A privacy-preserving distributed scheduling method and system for a dual-mode microgrid

Abstract

This invention provides a privacy-preserving distributed scheduling method and system for a dual-mode microgrid, belonging to the field of smart grid scheduling and privacy protection technology. The method includes: S1, establishing a distributed scheduling algorithm model for the dual-mode microgrid; S2, according to the operating mode instruction, at the initial moment of the distributed scheduling algorithm model's execution, injecting a set of disturbance values ​​satisfying zero-sum constraints into the state variables of each participating generator node; S3, each generator node, based on its disturbed state variables, executes the distributed scheduling algorithm model, exchanging and calculating state variables with neighboring nodes until the algorithm converges, obtaining the optimal scheduling scheme. This invention effectively protects the initial sensitive information of each generator node by injecting disturbances satisfying zero-sum constraints at the algorithm's initial moment. Because the disturbances satisfy zero-sum constraints, they cancel each other out in subsequent iterations, thus providing privacy protection while ensuring the algorithm's convergence accuracy and global optimality.

Description

Technical Field

[0001] This invention belongs to the field of smart grid scheduling and privacy protection technology, and specifically refers to a distributed scheduling method and system with privacy protection for a dual-mode microgrid. Background Technology

[0002] As an important carrier for integrating distributed renewable energy, microgrids mainly operate in two modes: grid-connected mode (where energy is exchanged between the main grid and the microgrid) and islanded mode (where the microgrid operates independently). In actual operation, microgrids need to dynamically and smoothly switch between these two modes based on the status of the main grid or dispatch instructions through devices such as energy routers to ensure the reliability and resilience of power supply. This dynamic switching capability is a major advantage of microgrids compared to traditional power grids.

[0003] Distributed scheduling algorithms achieve global optimization through local communication between agents (generator nodes) corresponding to distributed generators, and are a key technology supporting the efficient operation of microgrids. However, during the iterative process, distributed scheduling algorithms require each generator node to frequently exchange its internal state information (such as incremental cost, local power mismatch, etc.). This information is highly correlated with the nodes' private commercial data (such as generation cost parameters, real-time generation plans). Exchanging information in an open communication network poses significant privacy risks. For example, external eavesdroppers may intercept communication data and infer sensitive information of each participant for unfair competition or to launch cyberattacks. Even honest but curious legitimate nodes (i.e., participants who abide by the algorithm rules but attempt to spy on their neighbors' privacy) may use the received intermediate calculation results to deduce the private data of other nodes. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a distributed scheduling method and system with privacy protection for dual-mode microgrids, the technical solution of which is as follows: On the one hand, a privacy-preserving distributed scheduling method for a dual-mode microgrid is provided, the method comprising: S1. Establish a distributed scheduling algorithm model for a dual-mode microgrid; S2. According to the operation mode instruction, at the initial moment when the distributed scheduling algorithm model starts execution, inject a set of disturbance values ​​that satisfy the zero-sum constraint into the state variables of each power generation node participating in the scheduling. S3. Each power generation node executes the distributed scheduling algorithm model based on the disturbed state variables, and exchanges and calculates state variables with neighboring nodes until the algorithm converges to obtain the optimal scheduling scheme.

[0005] On the other hand, a privacy-preserving distributed scheduling system for a dual-mode microgrid is provided, the system comprising: Establish a module for building a distributed scheduling algorithm model for a dual-mode microgrid; The disturbance injection module is used to inject a set of disturbance values ​​that satisfy zero-sum constraints into the state variables of each power generation node participating in the scheduling at the initial moment when the distributed scheduling algorithm model starts execution, according to the operation mode instruction. The execution module is used by each power generation node to execute the distributed scheduling algorithm model based on the disturbed state variables. It exchanges and calculates state variables with neighboring nodes until the algorithm converges and obtains the optimal scheduling scheme.

[0006] On the other hand, an electronic device is provided, comprising a processor and a memory, wherein the memory stores at least one instruction, which is loaded and executed by the processor to implement the above-described privacy-preserving distributed scheduling method for a dual-mode microgrid.

[0007] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, the at least one instruction being loaded and executed by a processor to implement the above-described privacy-preserving distributed scheduling method for a dual-mode microgrid.

[0008] The beneficial effects of the technical solution provided by this invention include at least the following: This invention can effectively prevent external eavesdroppers and honest but curious legitimate nodes from obtaining real initial sensitive information, and does not affect the final convergence result of the distributed scheduling algorithm, ensuring that the globally optimal economic scheduling scheme is obtained. At the same time, it can adapt to various operating modes of microgrids, and has low computational and communication overhead, making it easy to deploy in actual microgrid control systems. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart of a privacy-protected distributed scheduling method for a dual-mode microgrid provided in an embodiment of the present invention. Figure 2 The physical structure and communication topology of the microgrid system in this embodiment of the invention; Figure 3 This is a schematic table for selecting power generation parameters provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the simulation results of adding a disturbance at the initial moment of the grid connection mode in an embodiment of the present invention; Figure 5 This is a schematic diagram of the simulation results of adding a perturbation at the initial moment of the island mode in an embodiment of the present invention; Figure 6 This is a schematic diagram of the simulation results against external eavesdroppers in an embodiment of the present invention; Figure 7 This is a schematic diagram of the simulation results for nodes that are internally honest but curious, according to an embodiment of the present invention. Figure 8 This is a block diagram of a privacy-protected distributed scheduling system for a dual-mode microgrid provided in an embodiment of the present invention. Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0011] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0012] This invention provides a privacy-protected distributed scheduling method for a dual-mode microgrid, which can be implemented by an electronic device, such as a terminal or a server. Figure 1 The diagram shown is a flowchart of the method. The processing flow may include the following steps: S1. Establish a distributed scheduling algorithm model for a dual-mode microgrid; Optionally, S1 specifically includes: Establish a distributed scheduling algorithm model and calculate the incremental cost of each power generation node. and power mismatch The iterative updates make the incremental costs of each power generation node in the microgrid more affordable. They are all equal to the same value, and the power mismatch is... When the value is 0, the total generation cost of the microgrid is minimized. The distributed scheduling algorithm model is as follows:

[0013] in, , This represents the number of generator nodes in a microgrid. It is a power generation node exist Incremental cost per moment It is a power generation node exist Local power mismatch at any given time. It is a power generation node The set of neighboring nodes, and For mode switching parameters, The time indicates the island mode. The time indicates the grid connection mode. As an acceleration factor, and , For consistent step size, For feedback gain coefficient, These are the adjacency matrix elements of the communication topology. To indicate the power generation node Whether it is connected to the leader node, where the index of the leader node is 0, representing the main power grid. The main grid electricity price, also known as the main grid incremental cost, For intermediate calculation variables, It is an auxiliary variable, only used at the switching time. The value is 1 at one time and 0 at other times. It is a power generation node exist The main power grid power distribution variables at time t, and It is a power generation node exist The compensation amount based on the power mismatch at any given time; The incremental cost Defined as:

[0014] in, For the first The output power of each power generation node All are the first The power generation cost coefficient of each power generation node.

[0015] Optionally, the intermediate calculation variable The update formula is: in, For consistent step size, It is a power generation node The power generation cost factor selected in grid-connected mode. It is a power generation node exist Output power at any given moment.

[0016] More specifically, S1 includes: S11. Establish a microgrid scheduling problem model; S12. Introduce the Lagrange multiplier method, construct the Lagrange function, and establish the Lagrange equation to solve the microgrid scheduling problem model. S13. Based on the Lagrange equation, the incremental cost and power mismatch of the generator are selected as consistency variables, and a distributed scheduling algorithm model is established. The distributed scheduling algorithm model is used to perform iterative calculations based on the received instructions representing the current operating mode, and the optimal power of each generator set is output. In grid-connected mode, the distributed scheduling algorithm model accelerates the consistency algorithm by introducing an acceleration factor. In islanded mode, the acceleration factor is disabled to adopt the traditional consistency algorithm.

[0017] More specifically, S11 includes: S111, Establishing the first [missing information] in microgrid dispatching Objective function for generator power generation cost :

[0018] in, , This represents the number of generator nodes in a microgrid. For the first The output power of the generator All are the first The power generation cost coefficient of a generator; S112. The microgrid scheduling problem model is established as follows:

[0019] Wherein, equation (1) is the objective function, equation (2) is the supply and demand balance constraint at the power generation and consumption ends, and equation (3) is the first equation. Upper and lower limits of the generator's output power. and These are mode switching parameters used in distributed scheduling algorithms. The time indicates the island mode. The time indicates the grid connection mode. The main grid electricity price, also known as the main grid incremental cost, It is the total power exchanged between the microgrid and the main grid. For total power requirements, It is the first The power requirements of the generator. For the first Minimum output power limit for generators. For the first Maximum output power limit of the generator.

[0020] More specifically, S12 includes: By introducing the Lagrange multiplier method, we construct the Lagrange function and establish the following Lagrange equation:

[0021] in It is a constructed Lagrange function. It is the introduced Lagrange multiplier, also known as incremental cost; The necessary and sufficient condition for minimizing power generation costs is that the first derivative of the Lagrange function is zero. Therefore, the Lagrange function... Calculate the power of each generator separately. Power exchange between microgrids and main grid and Lagrange multipliers The partial derivative of , and set it equal to zero:

[0022] At the same time, define the first The Lagrange multipliers of the generator are also the multipliers of each generator node. incremental cost for:

[0023] According to the formula The conclusion is that for the power generation cost to reach its minimum, the incremental cost of all generators must be equal, and equal to a uniform incremental cost. This is called the optimal solution. According to the formula Conclusion: When The optimal solution when the grid is connected. According to the formula The conclusion is that when the power generation cost reaches its minimum value, the supply and demand balance constraint on both the power generation and consumption sides is satisfied.

[0024] More specifically, in step S13, based on the Lagrange equation, the incremental cost and power mismatch of the generator are selected as consistency variables to establish a distributed scheduling algorithm model, specifically including: Incremental cost of selecting generator and power mismatch As a consistency variable, power mismatch Essentially, it's about each power generation node. Local load demand Compared with its current power generation capacity The estimator of the difference aims to be iteratively updated using a distributed algorithm to ensure that all nodes... When all nodes reach the same value, there is a power mismatch. Reaching 0, that is, all equal All The value is equal to 0, thus minimizing the power generation cost while satisfying the constraints. At this point, through... Calculate the optimal solution for output power. :

[0025] in, No. The incremental cost corresponding to the upper limit constraint on the output power of a generator. No. The incremental cost corresponding to the lower limit constraint of the generator output power. It is the optimal solution for incremental cost. It is the first The output power of a generator at which the cost is minimized; Incremental cost Update formula: in, It is the first A generator in Incremental cost per moment It is the first A generator in Local power mismatch at any time It is the first The set of neighboring nodes of the generator. and For mode switching parameters, As an acceleration factor, and , For consistent step size, For feedback gain coefficient, These are the adjacency matrix elements of the communication topology. To represent the power generation unit Whether it is connected to the leader node, where the index of the leader node is 0, representing the main power grid; Power mismatch Update formula: in, For intermediate calculation variables, It is an auxiliary variable, only used at the switching time. The value is 1 at one time and 0 at other times. It is a power generation unit exist The power distribution variables of the main power grid at any given time. and It is a power generation unit exist The amount of compensation based on power mismatch at all times; Intermediate calculation variables The update formula is:

[0026] in, For consistent step size, It is a power generation node Selection of power generation cost coefficient in grid-connected mode It is a power generation node exist Output power at any given moment; Power generation node exist Output power at time The update formula is:

[0027] Compensation based on power mismatch The update formula is: Compensation based on power mismatch The update formula is:

[0028] in It is the first The optimal output power of the generator when connected to the grid is determined by... It was calculated in time; Main grid power distribution variables The update formula is:

[0029] in, It is a power generation node exist Changes in power distribution across the main power grid at any given time; Optionally, in step S13, the distributed scheduling algorithm model is used to perform iterative calculations based on the received instructions representing the current operating mode, and the optimal power of each generator set is output, specifically including: S131, When the current operating mode is received (the operating mode is determined by the mode switching parameter) and Characterization, when the instruction of the running mode is When operating in grid-connected mode, the command for the operating mode is... When running in isolated mode, the instructions are: At this time, that is, in the isolated mode, at this time and The update formula is a classic, well-proven traditional consensus algorithm:

[0030] In islanded mode, the main grid power allocation variable in the algorithm Forced to zero, the power balance objective becomes a strict internal self-balancing equation: the total power generation of the microgrid must equal the total load when it is stable. The algorithm will seek a unified incremental cost that minimizes the total power generation cost of the entire microgrid by coordinating all distributed generators within the microgrid. In this state, the microgrid achieves true self-closed-loop management and does not depend on the external power grid; S132, When the received instruction for the current operating mode is At this time, it is in grid-connected mode. and Execute the formula to accelerate consistent updates:

[0031] S1321, Each power generation node Exchange incremental costs with neighboring nodes in the communication network. and power mismatch Each power generation node uses the received information to execute the accelerated consistency update formula, the core of which lies in the introduction of an acceleration factor. and historical status information and ,because The "leadership follower" item in the formula Once activated, generation nodes directly connected to the main grid will prioritize tracking the incremental costs of the main grid. And this goal is rapidly disseminated across the entire internet through the network, introducing a "momentum term". This significantly accelerates the convergence process, thereby driving up the incremental costs of all distributed generation nodes within the microgrid. Incremental cost of rapid convergence to the main grid ; S1322, Generating node connected to the main power grid Based on intermediate calculation variables Calculate the change in power distribution in the main power grid. ; S1323, Each power generation node is based on intermediate calculation variables Changes in power distribution in the main power grid And the introduced "momentum term" Update your own power mismatch value ; Repeat steps S1321 to S1323 until stability is achieved. At this point, all generating nodes... incremental cost The incremental costs are very close to those of the main power grid. And no further significant changes occur, the change in power distribution in the main power grid and power mismatch value All are 0, and at this time the power reaches a balanced state, that is, the total power generation of all power generation nodes plus the power allocation value of the main grid equals the total load power.

[0032] S2. According to the operation mode instruction, at the initial moment when the distributed scheduling algorithm model starts execution, inject a set of disturbance values ​​that satisfy the zero-sum constraint into the state variables of each power generation node participating in the scheduling. Optionally, S2 specifically includes: S2.1, Set the initial value for algorithm updates to:

[0033] in It is a power generation node The incremental cost at time 0, It is a power generation node The power mismatch at time 0, i.e., the power generation node initial value, It is the first The initial output power of the generator. For the first Minimum output power limit for generators. For the first The maximum output power limit of the generator is due to the grid-connected mode. and The update formula not only requires The value at time also needs to be determined. The value at time 0, therefore, the initial value that needs to be set for the update formula of the grid-connected mode is not only at time 0, but also at time -1, that is... and , It is the first The local load of the power generation node, and It is the first The transient balance guidance value set by the power generation node. It is a power generation node The power allocation variables of the main power grid at the initial moment; Optionally, The calculation formula is:

[0034] S2.2, Add a perturbation at the initial time: Each power generation node is based on the received operating mode instructions. and It iteratively communicates with neighboring nodes to update its own state variables, which include incremental cost. and local power mismatch Due to the initial value It contains sensitive information including initial local load and initial generation power, therefore, when sending the initial mismatch power to neighboring nodes... At that time, disturbances are added to mask them, at each power generation node. It will provide each of its communication neighbor nodes Generate a random perturbation value This value is confidential; only the node knows it. I know it myself, and at the same time, the node It also calculates an internal perturbation value for itself. It satisfies the following zero-sum constraint:

[0035] S2.3, Node Its true local initial power mismatch Adding this to the corresponding perturbation value forms the spoofed information transmitted externally:

[0036] node Simultaneously receives from all its neighbors Information after disturbance .

[0037] S3. Each power generation node executes the distributed scheduling algorithm model based on the disturbed state variables, and exchanges and calculates state variables with neighboring nodes until the algorithm converges to obtain the optimal scheduling scheme.

[0038] Optionally, S3 specifically includes: S3.1, Node Use the received disturbance-related neighbor information and internal disturbance values The first iteration is performed according to the distributed scheduling algorithm model, i.e. : S3.2, Starting from the second moment, that is All nodes stop injecting new disturbances and begin exchanging real, undisturbed state information: The distributed scheduling algorithm model continues to be executed. Since the zero-sum constraint is satisfied, these initial disturbances will cancel each other out in the subsequent iterations, so they do not affect the final convergence accuracy of the algorithm at all. The final convergence result of the algorithm is exactly the same as that without disturbance.

[0039] Optionally, an external eavesdropper intercepts data transmitted on the communication links between all nodes, obtaining the following information set:

[0040] in It refers to the topology of the entire communication network. For nodes Send to the node at the initial moment The protected initial power mismatch information, as observed by an external eavesdropper. It is the true initial value With random disturbances Because the information obtained by the external eavesdropper is the same under two different initial state values, the eavesdropper cannot determine which initial value the observed data comes from, and cannot deduce the true initial sensitive information of any node.

[0041] Alternatively, legitimate nodes that are honest but curious inside. While adhering to the algorithm protocol, the set of all received neighbor information is recorded as follows:

[0042] in, It refers to the topology of the entire communication network. For neighboring nodes Send to the node at the initial moment Protected initial power mismatch information, internally honest but curious legitimate nodes. Observed It is the true initial value With random disturbances The superposition of these elements, under two different initial state values, results in honest but curious legitimate nodes. The information available is the same, making it impossible to uniquely identify neighboring nodes from the information obtained. The true initial sensitive information.

[0043] When the operating mode switching command is triggered, a smooth transition between modes is achieved through a cooperative state transition mechanism to ensure power balance when stable.

[0044] Optionally, S3 further includes: At any moment Upon receiving the mode switching command, it uses auxiliary variables Achieving a coordinated state transition triggers a brief, one-time special calculation cycle to synchronously adjust the power mismatch. Status value: At time When the mode switching command is received, it is because Switch to That is, when switching from island mode to grid-connected mode, At any given time, the grid-connected mode is in operation, and the auxiliary variable is... exist The time is 1, and at this time... Activated This compensation amount undergoes a jump, ensuring convergence while maintaining power balance in subsequent iterations. Time, auxiliary variable The value is 0. At this point, the update formula is the accelerated consensus algorithm for grid-connected mode, and the values ​​of each power generation node are 0. Driven by the acceleration mechanism, it quickly converges to a unified state. Main power grid power distribution variables It is also dynamically adjusted to the optimal value based on the economics within the microgrid, while ensuring that the total power generation of all power generation nodes plus the total power allocation value of the main grid equals the total load power. At time When the mode switching command is received, it is because Switch to That is, when switching from grid-connected mode to islanded mode, In isolated mode, the auxiliary variable... exist The time is 1, and at this time... Activated This compensation amount undergoes a jump, ensuring convergence while maintaining power balance in subsequent iterations. Time, auxiliary variable The value is 0. In this case, the update formula is the classic, well-proven traditional consensus algorithm in islanded mode, where each generator node... It smoothly converges to a uniform incremental cost that minimizes the total generation cost of the entire microgrid. At the same time, it ensures that the total power generation of the power generation nodes is equal to the total load power.

[0045] In the embodiments of the present invention, the following are employed: Figure 2 The microgrid topology connection method shown in the present invention adopts the following approach for the method of this embodiment: Figure 3 The data shown was verified through simulation experiments.

[0046] Figure 2The diagram shows the physical structure and communication topology of a smart grid system. G1, G2, G3, G4, and G5 represent the first, second, third, fourth, and fifth generating nodes, respectively. MG represents the main grid, and ER represents the energy router, a key device connecting the microgrid and the main grid. A consistency step size is selected. Feedback gain coefficient , Incremental cost of the main power grid The local load is: , , , , , Initial output power: , , , , .

[0047] If in When selecting the grid connection mode, that is, selecting Substituting the above parameters into the initial value calculation formula, the initial value of the incremental cost can be obtained. , , , , Initial value of power mismatch , , , , , , , , , , set the initial value And with added perturbation Send it to neighboring nodes, perform initial iterations, and then... By iterating through the sent real values, the optimal value can be obtained. , , , , , , , , .

[0048] exist Switching is performed at any time, from grid-connected mode to islanded mode. , ,Bundle time , and the calculated compensation amount based on power mismatch. Substituted into and The update formula yields the optimal value and optimal solution. , , , , , .

[0049] exist Switching is performed at any time, from island mode to grid-connected mode. , ,Bundle time , and the calculated compensation amount based on power mismatch , , , , ,and of and The optimal value can be obtained from the update formula. , , , , , , , , The optimal solution obtained by adding a disturbance at the initial moment in the grid-connected mode is the same as the optimal solution obtained without a disturbance.

[0050] If in When selecting the island mode, that is, selecting Substituting the above parameters into the initial value calculation formula, the initial value of the incremental cost can be obtained. , , , , Initial value of power mismatch , , , , , set the initial value And with added perturbation Send it to neighboring nodes, perform initial iterations, and then... By iterating through the sent real values, the optimal solution for the isolated island can be obtained. , , , , , .

[0051] exist Switching is performed at any time, from island mode to grid-connected mode. , ,Bundle time , and the calculated compensation amount based on power mismatch , , , , ,and of and The optimal value can be obtained from the update formula. , , , , , , , , .

[0052] exist Switching is performed at any time, from grid-connected mode to islanded mode. , ,Bundle time , and the calculated compensation amount based on power mismatch. Substituted into and The optimal solution can be obtained from the update formula. , , , , , Here, the optimal solution obtained by adding a perturbation at the initial moment in the island mode is the same as the optimal solution obtained without the perturbation.

[0053] Figure 4 This is a simulation diagram with a disturbance added at the initial moment of the grid-connected mode. Figure 4 (a) is the convergence result of the cost increment of each power generation node at the initial moment of the grid-connected mode after adding a disturbance, according to Figure 4 (a) It can be seen that when a perturbation is added at the initial moment of the grid-connected mode, the incremental cost converges to the optimal value regardless of whether it is in grid-connected mode or switched to islanded mode. , Figure 4 (b) represents the convergence result of power mismatch, based on Figure 4 (b) It can be seen that the power mismatch result after the initial perturbation approaches zero. Figure 4 (c) represents the convergence of output power at each power generation node, based on... Figure 4 (c) It can be seen that the local active power converges to the optimal value after the initial perturbation. , Figure 4 (d) represents the sum of engine output power and main grid power, expressed as the result of load demand, based on... Figure 4 (d) It can be seen that even with the initial disturbance in the grid-connected mode, the entire network can still achieve supply and demand balance in steady state. Meanwhile, in and The microgrid's operating mode changed at a certain time. The simulation results show that all variables converged to the new optimal value, indicating that the proposed initial time perturbation does not affect the final convergence result of the distributed scheduling algorithm.

[0054] Figure 5 This is a simulation diagram with perturbations added at the initial moment of the island mode. Figure 5 (a) shows the convergence result of the cost increment of each power generation node after the initial perturbation, based on... Figure 5 (a) It can be seen that when a perturbation is added at the initial moment of the islanded mode, the incremental cost converges to the optimal value regardless of whether it is the islanded mode or the grid-connected mode. , Figure 5 (b) represents the convergence result of power mismatch, based on Figure 5 (b) It can be seen that the initial perturbation power mismatch in the islanded mode approaches zero. Figure 5 (c) represents the convergence of output power at each power generation node, based on... Figure 5 (c) It can be seen that the local active power converges to the optimal value when the perturbation is added at the initial moment of the islanding mode. , Figure 5(d) represents the sum of engine output power and main grid power, expressed as the result of load demand, based on... Figure 5 (d) It can be seen that even with the initial perturbation in the islanded mode, the entire network can still achieve supply and demand balance in steady state. Meanwhile, in and The microgrid's operating mode changed at a certain time. The simulation results show that all variables converged to the new optimal value, indicating that the proposed initial time perturbation does not affect the final convergence result of the distributed scheduling algorithm.

[0055] Figure 6 This is a diagram simulating the information set obtained by an external eavesdropper. , Figure 6 (a) and (b) are under different initial values Below value, Figure 6 (c) represents different initial values Below value, Figure 6 (d) The x-axis is under the original initial value. The following is The ordinate is another set of initial values. Below It is clear that the information set obtained by the external eavesdropper under two different initial values ​​is the same. Therefore, the external eavesdropper cannot determine which set of initial values ​​the observed data sequence comes from, and thus cannot infer the true initial local load or initial power generation information of any node with any precision.

[0056] Figure 7 This is a simulation diagram for internally honest but curious nodes, where the information set obtained by these nodes is... Suppose that nodes 4 and 5 collude to obtain the true initial value of node 1, and the information set they obtain is: and Node 2 is a legitimate neighbor of Node 1. Figure 7 (a) and (b) represent the results for node 1 with two sets of unequal initial values. Below value, Figure 7 (c) represents the two sets of unequal initial values ​​for node 1 and node 5. and Below value, Figure 7 (d) The x-axis is under the original initial value. and of and The ordinate is another set of initial values. and of and It is obvious that the information sets obtained by nodes 4 and 5 under the two unequal initial values ​​of nodes 1 and 2 are the same. Therefore, nodes 4 and 5 cannot uniquely determine neighbor 1 from their information sets. The true initial sensitive information.

[0057] like Figure 8 As shown, this embodiment of the invention also provides a privacy-protected distributed scheduling system for a dual-mode microgrid, the system comprising: Module 810 is established to build a distributed scheduling algorithm model for a dual-mode microgrid. The disturbance injection module 820 is used to inject a set of disturbance values ​​that satisfy zero-sum constraints into the state variables of each power generation node participating in the scheduling at the initial moment when the distributed scheduling algorithm model starts execution, according to the operation mode instruction. The execution module 830 is used by each power generation node to execute the distributed scheduling algorithm model based on the disturbed state variables, and to obtain the optimal scheduling scheme by exchanging and calculating state variables with neighboring nodes until the algorithm converges.

[0058] The present invention provides a privacy-protected distributed scheduling system for a dual-mode microgrid, the functional structure of which corresponds to the privacy-protected distributed scheduling method for a dual-mode microgrid provided in the present invention, and will not be described again here.

[0059] Figure 9 This is a schematic diagram of the structure of an electronic device 900 provided in an embodiment of the present invention. The electronic device 900 may vary considerably due to different configurations or performance. It may include one or more central processing units (CPUs) 901 and one or more memories 902. The memory 902 stores at least one instruction, which is loaded and executed by the processor 901 to implement the steps of the privacy-protected distributed scheduling method for the dual-mode microgrid described above.

[0060] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including instructions that can be executed by a processor in a terminal to complete the privacy-preserving distributed scheduling method for the dual-mode microgrid described above. For example, the computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, or optical data storage device, etc.

[0061] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A privacy-preserving distributed scheduling method for a dual-mode microgrid, characterized in that, The method includes: S1. Establish a distributed scheduling algorithm model for a dual-mode microgrid; S2. According to the operation mode instruction, at the initial moment when the distributed scheduling algorithm model starts execution, inject a set of disturbance values ​​that satisfy the zero-sum constraint into the state variables of each power generation node participating in the scheduling. S3. Each power generation node executes the distributed scheduling algorithm model based on the disturbed state variables, and exchanges and calculates state variables with neighboring nodes until the algorithm converges to obtain the optimal scheduling scheme.

2. The method according to claim 1, characterized in that, S1 specifically includes: Establish a distributed scheduling algorithm model and calculate the incremental cost of each power generation node. and power mismatch The iterative updates make the incremental costs of each power generation node in the microgrid more affordable. They are all equal to the same value, and the power mismatch is... When the value is 0, the total generation cost of the microgrid is minimized. The distributed scheduling algorithm model is as follows: in, , This represents the number of generator nodes in a microgrid. It is a power generation node exist Incremental cost per moment It is a power generation node exist Local power mismatch at any given time. It is a power generation node The set of neighboring nodes, and For mode switching parameters, The time indicates the island mode. The time indicates the grid connection mode. As an acceleration factor, and , For consistent step size, For feedback gain coefficient, These are the adjacency matrix elements of the communication topology. To indicate the power generation node Whether it is connected to the leader node, where the index of the leader node is 0, representing the main power grid. The main grid electricity price, also known as the main grid incremental cost, For intermediate calculation variables, It is an auxiliary variable, only used at the switching time. The value is 1 at one time and 0 at other times. It is a power generation node exist The main power grid power distribution variables at time t, and It is a power generation node exist The compensation amount based on the power mismatch at any given time; The incremental cost Defined as: in, For the first The output power of each power generation node All are the first The power generation cost coefficient of each power generation node.

3. The method according to claim 2, characterized in that, The intermediate calculation variables The update formula is: in, For consistent step size, It is a power generation node The power generation cost factor selected in grid-connected mode. It is a power generation node exist Output power at any given moment.

4. The method according to claim 3, characterized in that, S2 specifically includes: S2.1, Set the initial value for algorithm updates to: in It is a power generation node The incremental cost at time 0, It is a power generation node The power mismatch at time 0, i.e., the power generation node initial value, It is the first The initial output power of the generator. For the first Minimum output power limit for generators. For the first The maximum output power limit of the generator is due to the grid-connected mode. and The update formula not only requires The value at time also needs to be determined. The value at time 0, therefore, the initial value that needs to be set for the update formula of the grid-connected mode is not only at time 0, but also at time -1, that is... and , It is the first The local load of the power generation node, and It is the first The transient balance guidance value set by the power generation node. It is a power generation node The power allocation variables of the main power grid at the initial moment; S2.2, Add a perturbation at the initial time: Each power generation node is based on the received operating mode instructions. and It iteratively communicates with neighboring nodes to update its own state variables, which include incremental cost. and local power mismatch Due to the initial value It contains sensitive information including initial local load and initial generation power, therefore, when sending the initial mismatch power to neighboring nodes... At that time, disturbances are added to mask them, at each power generation node. It will provide each of its communication neighbor nodes Generate a random perturbation value This value is confidential; only the node knows it. I know it myself, and at the same time, the node It also calculates an internal perturbation value for itself. It satisfies the following zero-sum constraint: S2.3, Node Its true local initial power mismatch Adding this to the corresponding perturbation value forms the spoofed information transmitted externally: node Simultaneously receives from all its neighbors Information after disturbance .

5. The method according to claim 4, characterized in that, S3 specifically includes: S3.1, Node Use the received disturbance-related neighbor information and internal disturbance values The first iteration is performed according to the distributed scheduling algorithm model, i.e. : S3.2, Starting from the second moment, that is All nodes stop injecting new disturbances and begin exchanging real, undisturbed state information: The distributed scheduling algorithm model continues to be executed. Since the zero-sum constraint is satisfied, these initial disturbances will cancel each other out in the subsequent iterations, so they do not affect the final convergence accuracy of the algorithm at all. The final convergence result of the algorithm is exactly the same as that without disturbance.

6. The method according to claim 5, characterized in that, An external eavesdropper intercepts data transmitted on all communication links between nodes, obtaining the following information: in It refers to the topology of the entire communication network. For nodes Send to the node at the initial moment The protected initial power mismatch information, as observed by an external eavesdropper. It is the true initial value With random disturbances Because the information obtained by the external eavesdropper is the same under two different initial state values, the eavesdropper cannot determine which initial value the observed data comes from, and cannot deduce the true initial sensitive information of any node.

7. The method according to claim 5, characterized in that, Legitimate nodes that are honest but curious inside While adhering to the algorithm protocol, the set of all received neighbor information is recorded as follows: in, It refers to the topology of the entire communication network. Neighboring nodes Send to the node at the initial moment Protected initial power mismatch information, internally honest but curious legitimate nodes. Observed It is the true initial value With random disturbances The superposition of these elements, under two different initial state values, results in honest but curious legitimate nodes. The information available is the same, making it impossible to uniquely identify neighboring nodes from the information obtained. The true initial sensitive information.

8. A privacy-protected distributed scheduling system for a dual-mode microgrid, characterized in that, The system includes: Establish a module for building a distributed scheduling algorithm model for a dual-mode microgrid; The disturbance injection module is used to inject a set of disturbance values ​​that satisfy zero-sum constraints into the state variables of each power generation node participating in the scheduling at the initial moment when the distributed scheduling algorithm model starts execution, according to the operation mode instruction. The execution module is used by each power generation node to execute the distributed scheduling algorithm model based on the disturbed state variables. It exchanges and calculates state variables with neighboring nodes until the algorithm converges and obtains the optimal scheduling scheme.

9. An electronic device comprising a processor and a memory, wherein the memory stores at least one instruction, characterized in that, The at least one instruction is loaded and executed by the processor to implement the privacy-preserving distributed scheduling method for a dual-mode microgrid as described in any one of claims 1-7.

10. A computer-readable storage medium storing at least one instruction, characterized in that, The at least one instruction is loaded and executed by the processor to implement the privacy-preserving distributed scheduling method for a dual-mode microgrid as described in any one of claims 1-7.

Images