A coordinated load shedding control method, system, device and readable storage medium based on GOOSE communication

Abstract

The present invention discloses a low-frequency load shedding collaborative control method, system, device and readable storage medium based on GOOSE communication. The method specifically includes: after calling the topology analysis service to confirm that the local end is in the receiving-end power grid, sending a status query request to adjacent terminals, and receiving the action signals of adjacent terminals through GOOSE communication; performing load shedding operations based on the dynamic load shedding sequence; calculating the marginal cost functions of each terminal, optimizing the load shedding parameters after distributed consensus verification by combining blockchain technology, and sending the updates; after load shedding, locating the fault section through residual voltage detection and direction discrimination, generating a personalized reconstruction plan by combining topology identification and fault type matching, and restoring power supply to non-fault sections within a preset distance. The present invention realizes the rapid response, collaborative control, precise load shedding, intelligent optimization and rapid fault recovery of low-frequency load shedding, and significantly improves the frequency stability and power supply reliability of the power system in the scenario of high proportion of new energy access.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and particularly to a coordinated control method, system, device and readable storage medium for under-frequency load shedding based on GOOSE communication. Background Art

[0002] With the transformation of the global energy structure, the access ratio of new energy sources (such as wind power, photovoltaic power, etc.) in the power system is increasing day by day, forming an operating scenario of a power system with a high proportion of new energy access. However, new energy power generation has characteristics such as intermittency, volatility and uncertainty, posing a severe challenge to the frequency stability of the power system. When a large-scale disconnection of new energy power generation or a sudden change in load occurs, the power system frequency may drop sharply. If effective control measures are not taken in time, it is extremely easy to trigger a system frequency collapse, and then lead to large-scale power outages, seriously threatening the safe and reliable operation of the power system.

[0003] In the prior art, the under-frequency load shedding device, as a key device to ensure the frequency stability of the power system, mainly prevents the abnormal decline of frequency by automatically cutting off non-critical loads. However, most of the under-frequency load shedding devices in the current distribution network adopt a centralized control mode, with many technical defects and being difficult to meet the frequency emergency control requirements in the scenario of high proportion of new energy access. The specific manifestations are as follows:

[0004] 1. Response delay problem: Conventional under-frequency load shedding devices rely on the master station for centralized decision-making. It usually takes 300 - 500 ms from detecting frequency over-limit to performing the load shedding operation. In the scenario of millisecond-level frequency drop caused by new energy disconnection, this delay cannot meet the requirement of rapid control, may lead to the inability of the system frequency to recover in time, and then trigger more serious system oscillations or even collapse.

[0005] 2. Rigid strategy problem: The load shedding strategies of existing under-frequency load shedding devices mostly cut off loads in fixed rounds, lacking the ability to dynamically adjust the load priority. In actual operation, the importance and interruptibility of different loads are different. The fixed-round load shedding method is prone to over-cutting or under-cutting phenomena, not only wasting precious load resources, but also may exacerbate system oscillations and affect the stable operation of the power system.

[0006] 3. Poor coordination problem: In a distribution network with high penetration of distributed power sources, there is no effective coordination mechanism among multiple under-frequency load shedding terminals. When the system frequency is abnormal, each terminal may act simultaneously, resulting in over-cutting of loads and affecting power supply reliability. At the same time, there is no effective coordination between the existing feeder automation terminal and the under-frequency protection device, and it is difficult to synchronously execute fault location, isolation and load control operations, further reducing the stability and reliability of power supply.

[0007] 4. Problem of misoperation in single-criterion protection: Traditional under-frequency load shedding devices usually operate only based on the frequency criterion, without fully considering the impact of factors such as voltage disturbances on frequency measurement. In actual operation, voltage disturbances may cause false frequency fluctuations, leading to misoperation of the under-frequency load shedding device, cutting off unnecessary loads, and affecting the normal operation of the power system. Summary of the Invention

[0008] The purpose of the present invention is to provide a collaborative control method, system, device, and readable storage medium for under-frequency load shedding based on GOOSE communication, which realizes rapid response, collaborative control, precise load shedding, intelligent optimization, and fast fault recovery of under-frequency load shedding, significantly improves the frequency stability and power supply reliability of the power system in the scenario of high proportion of new energy access, and solves at least one of the above-mentioned existing technical problems.

[0009] In the first aspect, the present invention provides a collaborative control method for under-frequency load shedding based on GOOSE communication, and the method specifically includes:

[0010] By real-time monitoring of the frequency, voltage, and phase angle parameters of the power system, calculating the frequency change rate, and when the frequency change rate exceeds a preset change rate threshold, triggering the frequency protection thread;

[0011] After calling the topology analysis service to confirm that the local end is in the receiving-end power grid, sending a status query request to adjacent terminals, receiving the action signals of adjacent terminals through GOOSE communication, if at least two action signals of adjacent terminals are received, delaying the local load shedding action and starting topology verification, otherwise directly entering the load shedding execution stage;

[0012] Generating a dynamic load shedding sequence according to the built-in programmable load characteristic database, based on the dynamic load shedding sequence, performing load shedding operations through a preset multi-level load shedding logic, and simultaneously synchronously feeding back the action signals to the master station;

[0013] Calculating the marginal cost function of each terminal based on the game theory model at the master station, optimizing the load shedding parameters after distributed consensus verification in combination with blockchain technology, and issuing updates;

[0014] After load shedding, locating the fault section through residual voltage detection and direction discrimination, generating a personalized reconstruction plan in combination with topology recognition and fault type matching, and restoring power supply to non-fault sections within a preset distance.

[0015] In the second aspect, the present invention provides a collaborative control system for under-frequency load shedding based on GOOSE communication, and the system specifically includes:

[0016] A hardware frequency measurement module for real-time monitoring of the frequency, voltage, and phase angle parameters of the power system, calculating the frequency change rate, and triggering the frequency protection thread when the frequency change rate exceeds a preset change rate threshold;

[0017] A coordinated control module, which is used to call the topology analysis service to confirm that the local side is in the receiving-end power grid, then send a status query request to adjacent terminals, receive the action signals of adjacent terminals through GOOSE communication. If at least two action signals of adjacent terminals are received, the local load shedding action is delayed and topology verification is started; otherwise, it directly enters the load shedding execution stage.

[0018] A load priority decision module, which is used to generate a dynamic load shedding sequence according to the built-in programmable load characteristic database, and based on the dynamic load shedding sequence, execute the load shedding operation through the preset multi-level load shedding logic, and at the same time synchronously feedback the action signal to the master station.

[0019] A master station linkage module, which is used to calculate the marginal cost function of each terminal by the master station based on the game theory model, optimize the load shedding parameters after distributed consensus verification combined with blockchain technology, and issue updates.

[0020] A self-healing reconstruction module, which is used to locate the fault section through residual voltage detection and direction discrimination after load shedding is completed, generate a personalized reconstruction plan by combining topology recognition and fault type matching, and restore the power supply of non-fault sections within a preset distance.

[0021] Thirdly, the present invention provides a computer device, including: a memory, a processor, and a computer program stored on the memory. When the computer program is executed on the processor, it implements the low-frequency load shedding coordinated control method based on GOOSE communication as described in any one of the above methods.

[0022] Fourthly, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it implements the low-frequency load shedding coordinated control method based on GOOSE communication as described in any one of the above methods.

[0023] Compared with the prior art, the present invention has at least one of the following technical effects:

[0024] 1. The present invention realizes the rapid response, coordinated control, precise load shedding, intelligent optimization and rapid fault recovery of low-frequency load shedding, and significantly improves the frequency stability and power supply reliability of the power system in the scenario of high proportion of new energy access.

[0025] 2. Through real-time monitoring and distributed computing, the present invention can detect frequency anomalies and trigger protection actions within milliseconds, significantly shortening the response time and meeting the control requirements of the millisecond-level frequency drop scenario caused by new energy disconnection from the grid.

[0026] 3. Based on the dynamic load shedding sequence and multi-level load shedding logic, the present invention can accurately classify and cut off according to the importance and interruptibility of the load, avoiding over-cutting or under-cutting phenomena, improving the utilization efficiency of load resources, and reducing the risk of system oscillation.

[0027] 4. The present invention realizes information interaction and cooperative control among multiple terminals through GOOSE communication, ensuring that each terminal can synchronously execute fault location, isolation, and load control operations when the system frequency is abnormal, improving the reliability and stability of power supply.

[0028] 5. Combining the game theory model and blockchain technology, the present invention can dynamically optimize load shedding parameters, realize intelligent adjustment of load shedding strategies and distributed consensus verification, and improve the scientificity and accuracy of load shedding decisions.

[0029] 6. Through technical means such as residual voltage detection, direction discrimination, and topology identification, the present invention can quickly locate the fault section and generate a personalized reconstruction plan to realize the rapid restoration of power supply in the non-fault section, further improving the reliability and stability of power supply in the power system.

[0030] 7. By real-time collecting three-phase voltage signal parameters and accurately calculating the frequency change rate, the present invention can quickly and accurately trigger the frequency protection thread, improving the response speed and accuracy of low-frequency load shedding.

[0031] 8. The present invention uses topology analysis to confirm the receiving-end power grid and coordinates the action signals of adjacent terminals, which can avoid over-cutting of the load caused by the simultaneous action of multiple terminals, and improve the coordination and reliability of low-frequency load shedding.

[0032] 9. Based on the programmable load characteristic database, the present invention generates a dynamic load shedding sequence, which can achieve accurate hierarchical load shedding according to load characteristics, ensure the power supply of critical loads, and improve the rationality of load shedding.

[0033] 10. Based on the dynamic load shedding sequence, the present invention executes the multi-level load shedding logic and synchronously feeds back the action signals, which can realize accurate load cutting and real-time monitoring, providing data support for optimizing load shedding parameters.

[0034] 11. The master station of the present invention optimizes load shedding parameters based on the game theory model and blockchain technology, which can minimize the load shedding cost and dynamically adjust the parameters, improving the rationality and adaptability of load shedding strategies.

[0035] 12. After load shedding, the present invention generates a personalized reconstruction plan through residual voltage detection and topology identification, which can quickly restore power supply in the non-fault section, reduce the power outage range, and improve the reliability of power supply. Description of the Drawings

[0036] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0037] Figure 1 is a schematic flow chart of a low-frequency load shedding cooperative control method based on GOOSE communication provided by an embodiment of the present invention;

[0038] Figure 2 is a schematic structural diagram of a low-frequency load shedding cooperative control system based on GOOSE communication provided by an embodiment of the present invention;

[0039] Figure 3 is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0040] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures and technologies are presented to thoroughly understand the embodiments of the present application. However, those skilled in the art should understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0041] It should be understood that when used in the specification and appended claims of the present application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0042] It should also be understood that the term "and / or" used in the specification and appended claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0043] As used in the specification and appended claims of the present application, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if detecting [the described condition or event]" can be interpreted as meaning "once determined", "in response to determining", "once detecting [the described condition or event]", or "in response to detecting [the described condition or event]" according to the context.

[0044] In addition, in the description of the specification and the appended claims of this application, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0045] The reference to "one embodiment" or "some embodiments" etc. described in the specification of this application means that in one or more embodiments of this application, specific features, structures or characteristics described in connection with that embodiment are included. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments" etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0046] In the embodiments of this application, the execution subject of the process includes a terminal device. The terminal device includes but is not limited to: devices such as servers, computers, smart phones, and tablet computers that can execute the methods disclosed in this application. Figure 1 The flowchart of the low-frequency load shedding cooperative control method based on GOOSE communication disclosed in the first embodiment of the present invention is shown as follows and is described in detail below:

[0047] S101, by real-time monitoring of the frequency, voltage and phase angle parameters of the power system, calculate the frequency change rate. When the frequency change rate exceeds a preset change rate threshold, trigger the frequency protection thread.

[0048] In this embodiment, by real-time monitoring of the frequency, voltage and phase angle parameters of the power system, calculate the frequency change rate. When the frequency change rate exceeds a preset change rate threshold, trigger the frequency protection thread to achieve fast and accurate frequency control.

[0049] Specifically, deploy high-precision synchronized phasor measurement units (PMUs) at each terminal of the distribution network (such as substations, distributed power access points, etc.) to collect the frequency, voltage and phase angle parameters of the power system in real time. The sampling frequency of the PMU is not less than 10 kHz to ensure high-precision and real-time parameter measurement.

[0050] The terminal is built with a frequency change rate calculation module, which uses a sliding window algorithm to process the collected frequency data in real time. The specific calculation method is as follows: set the sliding window length to N (such as N = 100 sampling points), and the corresponding time window length is T = N / sampling frequency (such as T = 10 ms). In each time window, calculate the average value of the frequency, and calculate the frequency change rate between two adjacent time windows , where represents the average value of the frequency of the current time window, represents the average frequency of the previous time window, represents the time window length. According to the operation experience and simulation analysis of the power system, the threshold of the frequency change rate is set. When the frequency change rate exceeds this threshold, the frequency protection thread is triggered.

[0051] In this embodiment, through the high-precision PMU and the sliding window algorithm, the real-time calculation of the frequency change rate is realized, and the response time is shortened to the 10 ms level, which is much lower than the 300 - 500 ms delay of the traditional under-frequency load shedding device. The setting of the frequency change rate threshold is based on the actual operation characteristics of the power system, avoiding misoperation caused by false frequency fluctuations and improving the accuracy of frequency protection.

[0052] S102, after calling the topology analysis service to confirm that the local end is in the receiving-end power grid, send a status query request to the adjacent terminal, and receive the action signal of the adjacent terminal through GOOSE communication. If the action signals of at least two adjacent terminals are received, delay the local load shedding action and start topology verification, otherwise directly enter the load shedding execution stage.

[0053] In this embodiment, when multiple under-frequency load shedding terminals act independently, it is easy to cause over-cutting or under-cutting of the load, affecting the stable operation of the power system. Therefore, the present invention calls the topology analysis service to confirm the position of the local power grid and makes a collaborative decision in combination with the action signals of adjacent terminals to achieve precise load shedding control.

[0054] Specifically, a topology analysis service module is built in each terminal of the distribution network (such as substations, distributed power access points, etc.). This module obtains the topology structure information of the current power grid by performing real-time data interaction with the master station or adjacent terminals. The terminal judges whether the local end is in the receiving-end power grid according to the information returned by the topology analysis service. The specific judgment logic is as follows: If the power flow direction of the local power grid is the power receiving direction (that is, power flows into the local power grid from other regions), it is determined that the local end is in the receiving-end power grid; if the power flow direction of the local power grid is the power sending direction (that is, power flows out of the local power grid to other regions), it is determined that the local end is in the sending-end power grid.

[0055] After the terminal confirms that the local end is in the receiving-end power grid, it immediately sends a status query request to the adjacent terminal through GOOSE communication. GOOSE communication is a high-speed and reliable substation automation communication protocol, suitable for scenarios with high real-time requirements. After receiving the status query request, the adjacent terminal generates an action signal according to its own operating status (such as whether load shedding has been started, the current frequency value, etc.) and returns it to the requesting terminal through GOOSE communication. The action signal uses binary coding. For example, "1" indicates that load shedding has been started, and "0" indicates that load shedding has not been started.

[0056] After receiving the action signals of adjacent terminals, the requesting terminal makes a judgment according to the preset collaborative decision-making logic: If the action signals of at least two adjacent terminals are received (that is, at least two adjacent terminals have initiated load shedding), the load shedding action of this terminal is delayed, and topology verification is initiated; if the action signals of at least two adjacent terminals are not received, it directly enters the load shedding execution phase.

[0057] During the period of delaying the load shedding action, the terminal initiates the topology verification process, and the specific steps are as follows: The terminal collects the real-time operation data of this terminal and adjacent terminals, including frequency, voltage, power flow direction, etc.; According to the collected data, the terminal reconstructs the topology structure of the power grid and analyzes the position and role of this terminal's power grid in the overall power grid; According to the topology reconstruction result, the terminal adjusts the load shedding strategy of this terminal, such as reducing the amount of load shedding, adjusting the load shedding sequence, etc., to avoid over-cutting or under-cutting of the load. After the collaborative decision-making or topology verification is completed, the terminal performs the load shedding operation according to the adjusted load shedding strategy and synchronously feeds back the action signal to the master station through GOOSE communication.

[0058] In this embodiment, through the topology analysis service, the terminal can obtain the topology structure information of the power grid in real time, accurately judge the position of this terminal's power grid, and provide a basis for subsequent collaborative decision-making. Through GOOSE communication, the terminal can perform real-time information interaction with adjacent terminals to ensure the timeliness and accuracy of collaborative decision-making. Through the action signal interaction of adjacent terminals and the collaborative decision-making logic, the over-cutting of the load caused by multiple terminals acting simultaneously is avoided, and the coordination and stability of the system are improved. The topology verification process can adjust the load shedding strategy according to the real-time topology structure of the power grid, realize precise load shedding control, reduce resource waste and system oscillation. Through collaborative decision-making and topology verification, precise load shedding control is achieved, the risk of over-cutting the load is reduced, and the stability and power supply reliability of the system are improved.

[0059] S103, generate a dynamic load shedding sequence according to the built-in programmable load characteristic database, and based on the dynamic load shedding sequence, execute the load shedding operation through the preset multi-level load shedding logic, and at the same time synchronously feed back the action signal to the master station.

[0060] In this embodiment, traditional low-frequency load shedding devices usually adopt fixed load shedding strategies, which are difficult to meet the requirements of load dynamic changes and power grid topology structure complexity under high penetration of new energy. A single load shedding strategy may lead to over-cutting or under-cutting of the load, affecting the system stability and power supply reliability. To solve the above problems, the present invention generates a dynamic load shedding sequence through the built-in load characteristic database, combines the multi-level load shedding logic to achieve hierarchical load shedding, and synchronously feeds back the action signal to the master station to realize precise load shedding control.

[0061] Specifically, a programmable load characteristic database is built into the terminal device (such as substation RTU, distributed power controller, etc.). This database contains dynamic characteristic parameters of different types of loads, such as frequency-load response curves, recovery time, importance levels, etc. The database supports online updates. Through data interaction with the master station or adjacent terminals, it can obtain real-time information on changes in load characteristics (such as new energy access, load fluctuations, etc.) and dynamically adjust the content of the database.

[0062] The terminal device collects real-time operation data of the power grid, such as frequency, voltage, power flow, etc., and combines the load characteristic parameters in the programmable load characteristic database to evaluate the importance of the load. Based on the evaluation results, an optimization algorithm (such as genetic algorithm, particle swarm algorithm, etc.) is used to generate a dynamic load shedding sequence. The sequence contains information such as the priority, shedding order, and shedding amount of the loads to be shed.

[0063] According to the dynamic load shedding sequence, the terminal device executes the load shedding operation using a multi-level load shedding logic. The multi-level load shedding strategy includes but is not limited to: First-level load shedding: Shed the load with the lowest priority, triggering the frequency recovery threshold 1; Second-level load shedding: If the frequency does not recover to threshold 1, then shed the load with the second lowest priority, triggering the frequency recovery threshold 2; Third-level load shedding: And so on until the frequency recovers to the normal range.

[0064] The terminal device gradually sheds loads according to the multi-level load shedding strategy in the order of the dynamic load shedding sequence; after each level of load shedding, the terminal device monitors the change in the power grid frequency in real time to determine whether to enter the next level of load shedding.

[0065] After each level of load shedding is completed, the terminal device generates a corresponding action signal, which includes the load shedding level, the information of the shed load, the current power grid frequency, etc. The terminal device synchronously feeds back the action signal to the master station through GOOSE communication or IEC 61850 protocol. After receiving the signal, the master station conducts a global state evaluation and strategy optimization.

[0066] Exemplarily, assume that there are multiple load nodes in a distribution network, including industrial loads, commercial loads, residential loads, etc., and the proportion of new energy access is relatively high. When the system frequency drops, the terminal device executes a dynamic and precise load shedding method based on a programmable load characteristic database and a multi-level load shedding logic. First, the terminal device collects the grid frequency (such as 49.5 Hz) and load data in real time; combines with the programmable load characteristic database to evaluate the importance of each load; uses an optimization algorithm to generate a dynamic load shedding sequence, for example: preferentially cut off commercial load B (high priority, short recovery time); then cut off industrial load C (medium priority, medium recovery time); finally retain residential load A (low priority, long recovery time). Then, execute multi-level load shedding. Among them, for the first-level load shedding, cut off commercial load B and monitor the frequency to recover to 49.7 Hz; for the second-level load shedding, if the frequency does not recover to the threshold (such as 50 Hz), then cut off industrial load C and the frequency recovers to 49.9 Hz; for the third-level load shedding, if the frequency still does not recover, retain residential load A and do not perform further load shedding. After each level of load shedding is completed, the terminal device generates an action signal and feeds it back to the master station; after receiving the signal, the master station conducts a global state assessment to confirm the system frequency recovery situation.

[0067] In this embodiment, through the dynamic load shedding sequence and the multi-level load shedding logic, precise load shedding of the load is realized, important loads are preferentially retained, and the stability and power supply reliability of the system are improved. The programmable load characteristic database can adapt to the changes of different load characteristics and provide basic data support for the generation of the dynamic load shedding sequence. The dynamic load shedding sequence is generated in real time according to the load characteristics and the grid operation state, can accurately match the load shedding requirements, and avoid over-cutting or under-cutting of the load. The multi-level load shedding logic avoids the "one-size-fits-all" problem of the traditional load shedding strategy by shedding loads in stages, and improves the flexibility and accuracy of the load shedding control. The synchronous feedback of the action signal realizes the information sharing between the master station and the terminal device and provides data support for the master station to conduct global collaborative control.

[0068] S104. Calculate the marginal cost function of each terminal at the master station based on the game theory model, and optimize and issue updates of the load shedding parameters after distributed consensus verification by combining blockchain technology.

[0069] In this embodiment, the traditional method for optimizing load shedding parameters in a power system usually relies on centralized control, and has problems such as single-point failure risk, low calculation efficiency, and difficulty in adapting to multi-agent collaborative decision-making. With the wide application of distributed energy and intelligent terminals, there is an urgent need for a decentralized, secure, trustworthy and efficient method for optimizing load shedding parameters. To solve the above problems, the present invention realizes the optimization and secure issuance and update of the load shedding parameters by calculating the marginal cost function of each terminal at the master station based on the game theory model and performing distributed consensus verification by combining blockchain technology.

[0070] Specifically, in the master station, a game model including multiple terminal devices (such as distributed power controllers, load control terminals, etc.) is constructed. Each terminal is regarded as a game participant, and its strategy space includes the load shedding amount, load shedding timing, etc. According to the game model, the master station calculates the marginal cost function of each terminal based on information such as the load characteristics of each terminal, new energy power generation prediction, and power grid operation status. The marginal cost function reflects the cost change of the terminal during the load shedding process, including economic losses, equipment wear and tear, etc.

[0071] A blockchain network is constructed between the master station and each terminal. Each terminal acts as a blockchain node and is responsible for storing and verifying transaction information. The blockchain uses a consensus algorithm (such as PBFT, Raft, etc.) to ensure data consistency and security. The master station broadcasts the calculated marginal cost function of each terminal as transaction information to the blockchain network; each terminal node verifies the transaction information and reaches an agreement through the consensus algorithm to confirm the correctness of the marginal cost function.

[0072] Based on the marginal cost function of each terminal and the blockchain consensus verification result, the master station uses an optimization algorithm (such as particle swarm optimization, genetic algorithm, etc.) to calculate the optimal load shedding parameters. The optimal load shedding parameters include the load shedding amount and load shedding timing of each terminal, aiming to minimize the total system cost. The master station broadcasts the optimized load shedding parameters as new transaction information to the blockchain network; each terminal node verifies the new transaction information and reaches an agreement through the consensus algorithm to confirm the correctness of the parameter update; after verification, the master station sends the new load shedding parameters to each terminal, and the terminal performs the load shedding operation according to the new parameters.

[0073] In this embodiment, the marginal cost function based on game theory can accurately quantify the cost change of each terminal during the load shedding process, providing a scientific basis for subsequent parameter optimization. Blockchain technology realizes decentralized distributed consensus verification, avoids the risk of single-point failure, and improves the reliability and security of the system. Through the consensus algorithm, data consistency and immutability are ensured, providing a reliable data basis for subsequent parameter optimization. The parameter optimization method based on game theory and blockchain technology can achieve a globally optimal load shedding parameter configuration, improving the economy and stability of the system. Blockchain technology enables parameter updates to be quickly and securely communicated to each terminal, realizing dynamic adjustment of load shedding parameters. Based on game theory and blockchain technology, decentralized parameter optimization and update are realized, improving the reliability and economy of the system.

[0074] S105, after load shedding, detect the residual voltage and discriminate the fault section by direction, generate a personalized reconstruction plan by combining topology recognition and fault type matching, and restore power supply to the non-fault section within a preset distance.

[0075] In this embodiment, traditional power grid fault location and power supply restoration methods usually rely on centralized control or a single criterion, suffering from problems such as low fault location accuracy, long power supply restoration time, and difficulty in adapting to complex topological structures. To solve the above problems, the present invention accurately locates the fault section after load shedding, generates a personalized reconstruction plan by combining topological information and fault types, and realizes the rapid power supply restoration of non-fault sections.

[0076] Specifically, after the load shedding operation is completed, the master station collects the voltage data of each terminal (such as distribution switches, distributed power source controllers, etc.) and calculates the residual voltage value (i.e., the voltage amplitude after the fault). By comparing the residual voltage value with a preset threshold, the sections that may have faults are preliminarily screened out. Using the direction information of the fault current and combining the current phase differences of each terminal, the fault section is further located. By comparing the consistency between the fault current direction and the preset direction, the fault range is narrowed down.

[0077] The master station constructs a topological model of the power grid according to the power grid topological structure, combining the connection relationships and switch states of each terminal. Through topological identification technology, the isolation states of non-fault sections and fault sections are determined. According to information such as the residual voltage value and fault current characteristics of the fault section, a preset fault type library is matched to determine the fault type (such as short-circuit fault, grounding fault, etc.). Combining the topological model and the fault type, the master station uses an optimization algorithm (such as genetic algorithm, simulated annealing algorithm, etc.) to generate a personalized reconstruction plan. The plan includes the power supply restoration path of non-fault sections, the switch operation sequence, etc., aiming to minimize the power supply restoration time and the affected scope.

[0078] According to the personalized reconstruction plan, the master station selects the optimal power supply restoration path. The factors considered in path selection include path length, number of switch operations, load transfer capacity, etc. The master station issues switch operation instructions to relevant terminals, and the terminals execute switch operations according to the instructions to restore the power supply of non-fault sections. During the power supply restoration process, it is ensured that the length of the restoration path does not exceed a preset distance (such as 5 kilometers) to reduce the impact on other parts of the power grid.

[0079] Exemplarily, assume that a short - circuit fault occurs in a distribution network. After the master station stabilizes the system through load shedding operation, residual voltage detection and direction discrimination are first performed. For example, the residual voltage value of terminal A is 45V, and the residual voltage value of terminal B is 110V. It is preliminarily determined that a fault may exist in the section where terminal A is located; the fault current direction of terminal A is opposite to the preset direction, and the fault current direction of terminal B is consistent with the preset direction, determining that the fault section is downstream of terminal A. The topology model shows that there is a remotely controllable switch between terminal A and terminal B, and the switch is in the open state, obtaining the topology recognition result; the residual voltage value of the fault section is low and the fault current amplitude is large, matching it as a short - circuit fault, obtaining the fault type matching result; according to the topology recognition result and the fault type matching result, a personalized reconstruction plan is generated: close the standby switch between terminal C and terminal D to restore the power supply of the non - fault section. According to the reconstruction plan, select the path of terminal C → standby switch → terminal D → non - fault section. After receiving the command, terminal C and terminal D close the standby switch, and the non - fault section restores power supply. The restored path length is 3 kilometers, meeting the preset distance constraint.

[0080] In this embodiment, by combining residual voltage detection and direction discrimination, the fault section can be quickly and accurately located, improving the accuracy of fault location. Combining topology recognition and fault type matching can generate a personalized reconstruction plan for specific fault scenarios, improving the efficiency and adaptability of power supply restoration. Through the personalized reconstruction plan and the preset distance constraint, the power supply of the non - fault section can be restored in a short time, reducing the power outage time and impact range. Based on residual voltage detection, direction discrimination, topology recognition, and fault type matching, accurate positioning of the fault section and personalized reconstruction are realized, significantly improving the efficiency and adaptability of power supply restoration.

[0081] In some embodiments, in the above step S101, by real - time monitoring the frequency, voltage, and phase - angle parameters of the power system, calculating the frequency change rate, and when the frequency change rate exceeds the preset change rate threshold, triggering the frequency protection thread, specifically includes:

[0082] Real - time collect the three - phase voltage signal parameters of the power system, where the three - phase voltage signal parameters include frequency, voltage, and phase angle;

[0083] Based on the three - phase voltage signal parameters, calculate the instantaneous frequency by the equal - precision frequency measurement method;

[0084] Based on the instantaneous frequency, within multiple sliding time windows, obtain the frequency change rate through differential calculation;

[0085] Dynamically adjust the change rate threshold according to the system inertia constant and the new - energy penetration rate. When it is detected that the frequency change rate is greater than the change rate threshold, immediately trigger the frequency protection thread.

[0086] In this embodiment, as the proportion of new energy (such as wind energy and solar energy) in the power system continues to increase, the frequency stability of the power system faces new challenges. Traditional frequency protection methods usually adopt fixed thresholds and are difficult to adapt to the changes in system dynamic characteristics brought about by the access of new energy, which may lead to misoperation or refusal of protection. Therefore, there is an urgent need for a method that can dynamically adjust the rate-of-change threshold and improve the reliability of frequency protection. To solve the above problems, the present invention calculates the rate of change of frequency by real-time monitoring of the frequency, voltage and phase angle parameters of the power system, and dynamically adjusts the rate-of-change threshold according to the system inertia constant and new energy penetration rate, triggering the frequency protection thread to improve the accuracy and adaptability of frequency protection.

[0087] Specifically, high-precision voltage sensors are arranged at key nodes of the power system (such as power plants, substations, new energy connection points, etc.) to collect three-phase voltage signal parameters in real time, including frequency, voltage and phase angle. The equal-precision frequency measurement method is used to measure the frequency of the preprocessed voltage signal. The equal-precision frequency measurement method realizes high-precision frequency measurement by comparing the phase difference between the signal to be measured and the reference signal. Multiple sliding time windows (such as 10 ms, 50 ms, 100 ms, etc.) are set, and the rate of change of frequency is calculated within each time window. The inertia constant of the power system is obtained through system identification or online estimation methods. The new energy penetration rate is calculated according to the ratio of new energy generation power to the total generation power of the system. According to the system inertia constant and new energy penetration rate, the threshold of the rate of change of frequency is dynamically adjusted. The threshold adjustment formula is: , where represents the rate-of-change threshold, represents the proportionality coefficient, represents the system inertia constant, represents the new energy penetration rate weight coefficient, represents the new energy penetration rate. When it is detected that the rate of change of frequency is greater than the dynamically adjusted threshold, the frequency protection thread is immediately triggered to perform corresponding protection actions (such as generator tripping, load shedding, etc.).

[0088] Exemplarily, assume that in a certain regional power grid, the proportion of new energy generation power is 25% and the system inertia constant is 5 s. At a certain moment, due to the sudden decrease in new energy generation power, the system frequency begins to drop rapidly. Three-phase voltage sensors are arranged at the wind farm connection point to collect voltage signals in real time, and the sampling frequency is 10 kHz. The equal-precision frequency measurement method is used to calculate the instantaneous frequency, and the measurement accuracy is 0.001 Hz. The difference value of the instantaneous frequency is calculated within a 10 ms time window to obtain the rate of change of frequency. According to the system inertia constant of 5 s and the new energy penetration rate of 25%, the rate-of-change threshold is dynamically adjusted to 2.5 Hz / s. When it is detected that the rate of change of frequency is greater than 2.5 Hz / s, the frequency protection thread is immediately triggered to cut off part of the new energy generation power to prevent the system frequency from dropping further.

[0089] In this embodiment, by dynamically adjusting the change rate threshold, the accuracy and adaptability of frequency protection are improved, and the probabilities of protection maloperation and refusal operation are reduced. By real-time monitoring the frequency, voltage and phase angle parameters of the power system, calculating the frequency change rate, and dynamically adjusting the change rate threshold according to the system inertia constant and new energy penetration rate, the frequency protection thread is triggered, which significantly improves the accuracy and adaptability of frequency protection and provides a strong guarantee for the complex power system with high proportion of new energy access.

[0090] In some embodiments, in the above step S102, after calling the topology analysis service to confirm that the local end is in the receiving-end power grid, a status query request is sent to adjacent terminals, and the action signals of adjacent terminals are received through GOOSE communication. If the action signals of at least two adjacent terminals are received, the local load shedding action is delayed and topology checking is started, otherwise, it directly enters the load shedding execution stage, which specifically includes:

[0091] Call the topology analysis service to construct an equivalent power grid model based on the node admittance matrix and real-time measurement data;

[0092] By solving the equivalent power grid model, judge whether the local end is the receiving-end power grid. If it is the receiving-end power grid, send a status query request in GOOSE format to adjacent terminals, and the status query request includes the terminal ID, time stamp and action flag;

[0093] Receive the action signals of adjacent terminals through GOOSE communication, and count the number of actions. If the number of actions is greater than or equal to 2, delay the local load shedding action;

[0094] In the delay stage, calculate the topology deviation rate based on real-time measurement data. If the topology deviation rate is less than the preset deviation rate threshold, perform the load shedding operation, otherwise terminate the action and report to the master station.

[0095] In this embodiment, in the power system, when a fault or overload occurs, the receiving-end power grid needs to respond quickly and perform load shedding operations to maintain system stability. However, existing load shedding methods are mostly based on local measurement data and lack a global analysis of the power grid topology structure, which easily leads to misjudgment or uncoordinated load shedding actions. The present invention improves the accuracy and reliability of load shedding actions through global topology analysis and coordinated control of adjacent terminals.

[0096] Exemplarily, a certain regional power grid includes three adjacent substations A, B, and C. Substation A is the receiving-end power grid, and substations B and C are the sending-end power grids. When substation A detects overload, it needs to cooperate with adjacent substations to perform load shedding operations. The monitoring system of substation A collects the nodal admittance matrix (such as the admittance matrix Y_A of substation A) and measurement data (such as voltage U_A and current I_A) in real time. An equivalent model of substation A is constructed based on Y_A and the measurement data, and it is judged whether substation A is the receiving-end power grid through power flow calculation (such as the power inflow is greater than the outflow). If substation A is the receiving-end power grid, a status query request in GOOSE format is generated, including: Terminal ID: The unique identifier of substation A (such as "A_ID"); Timestamp: The current time (such as "2023-10-10 12:00:00"); Action flag: The initial state (such as "0" indicating no action). The request is sent to adjacent substations B and C through the GOOSE communication protocol. After receiving the request, if substations B and C detect overload, they reply with action signals (such as "1" indicating action). Substation A counts the number of received action signals. If the number of actions ≥ 2 (such as both substations B and C reply "1"), substation A delays the load shedding action and starts topology verification; otherwise, substation A directly enters the load shedding execution stage. During the delay stage (such as 100 ms), substation A recalculates the topology deviation rate based on real-time measurement data (such as |ΔP| / P_rated, where ΔP is the power deviation and P_rated is the rated power). If the deviation rate is less than the preset threshold (such as 5%), load shedding operations are performed (such as cutting off some loads); otherwise, the action is terminated and reported to the master station (such as sending an alarm message "Topology deviation is too large, load shedding terminated").

[0097] In this embodiment, the receiving-end power grid is globally judged through topology analysis to avoid misjudgment; the coordinated control of adjacent terminals based on GOOSE communication ensures the consistency of load shedding actions; the topology verification mechanism prevents misoperations caused by topology changes and guarantees the safe operation of the power grid.

[0098] In some embodiments, in the above step S103, the generating of the dynamic load shedding sequence according to the built-in programmable load characteristic database specifically includes:

[0099] Through the built-in programmable load characteristic database, the user importance level label, real-time power factor, and load type parameters of each load are called;

[0100] Based on the user importance level label, real-time power factor, and load type parameters, the score of each load is calculated through a priority scoring formula, and a dynamic load shedding sequence is generated in descending order of the score;

[0101] When it is detected that the score of the critical load is less than the preset score threshold, a safety alarm is triggered and the automatic load shedding is paused, waiting for manual confirmation;

[0102] Predict and update the label weights and dynamic coefficients of each load using an LSTM network, and send them to the terminal through the master station.

[0103] In this embodiment, in a power system, load shedding operations need to comprehensively consider the importance, power factor, and type of loads to avoid affecting critical loads. However, traditional load shedding methods are mostly based on fixed rules and lack the ability to dynamically adjust to load characteristics, easily leading to unreasonable load shedding strategies. The present invention realizes the dynamic optimization of load shedding strategies through a load characteristic database and machine learning prediction.

[0104] Exemplarily, a regional power grid includes multiple load nodes, including industrial loads (such as steel mills), commercial loads (such as shopping malls), and residential loads (such as residential areas). When the power grid detects overload, it is necessary to generate a dynamic load shedding sequence according to the load characteristics. Through the built-in programmable load characteristic database, call the user importance level labels (such as "high", "medium", "low"), real-time power factors (such as 0.85, 0.90, 0.95), and load type parameters (such as "industrial", "commercial", "residential") of each load. Calculate the priority score according to the user level (weight 0.5), power factor (weight 0.3), and load type (weight 0.2): score = user level weight × level value + power factor weight × power factor + load type weight × type value. Level value mapping: high = 3, medium = 2, low = 1; type value mapping: industrial = 3, commercial = 2, residential = 1. Among them, the score of load 1 = 0.5×3 + 0.3×0.95 + 0.2×3 = 2.835; the score of load 2 = 0.5×2 + 0.3×0.90 + 0.2×2 = 2.07; the score of load 3 = 0.5×1 + 0.3×0.85 + 0.2×1 = 1.305. Arrange in descending order of score to generate a dynamic load shedding sequence: load 1 → load 2 → load 3. If it is detected that the score of a critical load (such as load 1) is less than a preset score threshold (such as 2.5), trigger a safety alarm and suspend automatic load shedding. The operation and maintenance personnel confirm whether to continue load shedding or adjust the load shedding strategy through the master station interface. Use historical load data to train the LSTM network to predict the future change trend of load characteristics. For example, the predicted value of the future power factor of load 1 = 0.96, and the predicted value of the load type = industrial (remains unchanged). Update the label weights and dynamic coefficients of the load according to the prediction results (such as adjusting the user level weight of load 1 to 0.6, the power factor weight to 0.25, and the load type weight to 0.15), and send them to the terminal through the master station.

[0105] In this embodiment, through dynamic priority scoring, the importance of loads is accurately identified to avoid mis-shedding critical loads. The safety alarm mechanism prevents misoperation of critical loads and ensures the stable operation of the power grid. The LSTM prediction and parameter update mechanism enables the load shedding strategy to adapt to the dynamic changes of load characteristics and improves the load shedding effect.

[0106] In some embodiments, in the above step S103, based on the dynamic load shedding sequence, the load shedding operation is performed through a preset multi-level load shedding logic, and at the same time, the action signal is synchronously fed back to the master station, which specifically includes:

[0107] When the frequency drops and triggers the standard load shedding round, the load is shed according to the preset frequency threshold and load shedding ratio;

[0108] If the rate of change of frequency exceeds the dynamic threshold, activate the special load shedding round, calculate the dynamic load shedding amount through the slope sensitivity coefficient, and preferentially shed the interruptible load from the end of the dynamic load shedding sequence;

[0109] Package the time stamp, terminal ID, shed power, and average power factor of the load shedding action into a JSON message and synchronously feed it back to the master station through the MQTT protocol;

[0110] After receiving the multi-terminal data, the master station uses the LSTM network to predict the load shedding parameters in the next cycle.

[0111] In this embodiment, in the power system, when the frequency drops or the rate of change of frequency is abnormal, the load needs to be quickly shed to restore system stability. Traditional load shedding methods are mostly based on fixed rules and lack the ability to dynamically adjust the load characteristics, which easily leads to unreasonable load shedding strategies or mis-shedding of key loads. The present invention realizes the precise execution of the load shedding strategy through preset logic and dynamic parameter optimization.

[0112] Among them, the standard round (triggered by frequency threshold):

[0113] Preset 6 rounds of load shedding thresholds , and the load shedding amount for each round , where represents the load shedding amount in the m-th round of the standard round, represents the load shedding ratio coefficient of the standard round, represents the current total load power, represents the load shedding threshold in the m-th round of the standard round.

[0114] Among them, the special round (triggered by the rate of change of frequency):

[0115] Dynamically correct the load shedding amount according to the real-time rate of change of frequency , , where represents the load shedding amount in the n-th round of the special round, represents the slope sensitivity coefficient of the n-th round of the special round, represents the real-time rate of change of frequency, represents the instantaneous frequency within any sliding time window, represents the time step within any sliding time window.

[0116] Exemplarily, the grid frequency monitoring system detects that the frequency drops to 49.5 Hz (preset frequency threshold), triggering the standard load shedding round. According to the preset load shedding ratio (such as 10%), the amount of load to be shed is calculated. Based on the dynamic load shedding sequence, low-priority loads at the end of the sequence (such as residential loads) are preferentially shed. The grid frequency change rate monitoring system detects that the frequency change rate exceeds the dynamic threshold (such as 0.5 Hz / s), activating the special load shedding round. The dynamic load shedding amount is calculated through the slope sensitivity coefficient, and interruptible loads (such as commercial loads) are preferentially shed from the end of the dynamic load shedding sequence. The timestamp, terminal ID, shed power, and average power factor of the load shedding action are encapsulated into a JSON message, and the message is synchronously fed back to the master station through the MQTT protocol. The master station receives the load shedding data uploaded by multiple terminals to form a historical data set. The LSTM network is trained using the historical data set to predict the load shedding parameters for the next cycle (such as frequency threshold, load shedding ratio, slope sensitivity coefficient). The predicted load shedding parameters are sent from the master station to each terminal to achieve dynamic optimization of the load shedding strategy.

[0117] In this embodiment, through the multi-level load shedding logic and the dynamic load shedding sequence, low-priority loads are accurately identified and shed to avoid mis-shedding critical loads. The combination of the special load shedding round and the slope sensitivity coefficient quickly responds to abnormal frequency change rates, enhancing the stability of the power grid. The LSTM prediction and parameter update mechanism enables the load shedding strategy to adapt to changes in the power grid operation state, improving the load shedding effect.

[0118] In some embodiments, in step S104 above, when the master station calculates the marginal cost function of each terminal based on the game theory model, and after distributed consensus verification by combining blockchain technology, the load shedding parameters are optimized and sent for update, specifically including:

[0119] Regarding each terminal as a participant and taking the minimization of the total regional load shedding cost as the objective function, a non-cooperative game model is established;

[0120] Based on the non-cooperative game model, the master station iteratively solves the optimal load shedding amount and the unified marginal cost of each terminal through the KKT conditions;

[0121] Submitting the optimal load shedding amount and the unified marginal cost of each terminal to the blockchain network, and using the improved PBFT algorithm for distributed consensus verification. After the verification passes, a blockchain verification result is generated;

[0122] Based on the blockchain verification result, the load shedding ratio of the standard round and the slope sensitivity coefficient of the special round are dynamically corrected.

[0123] In this embodiment, in the load control of the power system, traditional load shedding methods are usually based on fixed rules or centralized optimization, lacking the ability to dynamically adjust the load characteristics of each terminal, and it is difficult to ensure the fairness and credibility of the load shedding parameters. The present invention realizes the dynamic optimization and reliable distribution of load shedding parameters through distributed game solving and blockchain consensus verification.

[0124] Specifically, the objective function satisfies , where represents the regional unified marginal cost, represents the load shedding amount of the i-th terminal, , and represent the cost coefficients fitted according to historical data, represents the critical load protection weight, represents the proportion of the critical load of the i-th terminal, represents the proportion of the real-time cuttable load of the i-th terminal.

[0125] The optimal load shedding amount of each terminal satisfies , where represents the optimal load shedding amount.

[0126] Exemplarily, a regional power grid includes 3 load terminals (Terminal A, Terminal B, and Terminal C), and the load characteristics of each terminal are different (such as power factor, load type, etc.). When the power grid experiences a sudden fault resulting in a frequency drop, load shedding is required to restore system stability. Regarding Terminal A, Terminal B, and Terminal C as game participants, the constraint conditions include: Total load shedding constraint: Load shedding of Terminal A + Load shedding of Terminal B + Load shedding of Terminal C = Total load shedding required by the power grid; Load shedding amount constraint for each terminal: Load shedding of Terminal A ≤ Maximum load shedding amount of Terminal A; Load shedding of Terminal B ≤ Maximum load shedding amount of Terminal B; Load shedding of Terminal C ≤ Maximum load shedding amount of Terminal C. Based on the non-cooperative game model, the master station iteratively solves the optimal load shedding amount and unified marginal cost of each terminal through the KKT conditions. Through iterative solution using the KKT conditions, the optimal load shedding amount and unified marginal cost of each terminal are obtained. For example, after iterative calculation, it is obtained that: Optimal load shedding amount of Terminal A = 80 kW; Optimal load shedding amount of Terminal B = 120 kW; Optimal load shedding amount of Terminal C = 50 kW; Unified marginal cost = 150 yuan / kW. The optimal load shedding amount and unified marginal cost of each terminal are submitted to the blockchain network. The improved PBFT algorithm is used for distributed consensus verification. After the verification passes, a blockchain verification result is generated. For example, multiple nodes in the blockchain network verify the submitted data, and after confirming the consistency and validity of the data, a blockchain verification result is generated. Based on the blockchain verification result, the load shedding ratio in the standard round and the slope sensitivity coefficient in the special round are dynamically corrected. For example, according to the verification result, the load shedding strategy is adjusted: The load shedding ratio in the standard round is adjusted from 10% to 12%, and the slope sensitivity coefficient in the special round is adjusted from 0.8 to 0.85. The corrected load shedding parameters are sent to each terminal. For example, the master station sends the updated load shedding parameters to each terminal: Terminal A: Load shedding ratio in the standard round 12%, slope sensitivity coefficient in the special round 0.85; Terminal B: Load shedding ratio in the standard round 12%, slope sensitivity coefficient in the special round 0.85; Terminal C: Load shedding ratio in the standard round 12%, slope sensitivity coefficient in the special round 0.85.

[0127] In this embodiment, through the game theory model, the optimal load shedding amount of each terminal is dynamically solved to minimize the total regional load shedding cost. The blockchain consensus verification mechanism ensures the fairness and credibility of the load shedding parameters, avoiding decision-making biases of a single centralized node. Dynamically correcting the load shedding ratio and slope sensitivity coefficient enables the load shedding strategy to adapt to changes in the power grid operation state and improves the load shedding effect. The immutability and distributed consensus mechanism of blockchain technology enhance the security and reliability of the load shedding parameter distribution.

[0128] In some embodiments, in the above step S105, after the load shedding is completed, the fault section is located by residual voltage detection and direction discrimination, and a personalized reconstruction plan is generated by combining topology recognition and fault type matching to restore power supply to the non-fault section within a preset distance, specifically including:

[0129] After the load shedding is completed, detect the residual voltage of each node in the power grid, mark the nodes with residual voltage lower than the preset residual voltage threshold as suspected fault nodes, and construct a preliminary screening section set;

[0130] For each section in the preliminary screening section set, calculate the phase angle difference and current ratio between the head and the end. If the phase angle difference between the head and the end is greater than the preset phase angle difference threshold and the current ratio is greater than the preset current ratio threshold, it is determined as a forward fault section, and a fault type matching result is obtained;

[0131] Based on the topology database, call the Dijkstra algorithm to generate the shortest power supply path from the non-fault section to the power source, select a reconstruction strategy in combination with the fault type matching result to generate a personalized reconstruction plan, and at the same time restore the power supply of the non-fault section according to the shortest power supply path from the non-fault section to the power source;

[0132] After the reconstruction is completed, send a restoration confirmation signal to the master station through GOOSE communication and update the topology database.

[0133] In this embodiment, traditional power grid fault restoration methods usually rely on global topology analysis or fixed rules, lack the ability to accurately locate fault sections and personalized restoration of non-fault sections, and are difficult to adapt to complex power grid topology changes. The present invention realizes fast power supply restoration of non-fault sections through dynamic fault location and topology optimization.

[0134] Exemplarily, a regional power grid includes 5 nodes (Node A, Node B, Node C, Node D, Node E) and 4 transmission lines (Line 1-2, Line 2-3, Line 3-4, Line 4-5). A short circuit fault occurs on Line 2-3, resulting in power loss in some sections of the power grid.

[0135] After the load shedding is completed, detect the residual voltage of each node. Set the preset residual voltage threshold to 0.6 times the rated voltage. The detection results are as follows: for Node A, the residual voltage is 0.95 times the rated voltage; for Node B, the residual voltage is 0.55 times the rated voltage; for Node C, the residual voltage is 0.40 times the rated voltage; for Node D, the residual voltage is 0.85 times the rated voltage; for Node E, the residual voltage is 0.90 times the rated voltage.

[0136] Mark Node B and Node C with residual voltage lower than the preset residual voltage threshold as suspected fault nodes, and construct a preliminary screening section set: Line 1-2 (including Node A and Node B), Line 2-3 (including Node B and Node C), Line 3-4 (including Node C and Node D).

[0137] For each section in the initial screening section set, calculate the phase angle difference and current ratio between the start and end points. Set the preset phase angle difference threshold to 30 degrees and the preset current ratio threshold to 1.2. The calculated results for line 1-2 are: phase angle difference between start and end points is 15 degrees, current ratio is 1.05; for line 2-3: phase angle difference between start and end points is 45 degrees, current ratio is 1.5; for line 3-4: phase angle difference between start and end points is 10 degrees, current ratio is 0.95. If the phase angle difference between the start and end points is greater than the preset phase angle difference threshold and the current ratio is greater than the preset current ratio threshold, it is determined as a positive fault section. The determination result is: for line 2-3: determined as a positive fault section; fault type matching result: short circuit fault. Based on the topology database, call the Dijkstra algorithm to generate the shortest power supply path from the non-fault section to the power source.

[0138] Assume that node A is the power source node, and the non-fault sections are node D and node E. The shortest power supply paths are: for node D: the shortest power supply path: node A → line 1-2 → line 2-3 (skipped) → line 3-4 → node D; for node E: the shortest power supply path: node A → line 1-2 → line 2-3 (skipped) → line 3-4 → line 4-5 → node E.

[0139] During actual reconstruction, skip the fault section line 2-3 and select alternative paths: for node D's alternative power supply path: node A → line 1-2 → node B (through other non-fault paths or standby lines) → node D; for node E's alternative power supply path: node A → line 1-2 → node B (through other non-fault paths or standby lines) → node D → line 4-5 → node E; (or directly connect node A and node E through other non-fault paths).

[0140] Combined with the fault type matching result (short circuit fault), select the reconstruction strategy to isolate the fault section and restore power supply to the non-fault section. Obtain reconstruction plan 1: isolate line 2-3 and restore power supply to node D and node E; reconstruction plan 2: if there are standby lines, enable the standby lines to connect node A with node D and node E.

[0141] After reconstruction, send a recovery confirmation signal to the master station through GOOSE communication to confirm that node D and node E have restored power supply. Update the topology database, mark the fault section line 2-3 as the fault status, and record the reconstructed power grid topology. The updated topology database includes: node A: connected to line 1-2; node B: connected to line 1-2 (upstream of the fault section), line 2-3 (fault section, isolated); node C: connected to line 2-3 (fault section, isolated), line 3-4; node D: connected to line 3-4, line 4-5; node E: connected to line 4-5.

[0142] In this embodiment, through residual voltage detection and direction discrimination, the faulty section is quickly located, reducing the fault troubleshooting time. Based on topology recognition and the generation of the shortest power supply path, the power supply to the non-faulty section is quickly restored, reducing the power outage time. Combining the fault type matching results, the optimal reconstruction strategy is selected to improve the adaptability and reliability of the reconstruction plan. Through GOOSE communication and topology update, the real-time nature and accuracy of the reconstruction information are ensured, enhancing the safety and stability of the power grid operation.

[0143] Referring to Figure 2 , an embodiment of the present invention provides a low-frequency load shedding collaborative control system 2 based on GOOSE communication. The system 2 specifically includes:

[0144] A hardware frequency measurement module 201, which is used to calculate the frequency change rate by real-time monitoring of the frequency, voltage, and phase angle parameters of the power system. When the frequency change rate exceeds a preset change rate threshold, it triggers a frequency protection thread;

[0145] A collaborative control module 202, which is used to send a status query request to adjacent terminals after calling the topology analysis service to confirm that the local end is in the receiving-end power grid, receive the action signals of adjacent terminals through GOOSE communication. If at least two adjacent terminal action signals are received, the local load shedding action is delayed and topology verification is started; otherwise, it directly enters the load shedding execution stage;

[0146] A load priority decision module 203, which is used to generate a dynamic load shedding sequence based on the built-in programmable load characteristic database, execute load shedding operations based on the dynamic load shedding sequence through a preset multi-level load shedding logic, and simultaneously synchronously feedback the action signals to the master station;

[0147] A master station linkage module 204, which is used to calculate the marginal cost function of each terminal based on the game theory model at the master station, optimize the load shedding parameters after distributed consensus verification in combination with blockchain technology, and issue updates;

[0148] A self-healing reconstruction module 205, which is used to locate the faulty section through residual voltage detection and direction discrimination after load shedding, generate a personalized reconstruction plan in combination with topology recognition and fault type matching, and restore the power supply to the non-faulty section within a preset distance.

[0149] It can be understood that the content in the embodiment of the low-frequency load shedding collaborative control method based on GOOSE communication as Figure 1 shown is applicable to the embodiment of this low-frequency load shedding collaborative control system based on GOOSE communication. The functions specifically implemented by this embodiment of the low-frequency load shedding collaborative control system based on GOOSE communication are the same as those in the embodiment of the low-frequency load shedding collaborative control method as Figure 1 shown, and the beneficial effects achieved are the same as those in Figure 1The beneficial effects achieved by the embodiments of the low-frequency load shedding cooperative control method based on GOOSE communication shown are the same.

[0150] It should be noted that for the information interaction, execution process, etc. between the above systems, since they are based on the same concept as the method embodiments of the present invention, their specific functions and the technical effects brought can be specifically referred to in the method embodiment part, and will not be elaborated here.

[0151] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiment and will not be elaborated here.

[0152] Referring to Figure 3 , an embodiment of the present invention further provides a computer device 3, including: a memory 302, a processor 301, and a computer program 303 stored on the memory 302. When the computer program 303 is executed on the processor 301, the low-frequency load shedding cooperative control method based on GOOSE communication as described in any one of the above methods is implemented.

[0153] The computer device 3 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art can understand that Figure 3 merely an example of the computer device 3 does not constitute a limitation to the computer device 3, and may include more or fewer components than shown, or combine some components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0154] The so-called processor 301 may be a Central Processing Unit (CPU), and this processor 301 may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0155] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as the hard disk or memory of the computer device 3. In some other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc., equipped on the computer device 3. Further, the memory 302 may also include both the internal storage unit and the external storage device of the computer device 3. The memory 302 is used to store an operating system, application programs, a Boot Loader, data, and other programs, such as the program code of the computer program, etc. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0156] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it implements the low-frequency load shedding collaborative control method based on GOOSE communication as described in any one of the above methods.

[0157] In this embodiment, when the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above embodiment methods of this application, a computer program can be used to instruct relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the photographing device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0158] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0159] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0160] In the embodiments disclosed in this application, it should be understood that the disclosed device / terminal device and method can be implemented in other ways. For example, the device / terminal device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in an electrical, mechanical or other form.

[0161] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Claims

1. A low-frequency load shedding cooperative control method based on GOOSE communication, characterized in that The method specifically includes: By real-time monitoring the frequency, voltage and phase angle parameters of the power system, calculating the frequency change rate, when the frequency change rate exceeds the preset change rate threshold, triggering the frequency protection thread; After calling the topology analysis service to confirm that the local end is in the receiving-end power grid, sending a status query request to adjacent terminals, receiving the action signals of adjacent terminals through GOOSE communication. If at least two action signals of adjacent terminals are received, delaying the local load shedding action and starting topology verification, otherwise directly entering the load shedding execution stage; Generating a dynamic load shedding sequence according to the built-in programmable load characteristic database, and based on the dynamic load shedding sequence, performing load shedding operations through the preset multi-level load shedding logic, and at the same time synchronously feeding back the action signals to the master station; At the master station, calculating the marginal cost function of each terminal based on the game theory model, optimizing the load shedding parameters after distributed consensus verification combined with blockchain technology and issuing updates; After load shedding, locating the fault section through residual voltage detection and direction discrimination, generating a personalized reconstruction plan by combining topology recognition and fault type matching, and restoring power supply to non-fault sections within a preset distance.

2. The method according to claim 1, characterized in that, The step of by real-time monitoring the frequency, voltage and phase angle parameters of the power system, calculating the frequency change rate, when the frequency change rate exceeds the preset change rate threshold, triggering the frequency protection thread specifically includes: Real-time collecting the three-phase voltage signal parameters of the power system, where the three-phase voltage signal parameters include frequency, voltage and phase angle; Based on the three-phase voltage signal parameters, calculating the instantaneous frequency by the equal-precision frequency measurement method; Based on the instantaneous frequency, obtaining the frequency change rate through differential calculation within multiple sliding time windows; Dynamically adjusting the change rate threshold according to the system inertia constant and new energy penetration rate. When it is detected that the frequency change rate is greater than the change rate threshold, immediately triggering the frequency protection thread.

3. The method according to claim 1, wherein The step of after calling the topology analysis service to confirm that the local end is in the receiving-end power grid, sending a status query request to adjacent terminals, receiving the action signals of adjacent terminals through GOOSE communication. If at least two action signals of adjacent terminals are received, delaying the local load shedding action and starting topology verification, otherwise directly entering the load shedding execution stage specifically includes: Calling the topology analysis service to construct an equivalent power grid model based on the node admittance matrix and real-time measurement data; By solving the equivalent power grid model, judging whether the local end is the receiving-end power grid. If it is the receiving-end power grid, sending a status query request in GOOSE format to adjacent terminals, where the status query request includes terminal ID, timestamp and action flag; Receiving the action signals of adjacent terminals through GOOSE communication, counting the number of actions. If the number of actions is greater than or equal to 2, delaying the local load shedding action; During the delay stage, calculating the topology deviation rate based on real-time measurement data. If the topology deviation rate is less than the preset deviation rate threshold, performing load shedding operations, otherwise terminating the action and reporting to the master station.

4. The method according to claim 1, wherein The step of generating a dynamic load shedding sequence according to the built-in programmable load characteristic database specifically includes: Through the built-in programmable load characteristic database, calling the user importance level label, real-time power factor and load type parameters of each load; Based on the user importance level tags, real-time power factor, and load type parameters, calculate the scores of each load through the priority scoring formula, and generate a dynamic load shedding sequence in descending order of scores; When it is detected that the score of the critical load is less than the preset score threshold, trigger a safety alarm and pause the automatic load shedding, waiting for manual confirmation; Use the LSTM network to predict and update the label weights and dynamic coefficients of each load, and send them to the terminal through the master station.

5. The method according to claim 4, wherein Based on the dynamic load shedding sequence, execute the load shedding operation through the preset multi-level load shedding logic, and at the same time synchronously feedback the action signal to the master station, specifically including: When the frequency drops and triggers the standard load shedding round, cut the load according to the preset frequency threshold and load shedding ratio; If the rate of change of frequency exceeds the dynamic threshold, activate the special load shedding round, calculate the dynamic load shedding amount through the slope sensitivity coefficient, and preferentially cut the interruptible load from the end of the dynamic load shedding sequence; Package the timestamp, terminal ID, cut power, and average power factor of the load shedding action into a JSON message, and synchronously feedback it to the master station through the MQTT protocol; After the master station receives the data of multiple terminals, use the LSTM network to predict the load shedding parameters for the next cycle.

6. The method according to claim 5, wherein Based on the game theory model at the master station, calculate the marginal cost function of each terminal, and optimize and issue updates to the load shedding parameters after distributed consensus verification by combining blockchain technology, specifically including: Regard each terminal as a participant, take minimizing the total regional load shedding cost as the objective function, and establish a non-cooperative game model; Based on the non-cooperative game model, the master station iteratively solves the optimal load shedding amount and unified marginal cost of each terminal through the KKT conditions; Submit the optimal load shedding amount and unified marginal cost of each terminal to the blockchain network, and use the improved PBFT algorithm for distributed consensus verification. After the verification passes, generate the blockchain verification result; Based on the blockchain verification result, dynamically correct the load shedding ratio of the standard round and the slope sensitivity coefficient of the special round.

7. The method according to claim 1, characterized in that, After the load shedding is completed, locate the fault section through residual voltage detection and direction discrimination, and generate a personalized reconstruction plan by combining topology recognition and fault type matching, and restore the power supply of the non-fault section within the preset distance, specifically including: After the load shedding is completed, detect the residual voltage of each node in the power grid, mark the nodes with residual voltage lower than the preset residual voltage threshold as suspected fault nodes, and construct a preliminary screening section set; For each section in the preliminary screening section set, calculate the phase angle difference and current ratio between the head and the end. If the phase angle difference between the head and the end is greater than the preset phase angle difference threshold and the current ratio is greater than the preset current ratio threshold, it is determined as a positive fault section, and the fault type matching result is obtained; Based on the topology database, call the Dijkstra algorithm to generate the shortest power supply path from the non-fault section to the power source, and select the reconstruction strategy in combination with the fault type matching result to generate a personalized reconstruction plan, and at the same time restore the power supply of the non-fault section according to the shortest power supply path from the non-fault section to the power source; After the reconstruction is completed, send a restoration confirmation signal to the master station through GOOSE communication and update the topology database.

8. A low-frequency load shedding collaborative control system based on GOOSE communication, characterized in that, The system specifically includes: A hardware frequency measurement module, which is used to calculate the rate of change of frequency by real-time monitoring the frequency, voltage, and phase angle parameters of the power system. When the rate of change of frequency exceeds the preset rate of change threshold, trigger the frequency protection thread; The coordinated control module is used to call the topology analysis service to confirm that the local side is in the receiving-end power grid, then send a status query request to adjacent terminals, receive the action signals of adjacent terminals through GOOSE communication. If at least two action signals of adjacent terminals are received, the local load shedding action is delayed and topology verification is started; otherwise, it directly enters the load shedding execution stage. The load priority decision module is used to generate a dynamic load shedding sequence according to the built-in programmable load characteristic database, and based on the dynamic load shedding sequence, execute the load shedding operation through the preset multi-level load shedding logic, and at the same time synchronously feedback the action signal to the master station. The master station linkage module is used to calculate the marginal cost function of each terminal based on the game theory model at the master station, optimize the load shedding parameters and issue updates after distributed consensus verification by combining blockchain technology. The self-healing reconstruction module is used to locate the fault section through residual voltage detection and direction discrimination after load shedding is completed, generate a personalized reconstruction plan by combining topology identification and fault type matching, and restore the power supply of non-fault sections within a preset distance.

9. A computer device, characterized in that, Including: A memory, a processor, and a computer program stored on the memory. When the computer program is executed on the processor, it implements the low-frequency load shedding coordinated control method based on GOOSE communication as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored thereon. When the computer program is run by the processor, it implements the low-frequency load shedding coordinated control method based on GOOSE communication as described in any one of claims 1 to 7.

Images