High-voltage distribution network self-healing strategy generation method, system and device and storage medium

Abstract

The invention discloses a high-voltage power distribution network self-healing strategy generation method, system and device and a storage medium, and relates to the technical field of power systems, and the method comprises the steps: deploying a multi-source data collection module, and collecting the voltage, current and switch state information of each node of a power grid in real time; and constructing a rule base containing a basic power recovery rule and a dynamic adjustment mechanism. According to the method, a complex multi-fault scene is decomposed into a quick response layer, an optimization adjustment layer and a global optimization layer through a hierarchical progressive strategy generation mechanism, the quick response layer realizes second-level preliminary treatment based on a predefined emergency rule, and delay caused by serial calculation in a traditional method is avoided; the optimization adjustment layer quickly eliminates local overload and topology conflicts; and the global optimization layer is used for disassembling a large-scale problem into multiple sub-problems for synchronous solving through parallelization processing, so that the solving time is greatly shortened.

Description

Technical Field

[0001] This invention relates to the technical field of power systems, and in particular to a method, system, device, and storage medium for generating self-healing strategies for high-voltage distribution networks. Background Technology

[0002] Traditional power grids often require manual inspection and repair when faults occur, resulting in long power outages and wide-ranging impacts. With the development of smart grid technology, "self-healing" has become one of the core objectives for improving power grid reliability. Main grid self-healing refers to the power system's ability to quickly identify and locate faults in the main grid (such as high-voltage transmission lines, substations, and other critical facilities) during operation, automatically isolate faulty areas, and restore power supply to non-faulty areas with little or no manual intervention, ensuring the safe, stable, and reliable operation of the power grid. Main grid self-healing strategies include: when a high-voltage transmission line experiences a short circuit due to lightning strikes, tree obstructions, etc., the self-healing system can automatically disconnect the faulty line and switch to a backup line for power supply; if a transformer or switching equipment in a substation fails, the system can quickly isolate the faulty equipment and activate backup equipment to maintain power supply; in disasters such as typhoons and earthquakes, the main grid self-healing system can quickly identify damaged areas and prioritize restoring power to important loads (such as emergency command centers and communication base stations). The core of generating a self-healing strategy for the main grid is to combine the real-time status of the power grid, fault characteristics, and operational constraints to achieve automated and intelligent generation of the strategy.

[0003] Existing methods for generating self-healing strategies for the main grid suffer from the following problems: Limited ability to handle multiple concurrent faults: Existing strategy generation methods based on rules or single optimization algorithms, such as MIP and genetic algorithms, often experience exponential increases in computational complexity when faced with multiple faults, such as simultaneous tripping of multiple lines or cascading equipment faults, leading to delays in strategy generation. For example, the traditional MIP model may take more than 10 seconds to solve multiple faults in a 500kV power grid, failing to meet the requirement of second-level self-healing. Difficulty in identifying new fault modes: AI models trained on historical data lack the ability to generalize to unseen fault modes, such as faults caused by new equipment defects or network attacks. When the power grid suffers from new malware attacks that cause switch malfunctions, RL agents relying on historical fault samples may not be able to generate effective strategies. Summary of the Invention

[0004] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a method, system, device, and storage medium for generating self-healing strategies for high-voltage distribution networks to address the problems of limited concurrent multi-fault processing capabilities and difficulty in identifying novel fault modes in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, embodiments of the present invention provide a method for generating a self-healing strategy for a high-voltage distribution network, comprising:

[0007] Real-time collection of voltage, current, and switch status information at various nodes of the power grid to obtain power grid fault information;

[0008] Based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, the priority and trigger threshold of each rule are dynamically adjusted to form a dynamic power restoration rule base.

[0009] A hierarchical progressive strategy generation mechanism is adopted to generate a power restoration strategy. The hierarchical progressive strategy generation mechanism initiates emergency handling based on the dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer.

[0010] The generated power restoration strategy is verified. If the verification is successful, self-healing control is executed, and the actual power restoration effect is fed back to the dynamic power restoration rule base to update the rule parameters, forming a closed-loop optimization system.

[0011] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method described in this invention, the method of dynamically adjusting the priority and trigger threshold of each rule by combining real-time power grid state variables and historical fault handling feedback includes: updating the dynamic priority and trigger threshold of each basic power restoration rule online based on the real-time power grid state variable set and historical fault handling feedback.

[0012] The state variables include load level, equipment health, and topology connectivity. After each rule execution, the weight coefficients and historical execution effect parameters are updated based on the actual power restoration effect and the forgetting factor, thus achieving adaptive evolution of the rule base.

[0013] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method described in this invention, a basic power restoration rule set is constructed based on the basic constraints of power grid operation and empirical power restoration principles. The basic power restoration rules include: prioritizing the restoration of important loads, avoiding equipment overload, and maintaining the integrity of the power grid topology.

[0014] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method described in this invention, the following is an online update of the dynamic priority and trigger threshold of each basic power restoration rule based on the real-time power grid state variable set and historical fault handling feedback: defining a real-time power grid state variable set, with each variable normalized to the [0,1] interval; the initial priority of each basic power restoration rule is set by expert experience, and its dynamic priority is corrected by weighted adjustment based on the real-time operating status of the power grid.

[0015] The trigger threshold of the basic power restoration rules is dynamically adjusted based on historical execution results and real-time risk coefficients. When the historical execution success rate of any basic power restoration rule is lower than the average level and the current risk is high, the system automatically raises its trigger threshold to prevent accidental activation; conversely, it lowers the threshold to improve response sensitivity.

[0016] The beneficial effects of this preferred technical solution are as follows: It enhances generalization capability through a dynamic power restoration rule base and feedback loop mechanism. The rule base reserves interfaces for fuzzy logic and expert systems to perform preliminary classification and strategy matching for unknown fault modes. Simultaneously, the historical data module records the characteristics and handling effects of unidentified faults in real time, continuously optimizing rule weights or generating new rules through reinforcement learning. When the power grid suffers a new type of malware attack, the system can quickly activate backup lines based on fuzzy rules and incorporate the fault mode into the rule base through subsequent feedback loops, avoiding the limitations of relying on historical samples and significantly improving adaptability to unknown faults.

[0017] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method described in this invention, the hierarchical progressive strategy generation mechanism initiates emergency handling based on a dynamic power restoration rule library at the rapid response layer, including: responding immediately to severe faults based on predefined emergency power restoration rules and initiating basic power restoration strategies.

[0018] When a sudden drop in bus voltage and loss of topology are detected, immediately disconnect the non-critical load switch on the de-energized bus; when the main power supply line fails, prioritize closing the backup power supply line circuit breaker to restore power to important loads; when a line or transformer experiences a momentary overload due to a fault, quickly disconnect the secondary load branch; identify equipment protection action signals and immediately isolate the fault area to prevent cascading tripping.

[0019] The beneficial effects of this preferred technical solution are as follows: through a hierarchical progressive strategy generation mechanism, complex multi-fault scenarios are decomposed into a rapid response layer, an optimization and adjustment layer, and a global optimization layer. The rapid response layer achieves preliminary handling within seconds based on predefined emergency rules, avoiding the delay caused by serial calculation in traditional methods. The optimization and adjustment layer quickly eliminates local overload and topology conflicts. The global optimization layer decomposes large-scale problems into multiple sub-problems for simultaneous solution through parallel processing, significantly shortening the solution time.

[0020] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method described in this invention, the verification of the generated power restoration strategy includes: taking minimizing load loss, minimizing equipment overload, and minimizing voltage deviation as objective functions, and taking equipment capacity constraints, voltage stability constraints, topology connectivity constraints, switching action number constraints, and economic constraints as constraints, and performing comprehensive optimization to obtain the optimal power restoration strategy.

[0021] The objective function is expressed as:

[0022] min(α·Load_Loss+β·Overload_Severity+γ·Voltage_Deviation)

[0023] Where α, β, and γ are weighting coefficients, and α+β+γ=1, Load_Loss is the unrecovered load, Overload_Severity is the degree of equipment overload, and Voltage_Deviation is the bus voltage deviation.

[0024] As a preferred embodiment of the high-voltage distribution network self-healing strategy generation method of the present invention, it further includes: the equipment capacity constraint is expressed as:

[0025]

[0026] Among them, P i Let be the active power of line i. Let i be the rated capacity of line i;

[0027] The voltage stability constraint is expressed as follows:

[0028]

[0029] Among them, V j V is the voltage amplitude at bus j. j min V j max Within the permitted range;

[0030] The topological connectivity constraint is expressed as follows:

[0031] Graph(G)is connected and acyclic(or with controlledloops)

[0032] Where Graph(G) is the topology graph of the power grid, connected means that there is a path between any two nodes, and acyclic means that the power grid should maintain a radial structure;

[0033] The constraint on the number of switching actions is expressed as follows:

[0034]

[0035] Where, ΔS k Let k be the number of times the switch is operated, and Switches be the set of all circuit breakers and disconnectors;

[0036] The economic constraints are expressed as follows:

[0037]

[0038] Among them, C l Let x be the operation cost of path l. l Choose a variable for the path; Paths is the set of all possible reconnection paths.

[0039] Secondly, the present invention provides a high-voltage distribution network self-healing strategy generation system, comprising:

[0040] The data acquisition module is used to collect voltage, current and switch status information of each node in the power grid in real time, and to obtain power grid fault information.

[0041] The dynamic power restoration rule base construction module is used to dynamically adjust the priority and trigger threshold of each rule based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, so as to form a dynamic power restoration rule base.

[0042] The layered progressive strategy generation module is used to generate a power restoration strategy using a layered progressive strategy generation mechanism. The layered progressive strategy generation mechanism initiates emergency handling based on the dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer.

[0043] The strategy verification and execution optimization module is used to verify the generated power restoration strategy. After the verification is successful, self-healing control is executed, and the actual power restoration effect is fed back to the dynamic power restoration rule base to update the rule parameters and form a closed-loop optimization system.

[0044] Thirdly, the present invention provides an electronic device, comprising:

[0045] Memory and processor;

[0046] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the high-voltage distribution network self-healing strategy generation method.

[0047] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the high-voltage distribution network self-healing strategy generation method.

[0048] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention decomposes complex multi-fault scenarios into a fast response layer, an optimization and adjustment layer, and a global optimization layer through a hierarchical progressive strategy generation mechanism. The fast response layer achieves preliminary handling within seconds based on predefined emergency rules, avoiding the delay caused by serial calculation in traditional methods. The optimization and adjustment layer quickly eliminates local overload and topology conflicts. The global optimization layer decomposes large-scale problems into multiple sub-problems for simultaneous solution through parallel processing, significantly shortening the solution time.

[0049] This invention enhances generalization capabilities through a dynamic power restoration rule base and a feedback loop mechanism. The rule base includes interfaces for fuzzy logic and expert systems, enabling preliminary classification and strategy matching of unknown fault modes. Simultaneously, a historical data module records the characteristics and handling effects of unidentified faults in real time, continuously optimizing rule weights or generating new rules through reinforcement learning. When the power grid is attacked by novel malicious software, the system can quickly activate backup lines based on fuzzy rules and incorporate the fault mode into the rule base through subsequent feedback loops, avoiding the limitations of relying on historical samples and significantly improving adaptability to unknown faults. Attached Figure Description

[0050] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0051] Figure 1 This is a schematic diagram of a method flow diagram for generating a self-healing strategy for a high-voltage distribution network according to an embodiment of the present invention;

[0052] Figure 2 This is a flowchart illustrating the correction algorithm of a high-voltage distribution network self-healing strategy generation method according to an embodiment of the present invention. Detailed Implementation

[0053] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0054] Example 1 (refer to) Figures 1-2 This is one embodiment of the present invention, which provides a method for generating a self-healing strategy for a high-voltage distribution network, comprising:

[0055] S100: Real-time acquisition of voltage, current and switch status information of each node in the power grid, and acquisition of power grid fault information;

[0056] S200: Based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, the priority and trigger threshold of each rule are dynamically adjusted to form a dynamic power restoration rule library.

[0057] S300: The layered progressive strategy generation mechanism generates the power restoration strategy. The layered progressive strategy generation mechanism initiates emergency handling based on the dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer.

[0058] S400: Verify the generated power restoration strategy. If the verification is successful, execute self-healing control and feed back the actual power restoration effect to the dynamic power restoration rule base to update the rule parameters and form a closed-loop optimization system.

[0059] It should be noted that existing methods for generating self-healing strategies for the main grid have the following problems: Limited ability to handle multiple concurrent faults: Existing strategy generation methods based on rules or single optimization algorithms, such as MIP and genetic algorithms, often suffer from exponential increases in computational complexity when facing multiple faults, such as simultaneous tripping of multiple lines or cascading equipment faults, leading to delays in strategy generation. For example, the traditional MIP model may take more than 10 seconds to solve multiple faults in a 500kV power grid, failing to meet the requirement of second-level self-healing; Difficulty in identifying new fault modes: AI models trained on historical data lack the ability to generalize to unseen fault modes, such as faults caused by new equipment defects or network attacks. When the power grid suffers from new malware attacks that cause switch malfunctions, RL agents relying on historical fault samples may not be able to generate effective strategies. This invention deploys a multi-source data acquisition module to collect voltage, current, and switch status information of each node in the power grid in real time; and constructs a rule base that includes basic power restoration rules and dynamic adjustment mechanisms. This invention uses a hierarchical progressive strategy generation mechanism to decompose complex multi-fault scenarios into a rapid response layer, an optimization and adjustment layer, and a global optimization layer. The rapid response layer achieves preliminary handling within seconds based on predefined emergency rules, avoiding the delays caused by serial computation in traditional methods. The optimization and adjustment layer quickly eliminates local overload and topology conflicts. The global optimization layer decomposes large-scale problems into multiple sub-problems for simultaneous solution through parallel processing, significantly shortening the solution time.

[0060] In this embodiment of the invention, in step S100, data cleaning technology is used to remove noise and outliers from the data and to initially extract fault features; the fault features include the time, location and type of the fault.

[0061] In this embodiment of the invention, step S200, which combines real-time power grid state variables and historical fault handling feedback to dynamically adjust the priority and trigger threshold of each rule, includes: updating the dynamic priority and trigger threshold of each basic power restoration rule online based on the real-time power grid state variable set and historical fault handling feedback.

[0062] The state variables include load level, equipment health, and topology connectivity. After each rule execution, the weight coefficients and historical execution effect parameters are updated based on the actual power restoration effect and the forgetting factor, thus achieving adaptive evolution of the rule base.

[0063] In this embodiment of the invention, in step S200, a basic power restoration rule set is constructed based on the basic constraints of power grid operation and empirical power restoration principles. The basic power restoration rules include: prioritizing the restoration of important loads, avoiding equipment overload, and maintaining the integrity of the power grid topology.

[0064] It should be noted that, in step S200, constructing the dynamic power restoration rule base requires first sorting out the basic constraints and empirical power restoration principles of power grid operation to form a basic rule set. Simultaneously, by accessing real-time power grid status data and historical fault handling feedback, a correction algorithm is used to dynamically correct the priority and trigger thresholds of the basic rules, and fuzzy matching rule templates for unknown fault modes are reserved. The correction algorithm achieves second-level response through state variables and risk coefficients; the feedback loop continuously optimizes weights and thresholds, improving the generalization ability for unknown faults; both weight and threshold adjustments are based on variables with clear physical meaning, facilitating dispatcher understanding and intervention.

[0065] In this embodiment of the invention, step S200, which updates the dynamic priority and trigger threshold of each basic power restoration rule online based on the real-time power grid state variable set and historical fault handling feedback, includes: defining a real-time power grid state variable set, with each variable normalized to the [0,1] interval; the initial priority of each basic power restoration rule is set by expert experience, and its dynamic priority is corrected by weighting the real-time operating status of the power grid.

[0066] The trigger threshold of the basic power restoration rules is dynamically adjusted based on historical execution results and real-time risk coefficients. When the historical execution success rate of any basic power restoration rule is lower than the average level and the current risk is high, the system automatically raises its trigger threshold to prevent accidental activation; conversely, it lowers the threshold to improve response sensitivity.

[0067] Specifically, dynamic priority is represented by real-time state weighted correction as follows:

[0068]

[0069] Among them, P j (t) represents the dynamic priority. The initial priority of the basic power restoration rules, wi The weights of the state variables on the rules are 0 ≤ w i ≤1, where n is the number of state variables.

[0070] Specifically, the trigger threshold for the basic power restoration rules is dynamically adjusted based on historical execution results and real-time risk coefficients, as follows:

[0071]

[0072] Among them, T j 0 is the initial threshold for the basic complexing rule, α1 and β1 are adjustment factors, and E avg The average execution effect of similar rules is given by λ(t), which is the risk coefficient calculated from the real-time fault severity. j This is based on the historical implementation results of the basic power restoration rules.

[0073] Furthermore, after each rule execution, the weight coefficients and historical execution effect parameters are updated based on the actual restoration effect and the forgetting factor, thus achieving adaptive evolution of the rule base, as shown below:

[0074]

[0075] Where γ and δ are forgetting factors and 0 < γ, δ < 1, This section describes the historical execution performance of the basic power restoration rules prior to this update. Based on the historical execution results after the basic power restoration rules were updated. State variable S i The weights prior to this update State variable S i Updated weights.

[0076] In this embodiment of the invention, the hierarchical progressive strategy generation mechanism in step S300, which initiates emergency response based on a dynamic power restoration rule base at the rapid response layer, includes: responding immediately to severe faults based on predefined emergency power restoration rules and initiating a basic power restoration strategy;

[0077] When a sudden drop in bus voltage and loss of topology are detected, immediately disconnect the non-critical load switch on the de-energized bus; when the main power supply line fails, prioritize closing the backup power supply line circuit breaker to restore power to important loads; when a line or transformer experiences a momentary overload due to a fault, quickly disconnect the secondary load branch; identify equipment protection action signals and immediately isolate the fault area to prevent cascading tripping.

[0078] It should be noted that the hierarchical progressive strategy generation mechanism divides the strategy generation mechanism into a fast response layer, an optimization adjustment layer, and a global optimization layer. Through parallel computing technology, it decomposes large-scale optimization problems into multiple sub-problems and solves them simultaneously, thus shortening the solution time.

[0079] Specifically, the rapid response layer responds immediately to severe faults and initiates basic power restoration measures based on predefined emergency power restoration rules; the optimization and adjustment layer uses a local search algorithm to perform preliminary optimization based on the rapid response, in order to reduce equipment overload and unreasonable topology problems during the power restoration process; and the global optimization layer uses parallelized mixed integer programming to optimize the global power restoration strategy for complex multi-fault scenarios.

[0080] In an optional embodiment, the fast response layer, optimization adjustment layer, and global optimization layer of the hierarchical progressive strategy generation mechanism achieve self-healing responses at the second to minute level through phased execution. The specific rules and execution timing are as follows:

[0081] The emergency power restoration rules of the fast response layer focus on handling severe power grid faults within seconds. The core rules include:

[0082] When a sudden drop in bus voltage and loss of topology are detected, immediately disconnect the non-critical load switches on the de-energized bus to prevent the fault from spreading; if the main power supply line is faulty, prioritize closing the backup power supply line circuit breaker (such as the bus tie switch) to quickly restore power supply to important loads; when a line or transformer experiences a momentary overload due to a fault, quickly disconnect the secondary load branches to avoid equipment damage; identify equipment protection action signals (such as bus differential protection), immediately isolate the fault area, and prevent cascading tripping;

[0083] The execution timing is as follows: it is executed within 0-3 seconds after the fault occurs, without waiting for global optimization calculations. It relies solely on local protection devices and real-time status data (such as voltage and switch status) to trigger the system, ensuring the rapid restoration of the power grid's basic power supply capacity.

[0084] The timing for the optimization and adjustment layer to be executed is 3-10 seconds after the rapid response layer completes the initial handling. At this time, the power grid has avoided collapse, but there may be problems such as local overload, unreasonable topology (such as ring network operation) or insufficient backup power capacity.

[0085] The optimization goal is to fine-tune the fast response results using lightweight algorithms (such as greedy algorithms and local search), for example: redistributing load to eliminate line overload; optimizing the switching operation sequence to reduce the power outage area; and verifying whether the capacity of the backup power supply line meets the remaining load demand. If it is insufficient, a load reduction strategy will be initiated.

[0086] The global optimization layer is executed when:

[0087] The optimization and adjustment layer is executed within 10 seconds to several minutes after completion, and global coordination is performed for complex multi-fault scenarios (such as multiple lines tripping at the same time or regional power grid disconnection).

[0088] The optimization objective is to generate the optimal power restoration strategy by using parallelized mixed-integer programming or an improved genetic algorithm, taking into account the following factors: minimizing power outage time and load loss; avoiding equipment overload and voltage collapse; maintaining the radial or ring network stability of the power grid topology; and coordinating the output support of distributed power sources (such as new energy sources).

[0089] The core advantage of the layered and progressive mechanism is that the rapid response layer (second-level) ensures survival, the optimization and adjustment layer (ten-second-level) prevents proliferation, and the global optimization layer (minute-level) seeks the optimal solution, forming a policy gradient in the time dimension. By layering, the NP-hard problem is decomposed into locally solvable subproblems, avoiding the exponential computational delay caused by multiple concurrent faults in traditional methods. The rapid response layer provides a safety net to ensure basic power supply, and subsequent layers optimize step by step. Even if the global optimization fails, the power grid can still maintain an acceptable operating state.

[0090] Furthermore, the hierarchical progressive strategy generation mechanism generates multi-layered self-healing strategies covering the second to minute timescales through a hierarchical design of time scale, computational complexity, and optimization objectives. The specific strategy content and hierarchical criteria are as follows: The strategy of the hierarchical progressive strategy generation mechanism:

[0091] The strategy types for the fast response layer (0-3 seconds) are: emergency handling strategies based on predefined rules; backup line switching: when the main power supply line fails, the backup power supply line circuit breaker (such as the bus tie switch) is automatically closed to restore power supply to critical loads; load instantaneous overload: when there is a momentary overload, the non-critical load branch is cut off to prevent equipment damage; fault area isolation: the faulty section switch is quickly disconnected according to the protection action signal (such as the bus differential protection) to prevent interlocking tripping; bus voltage support: distributed power sources or capacitor banks are put into operation to maintain the stability of the undervoltage bus voltage.

[0092] Optimization adjustment layer (3-10 seconds) strategy types: local optimization strategy based on lightweight algorithm; load redistribution: eliminate line overload and balance the load of each branch through greedy algorithm; topology optimization: adjust switch status (such as disconnecting tie switches in the ring network) to restore the radial operation of the power grid; standby capacity verification: check the remaining capacity of standby lines, and if insufficient, start load transfer or load reduction strategies; reactive power compensation adjustment: optimize capacitor bank switching and improve voltage quality.

[0093] Global optimization layer (10 seconds to several minutes) strategy types: global coordination strategy based on mixed integer programming (MIP) or improved genetic algorithm; multi-region load transfer: cross-regional coordination of backup power sources to minimize the scope of power outages; renewable energy output scheduling: integration of distributed power sources such as wind power and photovoltaics to support the power restoration process; long-term stability constraints: considering the grid N-1 safety criterion to avoid subsequent hidden dangers caused by the strategy; economic optimization: reducing operating costs (such as reducing the number of switching operations) while meeting reliability requirements.

[0094] First criterion for stratification: Time scale

[0095] Rapid Response Layer: Handles emergency events at the second level, relying on local protection devices and real-time status data (such as voltage and switch position), without requiring global calculations; Optimization and Adjustment Layer: Completes local optimization within ten seconds after initial handling, balancing real-time performance and computational efficiency; Global Optimization Layer: Performs global coordination at the minute level for complex scenarios, accepting higher computational complexity in exchange for the optimal solution.

[0096] Second criterion for layering: computational complexity

[0097] Fast Response Layer: Employs "if-then" rules or lookup table methods, with a computational complexity of O(1), ensuring a response time within seconds; Optimization and Adjustment Layer: Uses greedy algorithms or local search, with a complexity of O(n) (where n is the number of devices), suitable for decisions within ten seconds; Global Optimization Layer: Employs MIP or genetic algorithms, with a complexity of NP-hard, but can be completed within minutes through parallel processing and problem decomposition (such as decomposing the entire network problem into regional sub-problems).

[0098] Layering Criterion 3: Optimization Objectives

[0099] The rapid response layer aims to "avoid grid collapse," sacrificing some economic efficiency for survivability; the optimization and adjustment layer optimizes local operating efficiency (such as reducing load loss) while ensuring safety; and the global optimization layer integrates safety, economy, and long-term stability to generate the globally optimal power restoration plan.

[0100] It should be noted that by using a hierarchical progressive strategy generation mechanism, complex problems are decomposed into locally solvable subproblems, avoiding the delays caused by serial computation in traditional methods. A fast response layer ensures basic power supply, and subsequent layers continuously optimize rules through feedback loops, enhancing the generalization ability to unknown faults.

[0101] In this embodiment of the invention, in step S400, a strategy is executed according to a self-healing control method, which includes open-loop control, semi-closed-loop control, and closed-loop control.

[0102] In this embodiment of the invention, the verification of the generated power restoration strategy in step S400 includes: taking the minimization of load loss, the minimization of equipment overload and the minimization of voltage deviation as objective functions, and taking equipment capacity constraints, voltage stability constraints, topology connectivity constraints, switching action number constraints and economic constraints as constraints, and performing comprehensive optimization to obtain the optimal power restoration strategy.

[0103] The objective function is expressed as:

[0104] min(α·Load_Loss+β·Overload_Severity+γ·Voltage_Deviation)

[0105] Where α, β, and γ are weighting coefficients, and α+β+γ=1, Load-Loss is the unrecovered load, Overload-Severity is the degree of equipment overload, and Voltage-Deviation is the bus voltage deviation.

[0106] It should be noted that, in this embodiment of the invention, an overload level exceeding 1.0 is considered an overload, and the bus voltage deviation is defined as the degree of deviation from 1.0.

[0107] In this embodiment of the invention, step S400 further includes: the device capacity constraint is expressed as:

[0108]

[0109] Among them, P i Let be the active power of line i. Let i be the rated capacity of line i;

[0110] Voltage stability constraints are expressed as:

[0111]

[0112] Among them, V j V is the voltage amplitude at bus j. j min V j max Within the permitted range;

[0113] Topological connectivity constraints are expressed as:

[0114] Graph(G)is connected and acyclic(or with controlledloops)

[0115] Where Graph(G) is the topology graph of the power grid, connected means that there is a path between any two nodes, and acyclic means that the power grid should maintain a radial structure;

[0116] It should be noted that the power grid topology should maintain a radial or controllable ring network structure to avoid isolated nodes.

[0117] The constraint on the number of switching actions is expressed as follows:

[0118]

[0119] Where, ΔS k Let k be the number of times the switch is operated, and Switches be the set of all circuit breakers and disconnectors;

[0120] Economic constraints are expressed as:

[0121]

[0122] Among them, C l Let x be the operation cost of path l. l Choose a variable for the path; Paths is the set of all possible reconnection paths.

[0123] It should be noted that the switch operation count constraint is used to minimize the number of circuit breaker / disconnector operations and reduce equipment wear; the economic constraint is used to prioritize the use of low-cost power restoration paths.

[0124] Specifically, the verification process includes:

[0125] The action sequence (such as switching operation and load transfer) output by the hierarchical progressive strategy generation mechanism is converted into executable instructions for the simulation model.

[0126] Based on the current operating conditions of the power grid (such as load level and equipment status) and fault information (such as fault location and type), the initial conditions for simulation are generated.

[0127] The steady-state operating point of the power grid after the strategy is executed is calculated using the Newton-Raphson method or the fast decoupling method.

[0128] Verify that the power flow results satisfy all hard constraints; if soft constraints are violated, calculate the penalty value and update the objective function.

[0129] Generate strategy effectiveness scores (such as load recovery rate) and safety scores (such as the number of overloaded devices), and provide a conclusion on whether the validation is successful.

[0130] The verification algorithm adjusts the priorities of effectiveness, safety, and economy by adjusting the weight coefficients to adapt to the needs of different fault scenarios, which can ensure the basic operational safety of the power grid. Soft constraints guide strategy optimization through penalty terms.

[0131] Record detailed data for each self-healing process, including fault information, generated strategies, and execution results. Deeply mine historical data to discover potential problems, optimize power restoration rules, establish a feedback mechanism, and feed the actual execution effect back to the strategy generation layer to form a closed-loop optimization system.

[0132] It should be noted that this invention decomposes complex multi-fault scenarios into a rapid response layer, an optimization and adjustment layer, and a global optimization layer through a hierarchical progressive strategy generation mechanism. The rapid response layer achieves preliminary handling within seconds based on predefined emergency rules (such as switching backup power lines), avoiding the delay caused by serial calculation in traditional methods. The optimization and adjustment layer quickly eliminates local overload and topology conflicts. The global optimization layer decomposes large-scale problems into multiple sub-problems for simultaneous solution through parallel processing, significantly shortening the solution time.

[0133] It should also be noted that this invention enhances generalization capability through a dynamic power restoration rule base and a feedback loop mechanism. The rule base reserves interfaces for fuzzy logic and expert systems to perform preliminary classification and strategy matching for unknown fault modes (such as switch malfunctions caused by network attacks). Simultaneously, a historical data module records the characteristics and handling effects of unidentified faults in real time, continuously optimizing rule weights or generating new rules through reinforcement learning. For example, when the power grid suffers a new type of malware attack, the system can quickly activate backup lines based on fuzzy rules and incorporate the fault mode into the rule base through subsequent feedback loops, avoiding the limitations of relying on historical samples and significantly improving adaptability to unknown faults.

[0134] Example 2: The above example is an illustrative scheme for a high-voltage distribution network self-healing strategy generation method. It should be noted that the technical solution of this high-voltage distribution network self-healing strategy generation system belongs to the same concept as the technical solution of the above-described high-voltage distribution network self-healing strategy generation method. Details not described in detail in this example of the high-voltage distribution network self-healing strategy generation system can be found in the description of the above-described high-voltage distribution network self-healing strategy generation method.

[0135] This embodiment provides a high-voltage distribution network self-healing strategy generation system, comprising:

[0136] The data acquisition module is used to collect voltage, current and switch status information of each node in the power grid in real time, and to obtain power grid fault information.

[0137] The dynamic power restoration rule base construction module is used to dynamically adjust the priority and trigger threshold of each rule based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, so as to form a dynamic power restoration rule base.

[0138] The layered progressive strategy generation module is used to generate power restoration strategies using a layered progressive strategy generation mechanism. The layered progressive strategy generation mechanism initiates emergency handling based on a dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer.

[0139] The strategy verification and execution optimization module is used to verify the generated power restoration strategy. After successful verification, self-healing control is executed, and the actual power restoration effect is fed back to the dynamic power restoration rule base to update the rule parameters, forming a closed-loop optimization system.

[0140] This embodiment also provides an electronic device applicable to the self-healing strategy generation method for high-voltage distribution networks, including:

[0141] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the method for generating a self-healing strategy for high-voltage power distribution networks as proposed in the above embodiments.

[0142] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for generating a self-healing strategy for a high-voltage distribution network as proposed in the above embodiments.

[0143] The storage medium proposed in this embodiment and the method for generating a self-healing strategy for high-voltage distribution networks proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0144] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0145] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for generating a self-healing strategy for a high-voltage distribution network, characterized in that, include: Real-time collection of voltage, current, and switch status information at various nodes of the power grid to obtain power grid fault information; Based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, the priority and trigger threshold of each rule are dynamically adjusted to form a dynamic power restoration rule base. A hierarchical progressive strategy generation mechanism is adopted to generate a power restoration strategy. The hierarchical progressive strategy generation mechanism initiates emergency handling based on the dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer. The generated power restoration strategy is verified. If the verification is successful, self-healing control is executed, and the actual power restoration effect is fed back to the dynamic power restoration rule base to update the rule parameters, forming a closed-loop optimization system.

2. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 1, characterized in that, Combining real-time power grid state variables and historical fault handling feedback, the priority and trigger threshold of each rule are dynamically adjusted, including: updating the dynamic priority and trigger threshold of each basic power restoration rule online based on the real-time power grid state variable set and historical fault handling feedback; The state variables include load level, equipment health, and topology connectivity. After each rule execution, the weight coefficients and historical execution effect parameters are updated based on the actual power restoration effect and the forgetting factor, thus achieving adaptive evolution of the rule base.

3. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 2, characterized in that, A basic set of power restoration rules is constructed based on the fundamental constraints of power grid operation and empirical power restoration principles. These basic power restoration rules include: prioritizing the restoration of important loads, avoiding equipment overload, and maintaining the integrity of the power grid topology.

4. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 3, characterized in that, Based on the real-time power grid state variable set and historical fault handling feedback, the dynamic priority and trigger threshold of each basic power restoration rule are updated online, including: defining the real-time power grid state variable set, with each variable normalized to the [0,1] interval; the initial priority of each basic power restoration rule is set by expert experience, and its dynamic priority is corrected by weighting the real-time operating status of the power grid. The trigger threshold of the basic power restoration rules is dynamically adjusted based on historical execution results and real-time risk coefficients. When the historical execution success rate of any basic power restoration rule is lower than the average level and the current risk is high, the system automatically raises its trigger threshold to prevent accidental activation; conversely, it lowers the threshold to improve response sensitivity.

5. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 4, characterized in that, The hierarchical and progressive strategy generation mechanism initiates emergency response based on a dynamic power restoration rule base at the rapid response layer, including: responding immediately to severe faults and initiating basic power restoration strategies based on predefined emergency power restoration rules; When a sudden drop in bus voltage and loss of topology are detected, immediately disconnect the non-critical load switch on the de-energized bus; when the main power supply line fails, prioritize closing the backup power supply line circuit breaker to restore power to important loads; when a line or transformer experiences a momentary overload due to a fault, quickly disconnect the secondary load branch; identify equipment protection action signals and immediately isolate the fault area to prevent cascading tripping.

6. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 5, characterized in that, Verification of the generated power restoration strategy includes: taking minimizing load loss, minimizing equipment overload, and minimizing voltage deviation as objective functions, and taking equipment capacity constraints, voltage stability constraints, topology connectivity constraints, switching action count constraints, and economic constraints as constraints, and performing comprehensive optimization to obtain the optimal power restoration strategy; The objective function is expressed as: min(α·Load_Loss+β·Overload_Severity+γ·Voltage_Deviation) Where α, β, and γ are weighting coefficients, and α+β+γ=1, Load_Loss is the unrecovered load, Overload_Severity is the degree of equipment overload, and Voltage-Deviation is the bus voltage deviation.

7. The method for generating a self-healing strategy for a high-voltage distribution network as described in claim 6, characterized in that, Also includes: The equipment capacity constraint is expressed as follows: Among them, P i Let be the active power of line i. Let i be the rated capacity of line i; The voltage stability constraint is expressed as follows: Among them, V j V is the voltage amplitude at bus j. j min V j max Within the permitted range; The topological connectivity constraint is expressed as follows: Graph(G)is connected and acyclic(or with controlledloops) Where Graph(G) is the topology graph of the power grid, connected means that there is a path between any two nodes, and acyclic means that the power grid should maintain a radial structure; The constraint on the number of switching actions is expressed as follows: Where, ΔS k Let k be the number of times switch k is operated, and Switches be the set of all circuit breakers and disconnectors; The economic constraints are expressed as follows: Among them, C l Let x be the operation cost of path l. l Choose a variable for the path; Paths is the set of all possible reconnection paths.

8. A high-voltage distribution network self-healing strategy generation system, applied to the method described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect voltage, current and switch status information of each node in the power grid in real time, and to obtain power grid fault information. The dynamic power restoration rule base construction module is used to dynamically adjust the priority and trigger threshold of each rule based on the basic power grid restoration rules and combined with real-time power grid state variables and historical fault handling feedback, so as to form a dynamic power restoration rule base. The layered progressive strategy generation module is used to generate a power restoration strategy using a layered progressive strategy generation mechanism. The layered progressive strategy generation mechanism initiates emergency handling based on the dynamic power restoration rule base at the rapid response layer, and performs local and global optimization at the optimization adjustment layer and the global optimization layer. The strategy verification and execution optimization module is used to verify the generated power restoration strategy. After the verification is successful, self-healing control is executed, and the actual power restoration effect is fed back to the dynamic power restoration rule base to update the rule parameters and form a closed-loop optimization system.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the high-voltage distribution network self-healing strategy generation method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the high-voltage distribution network self-healing strategy generation method according to any one of claims 1 to 7.

Images