrypt and transmit local variables, generate encrypted transaction volume boundaries, and only transmit the encrypted transaction volume boundaries to make the original data irreversible.

[0028] As a preferred embodiment of the above, homomorphic encryption is an encryption scheme that allows calculations to be performed on encrypted data. Homomorphic encryption allows operations to be performed on data without decrypting it, and the correct result can be restored through decryption operations. Local variables (such as participants' energy trading volume, green certificate trading volume, stability constraints, etc.) are encrypted using homomorphic encryption protocols (such as Paillier encryption, BFV encryption scheme, etc.). Through homomorphic encryption, participants can encrypt their local variables and transmit them without exposing the content of the original data. Homomorphic encryption algorithms usually include public key and private key mechanisms. In this case, participants will use public keys to encrypt their local variables and transmit the encrypted data to other participants. Only participants with corresponding private keys can decrypt and obtain the original data. The trading volume boundary is used to constrain the trading volume between participants to ensure that the trading volume does not exceed an acceptable range. Through homomorphic encryption, the trading volume boundary can be calculated and transmitted in an encrypted state. Each participant first calculates its own trading volume boundary based on local optimization parameters and collaborative target constraints. The encrypted boundary data will be transmitted through the communication network. Due to encryption, other participants or the second The three parties cannot obtain specific transaction volume boundary information and can only see the encrypted value. During the peer-to-peer transaction process, encrypted data is transmitted through the communication network to ensure that the data cannot be accessed or tampered with by unauthorized third parties during transmission. After solving the local subproblem, each participant generates an encrypted transaction volume boundary and sends this encrypted data to other participants through a secure communication network (such as a VPN or TLS encrypted channel). To prevent man-in-the-middle attacks or data leaks, the communication protocol can further enhance security, such as using public-private key exchange protocols, digital signatures, and authentication technologies to ensure data integrity and authentication. After receiving the encrypted transaction volume boundary, participants can perform necessary calculations based on the homomorphic encryption protocol without decrypting the data. If they wish to share certain calculation results (such as transaction volume summation and balance) with other participants, they can use homomorphic encryption operations to perform operations with other participants without decrypting the data. For example, participant A can use the homomorphic encryption protocol to perform some calculation, such as addition, comparison, or other operations, on the encrypted transaction volume boundary with the encrypted boundaries of other participants to obtain the encrypted result.These encrypted results can be passed in subsequent transactions and calculations without exposing the original data; homomorphic encryption ensures that the encrypted data cannot be reverse-parsed to obtain the original data. Even during data transmission, third parties cannot restore the original data content by analyzing the encrypted data. The homomorphic encryption protocol ensures that only participants with private keys can decrypt the encrypted data and obtain real information. Even if the data is intercepted or stored during transmission, attackers cannot directly extract any useful original information from the encrypted data. For example, assuming that the transaction volume boundary of participant A is transmitted through homomorphic encryption and participant B cannot decrypt the data, then participant B can only perform calculations based on the encrypted data and cannot know the specific value of the transaction volume.

[0029] Furthermore, local subproblems are solved in parallel, including: Identify non-convex stability constraints in the collaborative objective constraints, where the non-convex stability constraints include at least one of voltage amplitude fluctuation limits, frequency deviation limits, or line power transmission capacity limits at distribution network nodes; Project the non-convex stability constraints to generate equivalent second-order cone programming constraints; A local optimization problem is constructed based on the constraints of the second-order cone programming form, and the local optimization problem is solved by a convex optimization algorithm to generate updated results of local variables that meet the stability requirements of the distribution network.

[0030] As a preferred embodiment of the above, non-convex stability constraints refer to constraints related to the stability of the distribution network. These constraints are mathematically non-convex, which usually make the optimization problem unsolvable or complex to solve. These constraints usually include voltage amplitude fluctuation limits, frequency deviation limits, line power transmission capacity limits, etc. at distribution network nodes; voltage amplitude fluctuation limits refer to that the voltage amplitude of each node in the distribution network may be affected by different power demands and power generation fluctuations, and the voltage fluctuation range needs to be limited; frequency deviation limits refer to that the frequency deviation of the power system cannot exceed the set limit, which is usually affected by changes in grid load and power generation; line power transmission capacity limits refer to that the power transmission capacity of each line in the distribution network is limited, and exceeding this limit may cause grid instability; the projection operation is used to convert non-convex constraints into equivalent convex constraints to ensure that the problem is transformed into a form that can be solved by the optimization algorithm. Through the projection operation, non-convex constraints such as voltage amplitude fluctuation limits, frequency deviation limits or line power transmission capacity limits at distribution network nodes can be converted into second-order cone constraints. After obtaining the constraints in the projected second-order cone programming form, they are used as constraints for the local optimization problem to construct a local optimization model. This optimization problem usually includes an objective function (such as minimizing cost and energy consumption) and a series of constraints (such as voltage, frequency, and power limits). These constraints include the original linear constraints and the converted second-order cone constraints to ensure that the stability of the power grid is not destroyed. The constructed local optimization problem is solved using modern convex optimization algorithms (such as interior point method and gradient descent method). The second-order cone programming problem can be efficiently solved using standard convex optimization solvers (such as MOSEK and CVXPY). The optimization process will be carried out in parallel in the local system of each participant. Each participant uses a convex optimization algorithm to solve the local subproblem based on its local variables and constraints and generates updated local variables. During the solution process, the second-order cone constraints are used to ensure that the local variables (such as voltage and frequency) do not exceed the allowable stability range. The results generated by the optimization algorithm meet the stability requirements of the distribution network.

[0031] Furthermore, the dual variable update uses dynamic penalty parameters, including: Calculate the original residual and dual residual of the current iteration round. The original residual represents the deviation amplitude of the local variable and the global consensus variable, and the dual residual represents the cumulative adjustment amplitude of the dual variable. The ratio of the original residual to the dual residual is compared with a preset threshold. If the ratio exceeds the preset threshold, the dynamic penalty parameter is increased by a preset proportional coefficient. If the ratio is lower than or equal to the preset threshold, the dynamic penalty parameter is reduced by a preset proportional coefficient; The dual variables are updated based on the adjusted dynamic penalty parameters to drive the convergence of local variables and global consensus variables.

[0032] As a preferred embodiment of the above, the original residual represents the deviation amplitude between the local variable and the global consensus variable. In each round of iteration, the larger the gap between the local variable and the global consensus variable, the larger the original residual, indicating that it has not converged yet; the dual residual reflects the cumulative adjustment amplitude of the dual variable, which measures the amount of correction of the global objective constraint by the dual variable during the optimization process. A larger dual residual indicates that the adjustment amplitude of the dual variable is larger, and further adjustment is still required to achieve convergence; the ratio of the original residual to the dual residual is calculated to evaluate the convergence of the current iteration. If the ratio is larger, it means that the gap between the local variable and the global consensus variable is larger, indicating that it has not converged yet; if the ratio is smaller, it means that it is close to convergence. The calculated ratio is compared with a preset threshold. The preset threshold is set according to the tolerance, usually a value less than 1. If the ratio exceeds the threshold, it indicates that a larger adjustment is required; if the ratio exceeds the preset threshold, it means that the current iteration is close to convergence. If the deviation between the local variable and the global consensus variable is large, the convergence speed needs to be accelerated. At this time, the dynamic penalty parameter is increased according to the preset proportional coefficient to enhance the adjustment of the dual variable, thereby accelerating the approach of the local variable and the global consensus variable. If the ratio is lower than or equal to the preset threshold, it means that it is close to convergence and the deviation is already small. Over-adjustment may cause instability. At this time, the dynamic penalty parameter is reduced according to the preset proportional coefficient to maintain stability and avoid over-adjustment. After the dynamic penalty parameter is adjusted, the dual variable is updated. This update will gradually reduce the deviation between the local variable and the global consensus variable. By continuously adjusting the penalty parameter, the optimization goal is gradually achieved while ensuring stability. By continuously adjusting the dynamic penalty parameter and updating the dual variable, the deviation between the local variable and the global consensus variable is gradually reduced, and finally they converge. In each iteration, with the appropriate adjustment of the penalty parameter, it can smoothly transition to the convergence state to ensure that the global optimization goal is achieved.

[0033] Furthermore, local variables are exchanged via the communication network, including: Define the asynchronous communication time window. The length of the asynchronous communication time window is set based on the network delay tolerance and the market clearing real-time requirements. For participants who fail to complete the local variable update within the current time window, the corresponding historical global consensus variable is retained as a temporary reference value and marked as delayed; The non-delayed participants continue to perform iterative calculations based on the latest global consensus variables until the delayed participants complete the update and synchronize to the current iteration round, or the deviation between the local variables of the non-delayed participants and the global consensus variables reaches the convergence standard; When a delayed participant rejoins, a compensatory update is performed based on the corresponding historical global consensus variable and the dual variable of the current iteration round.

[0034] As a preferred embodiment of the above, in the process of point-to-point transactions, due to the communication network delay between participants or the real-time requirement of market clearing, the local variable updates of participants may be asynchronous. Therefore, an asynchronous communication time window is defined to ensure that all participants can complete the update of local variables within the time window; the length of the asynchronous communication time window needs to be set according to two factors: one is the network delay tolerance, and the other is the real-time requirement of market clearing; the network delay tolerance determines the maximum acceptable communication delay within a certain time window; the real-time requirement of market clearing determines how much timeliness needs to be achieved in a trading cycle to ensure the normal operation of the market. For example, the window length can be set according to the load , communication conditions and the actual needs of the trading market; for those participants who fail to complete the update of local variables within the current time window, they will be marked as delayed. In this case, the participants in the delayed state will not update the global consensus variables. In order to ensure that these participants can resume synchronization with other participants in subsequent iterations, the historical global consensus variables of the participants in the delayed state are retained as temporary reference values, and continue to participate in the calculation based on this reference value until the participants in the delayed state can complete the update of local variables; for participants not marked as delayed, iterative calculations will continue to be performed based on the latest global consensus variables. Even if some participants fail to complete the update in time, other participants can still continue to use The latest global consensus variables are calculated; the iterative calculation will continue until two conditions are met: one is that the participants in the delayed state have completed the update of local variables and synchronized them to the current iteration round. At this time, the data of all participants will be completely synchronized; the other is that the deviation between the participants who are not delayed and the global consensus variables has reached the convergence standard, indicating that it is close to the optimal solution, and the updates of all participants have become relatively stable; when the participants in the delayed state reconnect and complete the update of local variables, they need to be updated through the compensation mechanism. The delayed participants will compensate according to their historical global consensus variables and the dual variables of the current iteration round. Specifically, the historical global consensus variables represent the reference data of the delayed participants in the previous iteration, and the current The dual variable is the current optimization result. Through this compensatory update, delayed participants can effectively adjust their update results to reduce the deviation with other participants and ensure the final convergence and global coordination. Throughout the process, the asynchronous communication mechanism enables each participant to participate in the global optimization according to its local computing process and network conditions. Participants in the delayed state can synchronize with other participants as quickly as possible in subsequent iterations by retaining historical global consensus variables and compensatory updates, reducing inconsistencies in the synchronization process. When all participants have completed the update and reached the convergence standard, the global optimization goal of the entire peer-to-peer transaction system will be achieved, and the local variables and global consensus variables of all participants will eventually converge to the optimal solution.

[0035] Furthermore, the collaborative goal constraints include: Energy trading supply and demand balance constraints: the total energy supply of each participant is dynamically matched with the total energy demand in each time period. The charging and discharging power of the energy storage equipment in the process is limited by the rated capacity and state of charge; Green certificate trading quota matching constraints: the green certificate generation of renewable energy sellers is linked to actual power generation and is subject to upper-level market quota restrictions. The green certificate purchase volume of renewable energy buyers must meet the consumption responsibility weight, and the transaction certificate must be delivered within the validity period; Distribution network stability constraints, including voltage amplitude, frequency deviation, and line power transmission capacity limitations.

[0036] As a preferred embodiment of the above, in each trading period, it is necessary to ensure that the total energy supply and total demand of all participants are dynamically balanced. Participants include power generators, load loaders, and energy storage devices with dual identities. Each participant needs to declare its power generation capacity or electricity demand, and the supply and demand situation is summarized based on the data of the entire network. The charging and discharging capacity of the energy storage device is constrained by the rated power of the device and the current state of charge (SOC, state parameter). The energy storage device cannot be charged beyond its capacity limit, nor can it be discharged when the SOC is too low. This part of the constraint imposes boundary conditions on the energy storage control variables in the local problem of each participant. The supply and demand balance is achieved by coordinating and matching the transaction volume declared by each participant in each round of iteration. And update the energy storage operation strategy in real time to respond to load fluctuations or uncertainties on the power generation side; renewable energy power generation participants generate green electricity certificates (green certificates) proportional to the power generation while generating electricity. The number of green certificates generated must match the market quota limit set by the superior energy regulatory department. The green certificate issuance mechanism can set different issuance ratios or coefficients according to the type of renewable energy (such as photovoltaic, wind power); green certificate buyers (usually electricity load side or traditional energy buyers) need to assume certain renewable energy absorption responsibilities according to national or regional policies, and need to purchase a certain proportion of green certificates to achieve compliance goals. The absorption responsibility is set by weight; all green certificate transactions must be completed within the validity period of the certificate, and green certificates that are not delivered after the expiration of the period will be delivered. Will expire, a time window should be set in the optimization model to ensure that the transaction flow and use are completed before the certificate expires. In the local optimization model, participants need to consider factors such as quota restrictions, actual power generation capacity, certificate validity period and the other party's trading needs when setting the green certificate trading volume. In the global coordination, the supply and demand of green certificates are matched, and priority is given to matching supply and demand and time-sensitive trading pairs; the voltage value of each distribution network node must be controlled within the specified range (for example, ±10% of the rated voltage) to prevent equipment damage or power outage risks caused by overvoltage or undervoltage. In the optimization model, voltage offset tolerance conditions are imposed on each trading plan. Trading plans that exceed the range will be judged as infeasible, and the power system frequency must be maintained within a small range above and below the standard value. Fluctuations within the range, excessive frequency deviation may trigger chain power grid accidents. The model needs to control the frequency disturbance caused by trading behavior, and restrict or impose penalty coefficients on participants with weak rapid adjustment capabilities. Each distribution line has a maximum safe transmission capacity. If the power flow caused by the transaction exceeds this capacity, there will be a risk of power grid overload. Line safety capacity constraints should be embedded in the local model and global coordination process to ensure that trading behavior does not violate the stability of the physical topology structure. All stability constraints can provide supporting data (such as power flow calculation results, node boundary conditions) through the real-time power grid status assessment system, which is embedded as input in each round of iterative optimization. The prediction model based on state estimation can be used to dynamically adjust the constraint range to improve flexibility.

[0037] Specifically, the deviation is less than a preset threshold, including: The magnitude of the raw residual is less than a preset first threshold, and the raw residual represents the direct deviation of the local variable from the global consensus variable; The magnitude of the dual residual is less than a preset second threshold, and the dual residual represents the accumulated deviation of the dual variable update; The degree of violation of the stability constraint does not exceed the safety tolerance value set by the distribution network operator. The safety tolerance value includes at least one of the number of times the node voltage amplitude exceeds the safe operating range, the duration of the line power exceeding the limit, or the accumulated value of the frequency deviation.

[0038] As a preferred embodiment of the above embodiment, the raw residual represents the direct deviation between the local variable and the global consensus variable. In the peer-to-peer transaction method, each participant generates a local variable based on the local optimization result, compares it with the global consensus variable, and calculates the residual. The raw residual is usually obtained by calculating the absolute difference between the local variable and the global consensus variable. The raw residual reflects whether the local variable is consistent with the global optimization goal. The preset first threshold is used to control the allowable deviation range of the raw residual. When the deviation between the local variable and the global consensus variable is greater than the threshold, it means that convergence has not been reached and optimization needs to be continued. If the raw residual is less than the preset first threshold, it means that the local variable is close to the global consensus variable, and the optimization process can be further performed, or it is considered to be close to convergence. In each iteration, the raw residuals of all participants are calculated and compared with the preset first threshold. When the raw residuals of all participants are less than the threshold, it can be considered that convergence has been achieved; the dual residual represents the cumulative deviation of the dual variable update. In peer-to-peer optimization, the dual variable is usually used to adjust the satisfaction of the global objective constraint. The dual residual reflects the historical update amplitude of the dual variable, that is, the relative value of the dual variable to the previous round in each round of iteration. The amount of change, the dual residual can be obtained by calculating the difference between the current dual variable and the previous round of dual variables. The preset second threshold is used to control the allowable range of change of the dual residual. If the amplitude of the dual residual exceeds the threshold, it means that the dual variable needs to be adjusted more and has not fully converged. When the amplitude of the dual residual is less than the preset second threshold, it means that the dual variable has stabilized and is close to convergence. In each iteration, the update amplitude of the dual variable (i.e., the dual residual) is checked and compared with the preset second threshold. If the dual residual is less than the threshold, it indicates that the dual variable has reached a sufficient convergence state. The optimization process can continue, or the system can be judged to be close to convergence. The stability constraints of the distribution network include voltage amplitude, frequency deviation, and line power transmission capacity. Each constraint has an allowable range. Exceeding this range may cause grid instability and affect the effectiveness of the transaction. The distribution network operator sets safety tolerances for voltage amplitude, frequency deviation, and line power. The degree of violation of each constraint must be controlled within these tolerances to ensure the safe operation of the distribution network. The voltage amplitude of each node in the distribution network should remain within a certain range. The operator will set a voltage deviation tolerance value (for example, ±10% of the rated voltage). If the node voltage exceeds this range, a voltage violation is recorded. The power transmitted by the line should not exceed the rated power of the equipment. If the line power exceeds the limit, the duration of the limit violation is recorded. When the limit violation lasts for more than the set maximum tolerance, it indicates that the stability constraint has been violated. Frequency deviation refers to the degree to which the actual frequency of the distribution network deviates from the standard frequency (such as 50Hz or 60Hz). The accumulated frequency deviation value reflects the total amount of frequency fluctuation during the entire transaction process.In each iteration, the system checks for violations of distribution network stability constraints, calculates voltage, power, and frequency deviations, and compares them with preset safety tolerances. If the degree of violation of certain distribution network constraints exceeds the set tolerance, the trading plan is automatically adjusted or necessary adjustments are made to ensure that the constraints are not seriously violated. When the degree of violation of all stability constraints does not exceed the tolerance, the distribution network stability is guaranteed and trading can continue.

[0039] Embodiment 2: like Figure 2 and Figure 3 As shown, based on the same inventive concept as the point-to-point transaction method in the aforementioned embodiment, the present invention also provides a green certificate transaction method, the method comprising: Obtain the local optimization parameters of the renewable energy seller, which include the seller's actual renewable energy generation, and obtain the collaborative target constraints of the renewable energy buyer, which include the consumption responsibility weight of the buyer's region; Based on local optimization parameters and collaborative objective constraints, the global optimization problem of the green certificate trading market is decomposed into local sub-problems for sellers and buyers, and a consensus variable for the green certificate trading volume quota associated with the global objective is defined; Initialize the green certificate trading volume quota consensus variable and the corresponding dual variable, establish a consensus alternating direction multiplication method to iteratively update the green certificate trading volume quota consensus variable; The seller and the buyer solve their respective local sub-problems in parallel based on the consensus variable of the green certificate trading volume quota, generating updated local variables. The seller's local variable includes the number of green certificates that can be sold, and the buyer's local variable includes the number of green certificates that need to be purchased. Calculate the global green certificate trading volume quota consensus variable based on the exchange of local variables over the communication network; According to the deviation between the global green certificate trading volume quota consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the green certificate trading matching plan is output. The green certificate trading matching plan includes the green certificate delivery quantity and the corresponding transaction price.

[0040] Specifically, the seller's local optimization parameters include its actual renewable energy generation (such as wind power, solar energy, etc.). These local optimization parameters determine the seller's supply capacity in the green certificate transaction. The seller also needs to provide energy storage capacity (if applicable) and other constraints, such as grid access capacity, project power generation scale, equipment efficiency, etc. All these factors will affect the number of green certificates it can sell. The seller will collect and transmit the total amount of renewable energy it produces, and provide this information to the trading platform as part of its local optimization model for subsequent calculations and optimization. The buyer's collaborative target constraints include the absorption responsibility weight, which indicates the proportion of green certificates the buyer needs to purchase. The absorption responsibility weight is usually determined based on the renewable energy absorption target of the region, aiming to ensure that the buyer fulfills the corresponding renewable energy absorption obligations. The absorption responsibility weight may be related to the buyer's electricity consumption, the region's renewable energy targets or regulatory requirements. The buyer determines the number of green certificates it needs to purchase by calculating its electricity consumption and the absorption responsibility weight. This information is used as the buyer's local optimization parameters for subsequent transaction calculations. The global optimization problem of the green certificate trading market is decomposed into local sub-problems of the seller and the buyer. The global goal is to coordinate the seller's available green certificates with the buyer's The seller's sub-problem is to make a green certificate sales decision based on its power generation (the number of green certificates available) and the price in the trading market. The buyer's sub-problem is to decide the number of green certificates to purchase based on its consumption responsibility weight to meet compliance requirements and optimize costs as much as possible. In order to coordinate the transactions between sellers and buyers, a green certificate transaction volume quota consensus variable is defined. The green certificate transaction volume quota consensus variable represents the overall transaction volume of the green certificate market that all participants (sellers and buyers) are concerned about. The green certificate transaction volume quota consensus variable helps ensure the supply and demand balance in the market and achieve the global optimal transaction matching solution. In the initial stage, for green The consensus variable for the trading volume quota is initialized with an initial value (e.g., zero or a reasonable guess). At the same time, the dual variable associated with this consensus variable is initialized. The dual variable is used to reflect the constraints of each participant in the transaction process and the impact of the global optimization goal. In subsequent iterations, these variables are updated using the consensus alternating direction multiplier method (ADMM). ADMM allows for efficient optimization in a distributed system, enabling sellers and buyers to coordinate while preserving their own data privacy. Each participant iteratively calculates the local subproblem based on the consensus variable, gradually improving the matching degree of the transaction plan.The seller and the buyer each solve their local sub-problems independently and in parallel based on the consensus variables. The seller calculates the number of green certificates that can be sold based on its actual renewable energy power generation, while the buyer calculates the number of green certificates to be purchased based on its consumption responsibility weight. The seller's local variable is the number of green certificates that can be sold, and the buyer's local variable is the number of green certificates that need to be purchased. Each participant makes the best decision based on the current consensus variables to achieve their respective goals. The seller calculates an optimal green certificate sales strategy based on market conditions, prices and demand, while the buyer calculates the optimal number of green certificates to be purchased based on its compliance requirements and budget. The seller and the buyer exchange their respective local variables (the number of green certificates that can be sold and the number of green certificates that need to be purchased) through the communication network. The global consensus variable for green certificate trading volume quotas (the total number of green certificates required to be purchased) can be calculated through information exchange, representing the total transaction volume of sellers and buyers. This global consensus variable is updated based on the exchange data between sellers and buyers to ensure a balance between supply and demand for green certificates in the global market. The dual variable is updated based on the deviation between the global consensus variable and each participant's local variable. The calculation of the deviation is a key step, determining the extent of adjustments made by each party. If the deviation is large, the adjustment of the dual variable is increased, and vice versa. When the deviation is less than a preset threshold, the transaction matching solution is considered to have converged, and the final green certificate transaction matching solution, including the green certificate delivery quantity and the corresponding transaction price, is output.

[0041] Embodiment 3; like Figure 4 and Figure 5 As shown, based on the same inventive concept as the point-to-point transaction method in the aforementioned embodiment, the present invention further provides a market clearing method, which includes: Obtain local optimization parameters of market participants, including energy demand data, green certificate quota data, and distribution network stability parameters, and obtain collaborative target constraints; Based on the local optimization parameters and collaborative objective constraints, the global optimization problem of market clearing is decomposed into local sub-problems for each participant, and consensus variables associated with the global objective are defined; Initialize the consensus variables and the corresponding dual variables, establish the consensus alternating direction multiplication method and iteratively update the consensus variables; Each market participant solves local sub-problems in parallel based on the consensus variables and generates updated local variables; Exchange local variables based on the communication network and calculate global consensus variables; According to the deviation between the global consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the market clearing result is output. The market clearing result includes the energy trading price, the green certificate trading matching plan and the injection adjustment instruction of the distribution network node power.

[0042] Specifically, market participants (including power generators, power users, energy storage equipment, etc.) provide their own local optimization parameters, including energy demand data, that is, the energy demand of the participants, which may involve load demand, power generation capacity, energy storage demand, etc., green certificate quota data, that is, the demand or supply capacity of the participants in the green certificate market, to determine their role in renewable energy transactions, distribution network stability parameters, including voltage, frequency deviation, power transmission restrictions, etc., to ensure that the transaction plan does not violate the stability constraints of the distribution network; all market participants also need to provide collaborative target constraints, which involve market rules, policy restrictions and transaction compliance to ensure that the behavior of each participant meets the global optimization goal, and each participant is based on its specific The local conditions of the entity (such as power generation, electricity demand, green certificate trading volume, etc.) provide optimization parameters, and ensure that these data can be transmitted to the market clearing system through a secure communication network; the global optimization problem of market clearing is decomposed into multiple local sub-problems. The global optimization problem involves the setting of energy trading prices, the matching of green certificate transactions, the stable operation of the distribution network and other goals. Each participant (such as power generators, electricity users, energy storage equipment, etc.) solves its local sub-problems according to local optimization parameters. In order to ensure the coordination of global optimization, consensus variables related to the global goals are defined. These consensus variables include energy trading prices, which determine the transaction prices of energy buyers and sellers, green certificate trading volume, which coordinates the supply and demand of green certificates of all parties, and power injection into the distribution network, which ensures the grid The power flow meets the stability requirements; the consensus variable, as part of the global goal, ensures that each participant, through coordination and update, jointly promotes the realization of the global optimization goal while satisfying their respective constraints; first initialize the consensus variable and the dual variable, and set an initial guess value. These variables include energy trading prices, green certificate trading volume, and distribution network power injection. The dual variable is used to reflect the impact of each participant on the constraints in the global optimization. It is the key variable required to coordinate the entire system and help each participant optimize its behavior to meet the global goal; the consensus variable is iteratively updated using the consensus alternating direction multiplier method (ADMM). ADMM is a problem suitable for distributed optimization. It gradually coordinates the local optimization problems of the participants. To achieve the global optimum, in each round of iteration, the local sub-problems and consensus variables of the participants are updated alternately; sellers, buyers and other participants (such as energy storage equipment, green certificate traders, etc.) solve local sub-problems in parallel based on the consensus variables, and sellers perform local optimization based on their power generation, energy demand and green certificate quota; buyers make green certificate purchase decisions based on the consumption responsibility weight and demand; distribution network nodes perform power scheduling based on power injection constraints, and each participant generates updated local variables based on local calculation results. For example, the seller's local variables include the amount of energy available for sale and the number of green certificates, etc., the buyer's local variables include the energy and number of green certificates to be purchased, and the local variables of distribution network participants include the amount of power injection that needs to be adjusted;All participants exchange their local variables through a secure communication network. These local variables include each participant's energy demand, supply, Green Certificate trading volume, and power injection. Global consensus variables are calculated based on the exchanged data. These consensus variables reflect the market's energy trading price, Green Certificate trading matching plan, and power injection adjustments for the distribution network. Dual variables are updated based on the deviation between the global consensus variables and each participant's local variables. The deviation reflects the gap between the participant's behavior and the global goal. The dual variables are adjusted based on the calculated deviation, gradually optimizing the global coordination mechanism. The magnitude of the adjustment depends on the magnitude of the deviation. When the deviation is large, the dual variable update amplitude is large, and vice versa. When the deviation is less than a preset threshold, the market clearing process is considered to have converged, and the final market clearing results are output. These results include the energy trading price (the final transaction price determined by market supply and demand), the Green Certificate trading matching plan (including the number of Green Certificates traded and the corresponding buyers and sellers), and the power injection adjustment instructions for distribution network nodes (power adjustment instructions to ensure grid stability).

[0043] Although the present application has been described with reference to specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and drawings are merely illustrative of the present application as defined herein and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, the present application is intended to include such modifications and variations as fall within the scope of the present application and its equivalents.

Claims

1. A point-to-point transaction method, characterized in that: The method comprises: Obtain local optimization parameters and collaborative objective constraints of multiple participants; Decomposing the global optimization problem into local sub-problems of each participant according to the local optimization parameters and the collaborative objective constraints, and defining consensus variables associated with the global objective; Initialize the consensus variable and the corresponding dual variable, establish a consensus alternating direction multiplication method to iteratively update the consensus variable; Each of the participants solves the local subproblems in parallel based on the initialized consensus variables to generate updated local variables; Calculating a global consensus variable by exchanging the local variables over the communication network; According to the deviation between the global consensus variable and the local variable, the dual variable is updated, and when the deviation is less than a preset threshold, the global optimization result is output.

2. The point-to-point transaction method according to claim 1, characterized in that: The consensus variables include: Energy transaction volume boundaries, used to constrain the executable scope of peer-to-peer energy transactions among the participants; Green certificate trading volume quota, used to balance the supply and demand of renewable energy certificates; Stability constraint boundaries are used to characterize the allowable deviation range of distribution network node voltage and distribution network node frequency.

3. The point-to-point transaction method according to claim 1, characterized in that: The communication network uses a homomorphic encryption protocol to encrypt and transmit the local variables, generate an encrypted transaction volume boundary, and only transmit the encrypted transaction volume boundary so that the original data cannot be reversely parsed.

4. The point-to-point transaction method according to claim 1, characterized in that: Solving the local subproblems in parallel includes: Identifying a non-convex stability constraint in the collaborative objective constraint, the non-convex stability constraint comprising at least one of a voltage amplitude fluctuation limit, a frequency deviation limit, or a line power transmission capacity limit at a distribution network node; Performing a projection operation on the non-convex stability constraint to generate an equivalent second-order cone programming formal constraint; A local optimization problem is constructed based on the second-order cone programming formal constraints, and the local optimization problem is solved by a convex optimization algorithm to generate an update result of the local variable that meets the stability requirements of the distribution network.

5. The point-to-point transaction method according to claim 1, characterized in that: The dual variable update adopts dynamic penalty parameters, including: Calculate the original residual and the dual residual of the current iteration round, the original residual represents the deviation amplitude of the local variable and the global consensus variable, and the dual residual represents the cumulative adjustment amplitude of the dual variable; Comparing the ratio of the original residual to the dual residual with the preset threshold, and if the ratio exceeds the preset threshold, increasing the dynamic penalty parameter by a preset proportional coefficient; If the ratio is lower than or equal to the preset threshold, reducing the dynamic penalty parameter by a preset proportional coefficient; The dual variable is updated based on the adjusted dynamic penalty parameter to drive the local variable and the global consensus variable to converge.

6. The point-to-point transaction method according to claim 1, characterized in that: Exchanging the local variables via a communication network comprises: Defining an asynchronous communication time window, the length of which is set based on network delay tolerance and market clearing real-time requirements; For the participants who fail to complete the local variable update within the current time window, the corresponding historical global consensus variable is retained as a temporary reference value and marked as a delayed state; The non-delayed participants continue to perform iterative calculations based on the latest global consensus variables until the delayed participants complete the update and synchronize to the current iteration round, or the deviation between the local variables of the non-delayed participants and the global consensus variables reaches the convergence standard; When the delayed participant reconnects, a compensatory update is performed based on the corresponding historical global consensus variable and the dual variable of the current iteration round.

7. The point-to-point transaction method according to claim 1, characterized in that: The collaborative target constraint conditions include: Energy trading supply and demand balance constraints: the total energy supply of each participant is dynamically matched with the total energy demand in each time period. The charging and discharging power of the energy storage equipment in the process is limited by the rated capacity and state of charge; Green certificate trading quota matching constraints: the green certificate generation of renewable energy sellers is linked to actual power generation and is subject to upper-level market quota restrictions. The green certificate purchase volume of renewable energy buyers must meet the consumption responsibility weight, and the transaction certificate must be delivered within the validity period; Distribution network stability constraints, including voltage amplitude, frequency deviation, and line power transmission capacity limitations.

8. The point-to-point transaction method according to claim 1, characterized in that: The deviation is less than a preset threshold, including: The magnitude of the raw residual is less than a preset first threshold, the raw residual representing a direct deviation of the local variable from the global consensus variable; The magnitude of the dual residual is less than a preset second threshold, the dual residual representing the accumulated deviation of the dual variable update; The degree of violation of the stability constraint does not exceed the safety margin value set by the distribution network operator, and the safety margin value includes at least one of the number of times the node voltage amplitude exceeds the safe operating range, the duration of the line power exceeding the limit, or the accumulated value of the frequency deviation.

9. A green certificate trading method, characterized in that: A point-to-point transaction method according to any one of claims 1 to 8 is used, wherein the method comprises: Obtaining the local optimization parameters of the renewable energy seller, the local optimization parameters including the seller's actual renewable energy power generation, and obtaining the synergistic target constraint conditions of the renewable energy buyer, the synergistic target constraint conditions including the consumption responsibility weight of the buyer's region; Decomposing the global optimization problem of the green certificate trading market into the local sub-problems of the seller and the buyer according to the local optimization parameters and the collaborative objective constraints, and defining a green certificate trading volume quota consensus variable associated with the global objective; Initialize the green certificate trading volume quota consensus variable and the corresponding dual variable, establish the consensus alternating direction multiplication method to iteratively update the green certificate trading volume quota consensus variable; The seller and the buyer solve their respective local sub-problems in parallel based on the green certificate trading volume quota consensus variable to generate updated local variables. The seller's local variable includes the number of green certificates that can be sold, and the buyer's local variable includes the number of green certificates that need to be purchased. Calculate the global green certificate trading volume quota consensus variable based on the exchange of the local variables over the communication network; According to the deviation between the global green certificate trading volume quota consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the green certificate trading matching plan is output. The green certificate trading matching plan includes the green certificate delivery quantity and the corresponding transaction price.

10. A market clearing method, characterized in that: A point-to-point transaction method according to any one of claims 1 to 8 is used, wherein the method comprises: Obtaining the local optimization parameters of market participants, including energy demand data, green certificate quota data, and distribution network stability parameters, and obtaining collaborative target constraint conditions; Decomposing the global optimization problem of market clearing into the local sub-problems of each participant according to the local optimization parameters and the collaborative objective constraints, and defining the consensus variables associated with the global objective; Initialize the consensus variable and the corresponding dual variable, establish the consensus alternating direction multiplier method to iteratively update the consensus variable; Each of the market participants solves the local sub-problems in parallel based on the consensus variables to generate updated local variables; Exchanging the local variables according to the communication network and calculating the global consensus variable; According to the deviation between the global consensus variable and the local variable, the dual variable is updated. When the deviation is less than the preset threshold, the market clearing result is output. The market clearing result includes the energy trading price, the green certificate trading matching plan and the injection adjustment instruction of the distribution network node power.

Images