Dynamic pricing method and system for virtual power plant information physical system under communication interruption

Abstract

The invention provides a dynamic pricing method and system for a virtual power plant information physical system under communication interruption, and belongs to the field of virtual power plant information physical system optimizing.The method comprises the steps that for transaction data between a virtual power plant and a load, according to a power supply rule, an optimal pricing strategy considering the interaction relation between the virtual power plants is adopted, and the transaction data between the virtual power plant and the load is obtained; determining a pricing problem of the virtual power plant, and converting the pricing problem of the virtual power plant into a Nikaido-Isoda function to obtain an optimal energy selling price; for point-to-point transaction between the two virtual power plants, selecting an optimal energy selling price of a seller based on an evolutionary game theory and a punishment mechanism; and under the condition of communication interruption, a path dynamic adjustment method is adopted, and interrupted transaction data is retransmitted to a destination by using a new path. According to the invention, reasonable energy selling prices between the virtual power plant and the load and between the two virtual power plants can be given under the condition of communication interruption.

Description

Technical Field

[0001] The present invention belongs to the field of virtual power plant information-physical system optimization, and specifically relates to a dynamic pricing method and system for a virtual power plant information-physical system under communication interruption. Background Art

[0002] In the prior art, renewable energy generation is random. In order to enable renewable energy to smoothly provide energy to the load, a virtual power plant (VPP) consisting of energy storage, renewable energy and distributed energy is usually developed. However, the energy supply of a single VPP is limited, and multiple VPPs can form a "multi-VPP" to provide energy to the load. In recent years, in order to improve the observability and controllability of multi-VPPs, information technology has been widely used to transform multi-VPPs into multi-VPP cyber-physical systems.

[0003] At present, energy transactions in multi-VPP cyber-physical systems include two categories: (1) managers in multi-VPP cyber-physical systems supply energy to loads, that is, multi-VPP cyber-physical system transactions; (2) one VPP supplies energy to another VPP, that is, P2P transactions. When a VPP sells energy to a load or another VPP, it wants to get a high return at a suitable sales price. Therefore, multi-VPP cyber-physical systems must face a series of pricing issues. In the above transactions, information including load demand or sales price is transmitted between buyers and sellers through communication networks. However, due to the strong dependence on pricing and data communication, communication issues cannot be ignored. Therefore, it is urgent to explore a dynamic pricing strategy for virtual power plant cyber-physical systems under communication interruptions, which has very broad application prospects.

[0004] Existing research on dynamic pricing strategies for cyber-physical systems of virtual power plants usually has the following problems:

[0005] 1. Existing pricing methods can be divided into two categories: optimization-based and game-based. Both optimization methods and game-based methods do not consider the interaction between VPPs. In fact, the selling price of each VPP will affect the revenue of other VPPs. Therefore, it is important to determine how to model the interaction relationship and consider it in the pricing strategy. Existing research lacks the design of a VPP cyber-physical system transaction pricing strategy;

[0006] 2. Existing methods all assume that both parties in the game are completely rational, that is, each party never considers the reaction of other parties to his decision and only chooses actions that maximize his benefits. In addition, a completely rational game requires that the parties make no mistakes and determine the optimal solution every time. Therefore, due to such strict requirements, the real-time performance of the game is poor, the game results are not always reasonable, and there is a lack of an effective P2P transaction pricing strategy design;

[0007] 3. The impact of communication equipment failure or incomplete communication data in existing research is considered to be network congestion, which can be solved by defense mechanisms. Once the communication path is directly interrupted or the data is completely lost, how to complete the pricing task in such a serious failure situation needs further research, and there is a lack of research on communication issues in the pricing process. Summary of the invention

[0008] In view of the deficiencies in the prior art, the present application proposes a dynamic pricing method and system for a virtual power plant cyber-physical system under communication interruption.

[0009] In a first aspect, the present application proposes a dynamic pricing method for a virtual power plant cyber-physical system under communication interruption, comprising:

[0010] Determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: a plurality of sellers and a buyer, wherein the seller is a virtual power plant, which is used to provide energy according to an energy selling price, wherein the plurality of virtual power plants all obtain maximum benefits according to the energy selling price, and the plurality of virtual power plants are in a competitive relationship, and the buyer is used to purchase corresponding energy according to the energy selling price;

[0011] In the virtual power plant cyber-physical system, the transaction data between the virtual power plant and the load and the point-to-point transaction data between two virtual power plants are collected;

[0012] Based on the generalized game theory and the power supply rules, the pricing problem of virtual power plants is determined by taking into account the transaction data between virtual power plants and loads. The pricing problem of virtual power plants is converted into the Nikaido-Isoda function to obtain the optimal energy selling price.

[0013] For peer-to-peer transactions between two virtual power plants, the optimal energy selling price of the seller is selected based on evolutionary game theory and penalty mechanism;

[0014] In the event of communication interruption, based on the principle of matching path performance with data importance, a dynamic path adjustment method is adopted to retransmit the interrupted transaction data to the destination using a new path.

[0015] The power supply rule is determined according to the pre-established virtual power plant cyber-physical system, and the calculation formula is as follows:

[0016]

[0017] Among them, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price n is the energy selling price of the nth virtual power plant, and N is the number of virtual power plants.

[0018] The transaction data between the virtual power plant and the load is based on the generalized game theory and the power supply rules, and the optimal pricing strategy considering the interaction between the virtual power plants is adopted to determine the pricing problem of the virtual power plant, and the pricing problem of the virtual power plant is converted into the Nikaido-Isoda function to obtain the optimal energy selling price, including:

[0019] Based on generalized game theory, according to the power supply rules, the pricing problem of virtual power plants is determined by adopting the optimal pricing strategy that takes into account the interaction between virtual power plants. The pricing problem is the game goal of each virtual power plant, wherein the game parties of generalized game theory are each virtual power plant in the virtual power plant cyber-physical system, the strategy of generalized game theory is the energy selling price of each virtual power plant, and the benefit of generalized game theory is the maximized benefit of each virtual power plant;

[0020] The pricing problem of virtual power plants is transformed into the Nikaido-Isoda function, and the Nikaido-Isoda function is solved to obtain the optimal energy selling price.

[0021] Based on the generalized game theory and according to the power supply rules, the pricing problem of virtual power plants is determined by adopting the optimal pricing strategy that considers the interaction between virtual power plants. The calculation formula is as follows:

[0022] maxU n (price n ,price -n ) = maxprice n ×Q n

[0023]

[0024]

[0025] The constraints are:

[0026] E n,n>>|E i,n |,i≠n,E i,n <0

[0027]

[0028] Among them, U n is the revenue of the nth virtual power plant in the transaction between the virtual power plant cyber-physical system and the load, price -n is the energy selling price of the remaining virtual power plants except the nth virtual power plant, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price n is the energy selling price of the nth virtual power plant, N is the number of virtual power plants, E is the interaction coefficient matrix, which represents the interaction relationship between virtual power plants, price nmin is the minimum energy selling price of the nth virtual power plant, price nmax is the maximum energy selling price of the nth virtual power plant, P load is the load power.

[0029] The pricing problem of the virtual power plant is converted into the Nikaido-Isoda function, and the Nikaido-Isoda function is solved to obtain the optimal energy selling price, which is calculated as follows:

[0030] The pricing problem of the virtual power plant is converted into the Nikaido-Isoda function, and the calculation formula is as follows:

[0031]

[0032] Constraints:

[0033] Where k represents the kth iteration, α is a positive constant, and u 1n For U n The elements in , f1 is the equivalent transformation form of the original optimization objective based on the logarithmic barrier method, and f2 is the supply and demand balance equality constraint.

[0034] When solving the Nikaido-Isoda function, the KKT condition (Karush-Kuhn-Tucker condition) for k iterations is:

[0035]

[0036] Among them, d k A vector representing the price change at the kth iteration;

[0037] Judgment condition|price k+1 -price k |≤ξ, where ξ is a constant. If it is satisfied, price k+1 The value of is the optimal energy selling price.

[0038] The point-to-point transaction between two virtual power plants is based on evolutionary game theory and penalty mechanism to select the optimal energy selling price of the seller, including:

[0039] For the point-to-point transaction between two virtual power plants, an evolutionary game model is constructed based on evolutionary game theory and penalty mechanism. The evolutionary game model adopts a set (N, S, P, U b ,U s ) represents, where N = (N b |N s ) is the game space, N b is the buyer set, N s is the seller collection; S = (S b |S s ) is the strategy space, S b is the buyer’s purchasing strategy, S s is the seller’s pricing strategy; P = (p1, p2, q1, q2, q3) is the game probability, where p1 is the probability of choosing the first purchase strategy, p2 is the probability of choosing the second purchase strategy; q1 is the probability of choosing the first purchase price, q2 is the probability of choosing the second purchase price, and q3 is the probability of choosing the third purchase price; is the buyer’s profit, where b ij is the payoff when the buyer chooses strategy i and the seller chooses strategy j; It indicates the penalty mechanism when the buyer chooses other strategies at will. If the buyer follows the normal game results, otherwise is the seller’s profit, where s ij is the profit when the seller chooses strategy j and the buyer chooses strategy i, It is a penalty mechanism for sellers to choose other prices at will. If buyers obey the game, the result will be normal. otherwise

[0040] According to the evolutionary game model, the evolutionary stable state is analyzed to simulate the changing trend of the strategy selection probability of the two virtual power plants;

[0041] In the results of the evolutionary stable state analysis, the original points existing on the boundary or surface of the feasible region correspond to the changes in the selection probability and select the optimal energy selling price of the seller.

[0042] The evolutionary game model is used to analyze the evolutionary stable state and simulate the changing trend of the strategy selection probability of the two virtual power plants, including:

[0043] Calculate the seller's expected profit and average profit when the seller obeys the game result;

[0044] During the game between two virtual power plants, the change of selection probability is described by the seller replication dynamic equation;

[0045] Calculate the expected payment and average payment of the seller when the buyer obeys the game result;

[0046] During the two virtual power plant game, the change of selection probability is described by the buyer's copy dynamic equation;

[0047] The party replication dynamic equation and the buyer replication dynamic equation are jointly solved to obtain the evolutionary stable state analysis result.

[0048] In the case of communication interruption, according to the principle of matching path performance with data importance, a dynamic path adjustment method is adopted to retransmit the interrupted transaction data to the destination using a new path, including:

[0049] According to the partial derivative of the virtual power plant pricing problem on the transaction data, the importance of the transaction data is calculated;

[0050] Determining the conditions of the planned communication path, wherein the conditions of the planned communication path are: matching the importance of the transaction data with the attributes of the channel, wherein the attributes of the channel include: security risk, congestion level, and transmission speed;

[0051] The minimum value of the sum of the attributes of all channels is used as the objective function of the dynamic adjustment method of the path;

[0052] According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load, and a new path is obtained in the case of communication interruption;

[0053] Retransmit the interrupted transaction data to the destination using a new path.

[0054] According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load, and a new path is obtained in the case of communication interruption, including:

[0055] Step S5.4.1: Divide the nodes in the communication network into leader nodes and follower nodes, take the central node of the virtual power plant as the leader node, and take the nodes other than the central node of the virtual power plant as follower nodes, wherein the leader node will not be affected by other nodes;

[0056] Step S5.4.2: In each iteration, a current node is selected from the follower nodes, and the status of all nodes is updated according to the status of the neighbor nodes of the current node using an update protocol based on deviation minimization, where the deviation minimization consists of the delay from other nodes to the leader node without considering queuing delay and the weights between nodes;

[0057] Step S5.4.3: judging whether the communication requirements are met according to the convergence value of the update protocol based on deviation minimization;

[0058] Step S5.4.4: If the communication requirements are met, the parent node of the current node is calculated, and the path composed of recursively connecting the parent nodes is used as the new path in case of communication interruption;

[0059] Step S5.4.5: If the communication requirement is not met, return to step S5.4.2, reselect other nodes from the follower nodes as the current node, and iterate again.

[0060] In a second aspect, the present application proposes a dynamic pricing system for a virtual power plant cyber-physical system under communication interruption, comprising:

[0061] A rule establishment module, used to determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: a plurality of sellers and a buyer, wherein the seller is a virtual power plant, used to provide energy according to an energy selling price, wherein the plurality of virtual power plants all obtain maximum benefits according to the energy selling price, wherein the plurality of virtual power plants are in a competitive relationship, and wherein the buyer is used to purchase corresponding energy according to the energy selling price;

[0062] A data collection module is used to collect transaction data between a virtual power plant and a load and point-to-point transaction data between two virtual power plants in a virtual power plant cyber-physical system;

[0063] The first pricing module is used to determine the pricing problem of the virtual power plant based on the transaction data between the virtual power plant and the load, based on the generalized game theory and the power supply rules, using the optimal pricing strategy that considers the interaction relationship between the virtual power plants, and converting the pricing problem of the virtual power plant into a Nikaido-Isoda function to obtain the optimal energy selling price;

[0064] The second pricing module is used to select the optimal energy selling price of the seller for the peer-to-peer transaction between two virtual power plants based on evolutionary game theory and penalty mechanism;

[0065] The data retransmission module is used to retransmit the interrupted transaction data to the destination using a new path in the event of communication interruption, based on the principle of matching path performance with data importance and adopting a dynamic path adjustment method.

[0066] Beneficial effects:

[0067] This application proposes a dynamic pricing method and system for the information-physical system of a virtual power plant under communication interruption. In the transaction between the power grid and the load, the interactive relationship between the power grid and the load is taken into account in the pricing, which improves the rationality of the pricing. At the same time, the generalized game theory is used to effectively simplify the calculation process of the optimal price. In P2P transactions, the boundary rationality characteristics of buyers and sellers are taken into account in the pricing process. At the same time, a penalty mechanism is designed in the game to enhance the rationality of pricing and ensure the evolutionary stability of the pricing results. Based on the idea of ​​"dynamic adjustment", the communication path is rerouted on the principle of matching the security of the path with the importance of the data to ensure the smooth transmission of data to the destination and the smooth execution of the pricing strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Flow chart of a dynamic pricing method for a virtual power plant cyber-physical system under communication interruption in an embodiment of the present application;

[0069] Figure 2 A schematic diagram of the overall architecture of a dynamic pricing method for a cyber-physical system of a virtual power plant under communication interruption according to an embodiment of the present application;

[0070] Figure 3 Schematic diagram of the overall pricing and power regulation process of the VPP cyber-physical system in an embodiment of the present application;

[0071] Figure 4 Schematic diagram of the communication topology used in the VPP cyber-physical system load transaction in the embodiment of the present application;

[0072] Figure 5 Schematic diagram of communication topology in P2P transaction in an embodiment of the present application;

[0073] Figure 6 A schematic diagram of a game tree according to an embodiment of the present application;

[0074] Figure 7 A schematic diagram of the execution process of the dynamic adjustment algorithm of an embodiment of the present application;

[0075] Figure 8 Schematic diagram of the dynamic pricing system of the cyber-physical system of a virtual power plant under communication interruption in an embodiment of the present application. DETAILED DESCRIPTION

[0076] The specific implementation of the present application is further described in detail below in conjunction with the drawings and examples.

[0077] In order to determine the reasonable pricing of the virtual power plant cluster, a cyber-physical system (CPS) was established using data communication technology to improve the observability of the virtual power plant. Then, the following work was performed in the CPS:

[0078] Physical layer:

[0079] (1) Based on generalized game theory, an optimal pricing strategy considering the interaction between virtual power plants is proposed for the transactions between virtual power plants and loads. The pricing problem is transformed into a Nikaido-Isoda function, which can easily obtain the optimal price.

[0080] (2) For peer-to-peer transactions between two virtual power plants, the optimal selling price of the seller is selected based on boundary rationality based on evolutionary game theory and penalty mechanism.

[0081] Network layer:

[0082] Aiming at the communication interruption problem in pricing (i.e., the game process), a dynamic path adjustment method is proposed based on the principle of matching path performance with data importance, that is, using a new path to retransmit the interrupted data to the destination.

[0083] Embodiment 1:

[0084] This embodiment proposes a dynamic pricing method for a virtual power plant cyber-physical system under communication interruption, such as Figure 1 , Figure 2 As shown, including:

[0085] Step S1: Determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: multiple sellers and one buyer, wherein the seller is a virtual power plant, which is used to provide energy according to an energy selling price, wherein the multiple virtual power plants all obtain maximum profits according to the energy selling price, and the multiple virtual power plants are in a competitive relationship, and the buyer is used to purchase corresponding energy according to the energy selling price;

[0086] In this embodiment, three problems are studied: "pricing problem in VPP cyber-physical system transactions", "pricing problem in P2P transactions" and "communication uncertainty problem". Among them, P2P transactions and VPP cyber-physical system transactions are carried out sequentially. VPP cyber-physical system transactions are carried out only after the P2P transaction between two VPPs is completed; reliable communication is the prerequisite for completing these two transactions, so the communication uncertainty problem needs to be solved during the transaction process. The structure studied is as follows Figure 3As shown, it includes the pricing part in VPP cyber-physical system transactions, the pricing part in P2P transactions, and the dynamic adjustment part of the communication path. The game algorithms are all included in the game center, which can support the operating environment of these algorithms.

[0087] In the VPP cyber-physical system, bandwidth and latency issues are considered, where the bandwidth is 100Mbps, the transmission speed is 12MB / s, and the allowed latency is 50ms. Since VPP transactions are many-to-one transactions, information from all participants is required, and pricing instructions are generated at the same time to facilitate supervision of all participants and ensure consistency. Therefore, it is appropriate to use centralized communication. The relevant topology is as follows Figure 4 The inputs include the selling price and load demand of the VPP; the outputs are the seller’s optimal selling price and the buyer’s actions.

[0088] The VPP cyber-physical system transaction involves multiple sellers and one buyer. Multiple VPPs strive to maximize their revenue and are in a competitive relationship. In addition, the revenue of each VPP is affected by the prices of other VPPs, and there are constraints between them. In summary, this game is similar to a generalized game. Considering that there are many VPPs that need to be priced, using the generalized game model, these VPPs need to follow a cooperation agreement: "When the price of a VPP is selected, the nth VPP is based on Bottom power supply.

[0089]

[0090] Among them, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price n is the energy selling price of the nth virtual power plant, and N is the number of virtual power plants.

[0091] In this case, since the price of a VPP is inversely proportional to the amount of energy it sells, all VPPs will not easily change their prices to avoid a decline in revenue.

[0092] Step S2: In the virtual power plant cyber-physical system, the transaction data between the virtual power plant and the load and the point-to-point transaction data between two virtual power plants are collected;

[0093] Step S3: Based on the generalized game theory and the power supply rules, the pricing problem of the virtual power plant is determined by adopting the optimal pricing strategy that considers the interaction between the virtual power plants for the transaction data between the virtual power plants and the loads. The pricing problem of the virtual power plant is converted into the Nikaido-Isoda function to obtain the optimal energy selling price, including:

[0094] Based on generalized game theory, according to the power supply rules, the pricing problem of virtual power plants is determined by adopting the optimal pricing strategy that takes into account the interaction between virtual power plants. The pricing problem is the game goal of each virtual power plant, wherein the game parties of generalized game theory are each virtual power plant in the virtual power plant cyber-physical system, the strategy of generalized game theory is the energy selling price of each virtual power plant, and the benefit of generalized game theory is the maximized benefit of each virtual power plant;

[0095] At the physical level, for transactions between VPPs and loads, based on generalized game theory, this embodiment proposes an optimal pricing strategy that takes into account the interactive relationship between VPPs, converts the pricing problem into a Nikaido-Isoda function, and conveniently obtains the optimal energy selling price.

[0096] A. Gaming goals of each VPP:

[0097] In order to solve the pricing problem in multi-VPP transactions and maximize the benefits of each VPP, the game objective of the nth VPP is set as:

[0098] maxU n (price n ,price -n ) = maxprice n ×Q n (2)

[0099]

[0100]

[0101] Among them, U n is the revenue of the nth virtual power plant in the transaction between the virtual power plant cyber-physical system and the load, price -n is the energy selling price of the remaining virtual power plants except the nth virtual power plant, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price nis the energy selling price of the nth virtual power plant, N is the number of virtual power plants, and E is the interaction coefficient matrix, which represents the interaction relationship between virtual power plants. The influence of each VPP itself is dominant, so these coefficients need to satisfy:

[0102] E n,n >>|E i,n |,i≠n,E i,n <0 (5)

[0103] In the game, there should be:

[0104]

[0105] Price nmin is the minimum energy selling price of the nth virtual power plant, price nmax is the maximum energy selling price of the nth virtual power plant, P load is the load power.

[0106] B. The game model has three elements:

[0107] (1) Players: each VPP in the VPP cyber-physical system; (2) Strategy: the selling price of each VPP; (3) Profit U n (price n ,price -n ).when Vector It is called the generalized Nash equilibrium solution; E ij By E ij =(ΔP ricej / P ricej ) / (ΔP ricei / P ricei ) is calculated. In the game, pricing can be regarded as a complete information game, and the matrix E is a known quantity.

[0108] C.Solution to the game problem:

[0109] In order to find the optimal price * , by constructing the Nikaido-Isoda function, the generalized game problem is transformed into a convex optimization problem:

[0110]

[0111] Where Φ(price) is the energy selling price of the nth virtual power plant, which is changed from Change to When , the increased benefit, k represents the kth iteration, and α is a positive constant. Correspondingly, the objective function is:

[0112]

[0113] Among them, α is a positive constant. Therefore, we should have:

[0114]

[0115] When considering the constraints in (6), based on the logarithmic barrier method, the solution of problem (8) is transformed into finding the solution of the problem

[0116]

[0117] Among them, u 1n For U n The relevant KKT conditions are: Where v is the auxiliary coefficient and f1 is the original optimization target is an equivalent transformation form based on the logarithmic barrier method, and f2 is the supply and demand balance equation constraint.

[0118] For convex optimization problems, the Hessian matrix is ​​negative definite, and the optimal goal is to find the maximum value. Therefore, the application conditions of Newton's method can be met. Accordingly, the following formula needs to be calculated:

[0119] minf1(price k +d k )stf2(price k +d k )=0 (11)

[0120] According to Taylor's formula, the KKT condition for the kth iteration is:

[0121]

[0122] According to formula (12), d k can be obtained, k A vector representing the price change at the kth iteration, price k+1 =price + +d k can also be obtained. At the same time, the judgment condition|price k+1 -price k |≤ξ is satisfied, ξ is a constant. If satisfied, price k+1 The value of is the optimal solution. If not, set k = k + 1 and return to convert the solution to equation (12).

[0123] For VPP cyber-physical systems, when a VPP is added or removed from the system, only the output value Q needs to be updated. n And the coefficient matrix E.

[0124] Step S4: For the point-to-point transaction between two virtual power plants, based on the evolutionary game theory and penalty mechanism, the optimal energy selling price of the seller is selected, including:

[0125] Step S4.1: for the point-to-point transaction between two virtual power plants, an evolutionary game model is constructed based on evolutionary game theory and penalty mechanism;

[0126] Step S4.2: According to the evolutionary game model, the evolutionary stable state is analyzed to simulate the changing trend of the strategy selection probability of the two virtual power plants, including:

[0127] Step S4.2.1: Calculate the expected profit and average profit of the seller when the seller obeys the game result;

[0128] Step S4.2.2: During the game between two virtual power plants, the change of selection probability is described by the seller replication dynamic equation;

[0129] Step S4.2.3: Calculate the expected payment and average payment of the seller when the buyer obeys the game result;

[0130] Step S4.2.4: During the game between two virtual power plants, the change of selection probability is described by the buyer replication dynamic equation;

[0131] Step S4.2.5: Solve the party replication dynamic equation and the buyer replication dynamic equation together to obtain the evolutionary stable state analysis result.

[0132] Step S4.3: In the results of the evolutionary stable state analysis, the original points existing on the boundary or surface of the feasible region correspond to the change in selection probability, and the optimal energy selling price of the seller is selected.

[0133] In this embodiment, at the physical level, for a point-to-point transaction between two VPPs, the optimal selling price of the seller is selected based on boundary rationality based on evolutionary game theory and penalty mechanism.

[0134] In P2P transactions, both parties have boundaries and can change their game strategies instead of completing them all at once, showing the characteristics of dynamic games. In short, this type of game is similar to evolutionary games. Figure 5 In order to solve the pricing problem in P2P transactions and maximize the interests of both parties, an evolutionary game model is first constructed; then the evolutionary stable state is analyzed to simulate the changing trend of the strategy selection probability of the two VPPs; finally, the optimal pricing strategy is selected for the seller.

[0135] A. Building an Evolutionary Game Model

[0136] The constructed game tree is as follows Figure 6 As shown, firstly, relevant definitions are given to construct the game model.

[0137] The evolutionary game model can be represented by a set (N, S, P, U b ,U s ) is used to represent the parameters. The meanings of the parameters are as follows: 1) N = (N b |N s ) is the game space, where N b is the buyer set; N s is the seller set; 2) S = (S b |S s ) is the strategy space, where S b Indicates the buyer's purchasing strategy; S s represents the seller’s pricing strategy; 3) P = (p1, p2, q1, q2, q3) represents the game probability, where p1 and p2 represent the probability of choosing two purchasing strategies; q1, q2, q3 represent the probability of choosing three prices respectively; 4) represents the buyer's gain, where b ij represents the profit when the buyer chooses strategy i and the seller chooses strategy j; It indicates the penalty mechanism when the buyer chooses other strategies at will; if the buyer follows the normal game results, otherwise represents the seller’s revenue, where s ij represents the profit when the seller chooses strategy j and the buyer chooses strategy i; It indicates the penalty mechanism when the seller chooses other prices at will; if the buyer obeys the game result, otherwise

[0138] B. Evolutionary Stable State Analysis

[0139] (1) Calculation of dynamic equations copied by the seller

[0140] Under normal circumstances, the seller obeys the game results, expected returns and average returns for:

[0141]

[0142]

[0143]

[0144]

[0145] It should be noted that q3=1-q1-q2, and the replication kinetic equation of q3 is omitted in the present invention.

[0146] (2) Calculation of the buyer's copy dynamic equation

[0147] Similarly, the buyer's expected payment U b1 , U b2 and average payment Should first be calculated as:

[0148]

[0149]

[0150] because therefore The calculation is as follows

[0151]

[0152] (3) Joint solution

[0153] If we solve (14) = 0 and (16) = 0 simultaneously, we can get several sets of solutions (p1, q1, q2). In these solutions, some terms like or The probability that the buyer or seller chooses to buy or sell energy is 1 or 0 respectively; other solutions indicate that the buyer and seller choose to buy or sell energy with different probabilities. For the original points (p1, q1, q2) that exist on the boundary or surface of the feasible region, the change of the selection probability is directly analyzed.

[0154] Step S5: In the case of communication interruption, according to the principle of matching path performance with data importance, a dynamic path adjustment method is adopted to retransmit the interrupted transaction data to the destination using a new path, including:

[0155] Step S5.1: Calculate the importance of the transaction data based on the partial derivative of the virtual power plant pricing problem with respect to the transaction data;

[0156] Step S5.2: determining the conditions of the planned communication path, wherein the conditions of the planned communication path are: matching the importance of the transaction data with the attributes of the channel, wherein the attributes of the channel include: security risk, congestion level, and transmission speed;

[0157] Step S5.3: taking the minimum value of the sum of the attributes of all channels as the objective function of the dynamic adjustment method of the path;

[0158] Step S5.4: According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load, and a new path is obtained in the case of communication interruption, including:

[0159] Step S5.4.1: Divide the nodes in the communication network into leader nodes and follower nodes, take the central node of the virtual power plant as the leader node, and take the nodes other than the central node of the virtual power plant as follower nodes, wherein the leader node will not be affected by other nodes;

[0160] Step S5.4.2: In each iteration, a current node is selected from the follower nodes, and the status of all nodes is updated according to the status of the neighbor nodes of the current node using an update protocol based on deviation minimization, where the deviation minimization consists of the delay from other nodes to the leader node without considering queuing delay and the weights between nodes;

[0161] Step S5.4.3: judging whether the communication requirements are met according to the convergence value of the update protocol based on deviation minimization;

[0162] Step S5.4.4: If the communication requirements are met, the parent node of the current node is calculated, and the path composed of recursively connecting the parent nodes is used as the new path in case of communication interruption;

[0163] Step S5.4.5: If the communication requirement is not met, return to step S5.4.2, reselect other nodes from the follower nodes as the current node, and iterate again.

[0164] Step S5.5: Retransmit the interrupted transaction data to the destination using the new path.

[0165] In this embodiment, at the network layer, in response to the communication interruption problem in pricing (i.e., the game process), a dynamic path adjustment method is proposed based on the principle of matching path performance with data importance, and a new path is used to retransmit the interrupted data to the destination.

[0166] In order to solve the problem of communication interruption during the transaction process and improve the reliability of centralized communication, a dynamic adjustment algorithm is proposed to reconstruct the path and retransmit the data to the destination through the reconstructed path. In this method, "dynamic" means that under the interference of communication interruption, the communication path will change passively and show dynamic characteristics. The core ideas of this algorithm are "determining the importance of data", "constructing the objective function" and "dynamic adjustment of the path". The adjustment process is as follows Figure 7 shown.

[0167] exist Figure 7 In, there are N VPPs, communication channels, where A represents an array. For example, in a transaction between a VPP and a load, once the communication path from the VPP to the gaming center is interrupted, the transmission of the original price and elasticity coefficient will be interrupted. The dynamic adjustment process is as follows: First, the current communication state of the network is passed to the dynamic adjustment component to complete state identification and path planning; the generated planning instructions are sent to VPP1 to complete the path reconstruction, that is, the data is directly transmitted from VPP 1 to VPP 3, and then VPP 3 transmits the data from VPP 1 back to the gaming center; finally, based on the returned data, the gaming center will perform the pricing task. The specific path planning algorithm is divided into the following steps.

[0168] Step 5.1: Data importance assessment

[0169] According to the partial derivative of the virtual power plant pricing problem with respect to the transaction data, the importance of the transaction data is calculated:

[0170]

[0171] The relevant equation is as follows:

[0172]

[0173]

[0174] In addition, the importance of the data can be calculated as For P2P transactions, according to equations (13) to (17), equation (14) = 0, equation (16) = 0, and the relative importance of the data can also be obtained. In the design of the present invention, for the transmission of important data, the path with better performance is preferentially selected.

[0175] Step 5.2: Determine the conditions of the planned communication path, wherein the conditions of the planned communication path are: the importance of the transaction data matches the attributes of the channel;

[0176] The communication path should be planned so that the importance of the data matches the properties of the channel.

[0177] Step S5.3: taking the minimum value of the sum of the attributes of all channels as the objective function of the dynamic adjustment method of the path;

[0178] Taking the minimum value as the optimal goal, we get the following design. If we want to add the attributes of all channels together, we have:

[0179]

[0180] Where N is the number of VPPs; min means minimum; Att ij,kis the kth attribute of the channel from VPP i to VPP j, including security risk, congestion level, and transmission speed. ij Indicates the link between the i-th VPP and the j-th VPP. If connected, it is equal to 1, otherwise it is equal to 0.

[0181] If you only need the maximum value of the attribute to satisfy the condition:

[0182] Z=min{max[Att ij.k ×sign(a ij )]},j∈N i (twenty one)

[0183] Step S5.4: According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load to obtain a new path in the event of communication interruption.

[0184] Step S5.4.1: Divide the nodes in the communication network into leader nodes and follower nodes, take the central node of the virtual power plant as the leader node, and take the nodes other than the central node of the virtual power plant as follower nodes, wherein the leader node will not be affected by other nodes;

[0185] Step S5.4.2: In each iteration, a current node is selected from the follower nodes, and the status of all nodes is updated according to the status of the neighbor nodes of the current node using an update protocol based on deviation minimization, where the deviation minimization consists of the delay from other nodes to the leader node without considering queuing delay and the weights between nodes;

[0186] Step S5.4.3: judging whether the communication requirements are met according to the convergence value of the update protocol based on deviation minimization;

[0187] Step S5.4.4: If the communication requirements are met, the parent node of the current node is calculated, and the path composed of recursively connecting the parent nodes is used as the new path in case of communication interruption;

[0188] Step S5.4.5: If the communication requirement is not met, return to step S5.4.2, reselect other nodes from the follower nodes as the current node, and iterate again.

[0189] This embodiment proposes a distributed deviation minimization consensus algorithm for path planning between the VPP center and the load. The nodes in the communication network are divided into two types: leader nodes i∈S1 and follower nodes i∈S2. The VPP center node is regarded as the leader node. Its state remains unchanged and will not be affected by other nodes. Other nodes are determined as follower nodes. In each iteration, their states will be updated according to the conditions of their neighbors. The update protocol is:

[0190]

[0191] Among them, y i (k) represents the delay from node i to the VPP center node without considering the queuing delay. ij is a bias term, which represents the weight between node i and node j. i is the neighbor node directly connected to node i.

[0192] When the algorithm converges, y i The value will reach

[0193]

[0194] Then, by recursively connecting the parent nodes, the optimal path between the VPP center and the load is found. For node i, its parent node P i for:

[0195]

[0196] The proposed distributed deviation minimization consensus algorithm is used to benchmark the path planning solution between VPP centers and loads.

[0197] Step S5.5: Retransmit the interrupted transaction data to the destination using the new path.

[0198] Simulation results:

[0199] This example uses a VPP cyber-physical system containing three VPPs as the experimental object and conducts relevant simulation tests. The scales of the VPPs are 95kWh, 100kWh, and 105kWh respectively; the initial selling prices are 200$ / (MWh), 180$ / (MWh), and 160$ / (MWh); the load demand is 150(MWh). The allowable range of selling prices is "0-700$ / (MWh)", where the elasticity coefficient matrix is:

[0200]

[0201] The simulation results are shown in the table:

[0202] Table 1. P2P pricing iteration results of three VPPs

[0203] Iterations VPP1 Price VPP2 Price VPP3 Price Iteration 1 502.5765 505.0883 507.484 Iteration 2 491.3366 505.1858 518.6136 Iteration 3 485.7165 505.23456 524.1813 Iteration 4 482.9065 505.2589 526.9642 Iteration 5 481.5096 505.2792 528.3638 Iteration 6 480.8071 505.2853 529.0595

[0204] Table 2. Three VPP and load pricing iteration results

[0205] Iterations VPP1 Price VPP2 Price VPP3 Price Iteration 1 552.5743 555.0123 547.3574 Iteration 2 505.1264 502.1221 509.9565 Iteration 3 492.6767 490.4340 495.0129 Iteration 4 480.9978 480.5812 484.3394 Iteration 5 471.9000 472.4549 470.6508 Iteration 6 471.9001 472.4549 470.6500

[0206] Tables 1 and 2 show the dynamic pricing results of the virtual power plant (VPP) cyber-physical system in the case of communication interruption. By comparing the simulation results in the two scenarios, it can be observed that the pricing method proposed in this embodiment shows obvious superiority in optimizing the benefits of VPP and responding to user needs.

[0207] Table 1 shows the pricing iteration results of the three VPPs in the case of peer-to-peer (P2P) transactions. As the number of pricing iterations increases, the pricing of VPP1, VPP2, and VPP3 gradually stabilizes, and the pricing fluctuation decreases. This shows that the proposed dynamic pricing method can effectively cope with the complex competition environment under communication interruption, and gradually find a balance point, so that each VPP can achieve relatively stable benefits in the competition.

[0208] Table 2 shows the pricing selection costs of three VPPs and loads in trading situations. Compared with the P2P trading in Table 1, the pricing changes in the load trading situation are more significant, especially in the initial few cost selection processes, where the pricing fluctuates greatly. This further illustrates the ability of the proposed method to dynamically adjust pricing to adapt to different trading environments in the event of communication interruption.

[0209] In summary, the proposed dynamic pricing method can effectively maintain the trading balance between virtual power plants and between virtual power plants and loads in the event of communication interruption, avoiding the pricing failure problem caused by communication interruption in traditional methods, and ultimately achieving stability and maximum profit in a competitive environment, fully demonstrating the practical application value of this technology.

[0210] The dynamic pricing method of the information-physical system of the virtual power plant under communication interruption proposed in this embodiment includes: first, in order to determine the reasonable pricing of the VPP cluster, a information-physical system (CPS) is established using data communication technology to improve the observability of the VPP. Then, the following work is performed in the CPS. In terms of the physical layer: (1) For transactions between VPPs and loads, based on generalized game theory, an optimal pricing strategy that considers the interaction relationship between VPPs is proposed, and the pricing problem is converted into a Nikaido-Isoda function to conveniently obtain the optimal price; (2) For point-to-point transactions between two VPPs, based on evolutionary game theory and penalty mechanism, the optimal selling price of the seller is selected based on boundary rationality. In terms of the network layer: In view of the communication interruption problem (i.e., the game process) in pricing, based on the principle of matching path performance with data importance, a dynamic adjustment method of the path is proposed, that is, a new path is used to retransmit the interrupted data to the destination.

[0211] Embodiment 2:

[0212] This embodiment proposes a dynamic pricing system for a virtual power plant cyber-physical system under communication interruption, such as Figure 8 As shown, it includes: a rule establishment module, a data collection module, a first pricing module, a second pricing module and a data retransmission module, the rule establishment module is connected to the data collection module, the data collection module is respectively connected to the first pricing module and the second pricing module, the first pricing module and the second pricing module are respectively connected to the data retransmission module;

[0213] A rule establishment module, used to determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: a plurality of sellers and a buyer, wherein the seller is a virtual power plant, used to provide energy according to an energy selling price, wherein the plurality of virtual power plants all obtain maximum benefits according to the energy selling price, wherein the plurality of virtual power plants are in a competitive relationship, and wherein the buyer is used to purchase corresponding energy according to the energy selling price;

[0214] A data collection module is used to collect transaction data between a virtual power plant and a load and point-to-point transaction data between two virtual power plants in a virtual power plant cyber-physical system;

[0215] The first pricing module is used to determine the pricing problem of the virtual power plant based on the transaction data between the virtual power plant and the load, based on the generalized game theory and the power supply rules, using the optimal pricing strategy that considers the interaction relationship between the virtual power plants, and converting the pricing problem of the virtual power plant into a Nikaido-Isoda function to obtain the optimal energy selling price;

[0216] The second pricing module is used to select the optimal energy selling price of the seller for the peer-to-peer transaction between two virtual power plants based on evolutionary game theory and penalty mechanism;

[0217] The data retransmission module is used to retransmit the interrupted transaction data to the destination using a new path in the event of communication interruption, based on the principle of matching path performance with data importance and adopting a dynamic path adjustment method.

[0218] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0219] The protection scope of the present application is not limited to the above-mentioned embodiments. Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the scope and spirit of the present disclosure. If these changes and modifications fall within the scope of the claims of the present disclosure and their equivalents, the intention of the present disclosure also includes these changes and modifications.

Claims

1. A dynamic pricing method for a virtual power plant cyber-physical system under communication interruption, characterized in that: include: Determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: a plurality of sellers and a buyer, wherein the seller is a virtual power plant, which is used to provide energy according to an energy selling price, wherein the plurality of virtual power plants all obtain maximum benefits according to the energy selling price, and the plurality of virtual power plants are in a competitive relationship, and the buyer is used to purchase corresponding energy according to the energy selling price; In the virtual power plant cyber-physical system, the transaction data between the virtual power plant and the load and the point-to-point transaction data between two virtual power plants are collected; Based on the generalized game theory and the power supply rules, the pricing problem of virtual power plants is determined by taking into account the transaction data between virtual power plants and loads. The pricing problem of virtual power plants is converted into the Nikaido-Isoda function to obtain the optimal energy selling price. For peer-to-peer transactions between two virtual power plants, the optimal energy selling price of the seller is selected based on evolutionary game theory and penalty mechanism; In the event of communication interruption, based on the principle of matching path performance with data importance, a dynamic path adjustment method is adopted to retransmit the interrupted transaction data to the destination using a new path.

2. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 1 is characterized in that: The power supply rule is determined according to the pre-established virtual power plant cyber-physical system, and the calculation formula is as follows: Among them, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price n is the energy selling price of the nth virtual power plant, and N is the number of virtual power plants.

3. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 1 is characterized in that: The transaction data between the virtual power plant and the load is based on the generalized game theory and the power supply rules, and the optimal pricing strategy considering the interaction between the virtual power plants is adopted to determine the pricing problem of the virtual power plant, and the pricing problem of the virtual power plant is converted into the Nikaido-Isoda function to obtain the optimal energy selling price, including: Based on generalized game theory, according to the power supply rules, the pricing problem of virtual power plants is determined by adopting the optimal pricing strategy that takes into account the interaction between virtual power plants. The pricing problem is the game goal of each virtual power plant, wherein the game parties of generalized game theory are each virtual power plant in the virtual power plant cyber-physical system, the strategy of generalized game theory is the energy selling price of each virtual power plant, and the benefit of generalized game theory is the maximized benefit of each virtual power plant; The pricing problem of virtual power plants is transformed into the Nikaido-Isoda function, and the Nikaido-Isoda function is solved to obtain the optimal energy selling price.

4. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 3 is characterized in that: Based on the generalized game theory and according to the power supply rules, the pricing problem of virtual power plants is determined by adopting the optimal pricing strategy that considers the interaction between virtual power plants. The calculation formula is as follows: maxU n (stories n stories -n )=maxprice n ×Q n The constraints are: AND n ,n>>|And i,n |,i≠n,E i,n <0 price nmin ≤price n ≤price nmax ,n=1,2,...,N; Among them, U n is the revenue of the nth virtual power plant in the transaction between the virtual power plant cyber-physical system and the load, price -n is the energy selling price of the remaining virtual power plants except the nth virtual power plant, Q n is the load demand of the nth virtual power plant, Q n.max is the maximum load demand of the nth virtual power plant, E i,n is the interaction coefficient between the energy sold by the nth virtual power plant and the sales price of the ith virtual power plant, price n is the energy selling price of the nth virtual power plant, N is the number of virtual power plants, E is the interaction coefficient matrix, which represents the interaction relationship between virtual power plants, price nmin is the minimum energy selling price of the nth virtual power plant, price nmax is the maximum energy selling price of the nth virtual power plant, P load is the load power.

5. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 3 is characterized in that: The pricing problem of the virtual power plant is converted into the Nikaido-Isoda function, and the Nikaido-Isoda function is solved to obtain the optimal energy selling price, which is calculated as follows: The pricing problem of the virtual power plant is converted into the Nikaido-Isoda function, and the calculation formula is as follows: Constraints: Where k represents the kth iteration, α is a positive constant, and u 1n For U n The element f1 in is the equivalent transformation form of the original optimization objective based on the logarithmic barrier method, and f2 is the supply-demand balance equation constraint; When solving the Nikaido-Isoda function, the KKT condition for k iterations is: Among them, d k A vector representing the price change at the kth iteration; Judgment condition|price k+1 -price k |≤ξ, where ξ is a constant. If it is satisfied, price k+1 The value of is the optimal energy selling price.

6. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 1 is characterized in that: The point-to-point transaction between two virtual power plants is based on evolutionary game theory and penalty mechanism to select the optimal energy selling price of the seller, including: For the point-to-point transaction between two virtual power plants, an evolutionary game model is constructed based on evolutionary game theory and penalty mechanism. The evolutionary game model adopts a set (N, S, P, U b ,U s ) represents, where N = (N b |N s ) is the game space, N b is the buyer set, N s is the seller collection; S = (S b |S s ) is the strategy space, S b is the buyer’s purchasing strategy, S s is the seller’s pricing strategy; P = (p1, p2, q1, q2, q3) is the game probability, where p1 is the probability of choosing the first purchase strategy, p2 is the probability of choosing the second purchase strategy; q1 is the probability of choosing the first purchase price, q2 is the probability of choosing the second purchase price, and q3 is the probability of choosing the third purchase price; is the buyer’s profit, where b ij is the payoff when the buyer chooses strategy i and the seller chooses strategy j; It indicates the penalty mechanism when the buyer chooses other strategies at will. If the buyer follows the normal game results, otherwise is the seller’s profit, where s ij is the profit when the seller chooses strategy j and the buyer chooses strategy i, It is a penalty mechanism for sellers to choose other prices at will. If buyers obey the game, the result will be normal. otherwise According to the evolutionary game model, the evolutionary stable state is analyzed to simulate the changing trend of the strategy selection probability of the two virtual power plants; In the results of the evolutionary stable state analysis, the original points existing on the boundary or surface of the feasible region correspond to the changes in the selection probability and select the optimal energy selling price of the seller.

7. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 5 is characterized in that: The evolutionary game model is used to analyze the evolutionary stable state and simulate the changing trend of the strategy selection probability of the two virtual power plants, including: Calculate the seller's expected profit and average profit when the seller obeys the game result; During the game between two virtual power plants, the change of selection probability is described by the seller replication dynamic equation; Calculate the expected payment and average payment of the seller when the buyer obeys the game result; During the two virtual power plant game, the change of selection probability is described by the buyer's copy dynamic equation; The party replication dynamic equation and the buyer replication dynamic equation are jointly solved to obtain the evolutionary stable state analysis result.

8. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 5 is characterized in that: In the case of communication interruption, according to the principle of matching path performance with data importance, a dynamic path adjustment method is adopted to retransmit the interrupted transaction data to the destination using a new path, including: According to the partial derivative of the virtual power plant pricing problem on the transaction data, the importance of the transaction data is calculated; Determining the conditions of the planned communication path, wherein the conditions of the planned communication path are: matching the importance of the transaction data with the attributes of the channel, wherein the attributes of the channel include: security risk, congestion level, and transmission speed; The minimum value of the sum of the attributes of all channels is used as the objective function of the dynamic adjustment method of the path; According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load, and a new path is obtained in the case of communication interruption; Retransmit the interrupted transaction data to the destination using a new path.

9. The dynamic pricing method for a virtual power plant cyber-physical system under communication interruption according to claim 5, characterized in that: According to the objective function of the dynamic adjustment method of the path, a distributed deviation minimization consensus algorithm is used to plan the path between the central node of the virtual power plant and the load, and a new path is obtained in the case of communication interruption, including: Step S5.4.1: Divide the nodes in the communication network into leader nodes and follower nodes, take the central node of the virtual power plant as the leader node, and take the nodes other than the central node of the virtual power plant as follower nodes, wherein the leader node will not be affected by other nodes; Step S5.4.2: In each iteration, a current node is selected from the follower nodes, and the status of all nodes is updated according to the status of the neighbor nodes of the current node using an update protocol based on deviation minimization, where the deviation minimization consists of the delay from other nodes to the leader node without considering queuing delay and the weights between nodes; Step S5.4.3: judging whether the communication requirements are met according to the convergence value of the update protocol based on deviation minimization; Step S5.4.4: If the communication requirements are met, the parent node of the current node is calculated, and the path composed of recursively connecting the parent nodes is used as the new path in case of communication interruption; Step S5.4.5: If the communication requirement is not met, return to step S5.4.2, reselect other nodes from the follower nodes as the current node, and iterate again.

10. A dynamic pricing system for a virtual power plant cyber-physical system under communication interruption, characterized in that: include: A rule establishment module, used to determine a power supply rule according to a pre-established virtual power plant cyber-physical system, wherein the pre-established virtual power plant cyber-physical system includes: a plurality of sellers and a buyer, wherein the seller is a virtual power plant, used to provide energy according to an energy selling price, wherein the plurality of virtual power plants all obtain maximum benefits according to the energy selling price, wherein the plurality of virtual power plants are in a competitive relationship, and wherein the buyer is used to purchase corresponding energy according to the energy selling price; A data collection module is used to collect transaction data between a virtual power plant and a load and point-to-point transaction data between two virtual power plants in a virtual power plant cyber-physical system; The first pricing module is used to determine the pricing problem of the virtual power plant based on the transaction data between the virtual power plant and the load, based on the generalized game theory and the power supply rules, using the optimal pricing strategy that considers the interaction relationship between the virtual power plants, and converting the pricing problem of the virtual power plant into a Nikaido-Isoda function to obtain the optimal energy selling price; The second pricing module is used to select the optimal energy selling price of the seller for the peer-to-peer transaction between two virtual power plants based on evolutionary game theory and penalty mechanism; The data retransmission module is used to retransmit the interrupted transaction data to the destination using a new path in the event of communication interruption, based on the principle of matching path performance with data importance and adopting a dynamic path adjustment method.

Images