Method and system for compiling dispatch accident plan based on anti-accident simulation

Abstract

A method and system for writing a dispatch accident plan based on anti-accident simulation, relating to the general field of control or regulation systems, comprising: extracting historical fault data of a power grid to generate a standard fault data set; extracting fault feature vectors from the fault data set; extracting key nodes, operation sequences, and transfer paths in the handling process of the fault data set, and characteristically correlating the fault feature vectors with the handling process to generate a plan knowledge base; performing simulation combination based on the fault data set to generate an extended fault scenario; extracting target feature vectors from the extended fault scenario, and generating a plan framework in combination with the plan knowledge base; performing load flow calculation on the transfer scheme in the plan framework, and obtaining a target plan according to the equipment load constraint calculation; and storing the target plan in the plan knowledge base after verifying that the plan execution data of the target plan meets the expected execution result. Implementation of this application can improve the scenario adaptability of the dispatch accident plan and ensure the effectiveness of the plan.

Description

Technical Field

[0001] The present application relates to the general field of control or regulation systems, and in particular to a method and system for compiling a scheduling accident plan based on anti-accident simulation. Background Art

[0002] As power systems grow larger and more complex, and the proportion of renewable energy connected to the grid continues to increase, the uncertainty facing grid operations is increasing. Power system dispatchers must respond to a variety of emergencies, including equipment failures and natural disasters, to ensure power supply reliability and safe and stable system operation. Especially in the context of large-scale interconnected power grids, localized failures can trigger chain reactions, causing widespread power outages. Therefore, a scientific and efficient dispatch contingency plan system is essential.

[0003] In related technologies, the development of emergency response plans for power system dispatching primarily relies on expert judgment combined with simple simulation verification. This involves establishing a basic power system model, presupposing several typical fault conditions, and analyzing the system's response characteristics through simulation. Dispatching experts then design corresponding control strategies and emergency measures based on their experience. Furthermore, regular emergency drills are conducted to verify the feasibility of the plans and to refine and improve them.

[0004] However, due to the limitations of computing resources and time costs, related technologies can generally only simulate and analyze a limited number of accident scenarios when formulating plans, making it difficult to cover various complex fault scenarios that may occur in the actual operation of the power system. As a result, the scenarios in which the plans generated by related technologies are not adaptable enough, making it difficult to deal with fault events that have not been simulated. Summary of the Invention

[0005] The present application provides a method and system for compiling a dispatch accident plan based on anti-accident simulation, which is used to improve the scenario adaptability of the dispatch accident plan and ensure the effectiveness of the plan.

[0006] In the first aspect, the present application provides a method for writing a dispatching accident plan based on anti-accident simulation, which is applied to an intelligent decision support system. The method includes: extracting the fault type, faulty equipment, impact range and handling process of historical fault data of the power grid to generate a standard fault data set; extracting the power grid topology, equipment operating status and load distribution parameters in the fault data set to generate a fault feature vector; extracting the key nodes, operation sequences and transfer paths of the fault data set in the handling process, feature-associating the fault feature vector with the handling process, and generating a plan knowledge base; performing simulation combination based on the fault data set to generate an extended fault scenario; extracting the target feature vector in the extended fault scenario, and generating a plan framework in combination with the plan knowledge base; performing load flow calculation on the transfer plan in the plan framework, and obtaining the target plan according to the equipment load constraint calculation; after verifying that the plan execution data of the target plan meets the expected execution results, storing the target plan in the plan knowledge base.

[0007] In the above embodiment, the intelligent decision support system generates standard data sets and feature vectors by extracting historical fault data, establishes a plan knowledge base and performs fault scenario simulation expansion, and verifies the feasibility of the plan in combination with load flow calculation; historical experience is digitized and standardized, and coverage of more fault scenarios is achieved through simulation expansion, so that the generated plan has stronger scenario adaptability and reliability, thereby improving the efficiency and accuracy of power grid fault handling.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the steps of extracting key nodes, operation sequences and transfer paths of the fault data set in the handling process, characteristically associating the fault feature vector with the handling process, and generating a plan knowledge base specifically include: reading the equipment state change points from the handling process of the fault data set, reordering the key nodes according to the power grid topology structure, and obtaining a corrected state change sequence; reorganizing the operation instructions according to electrical causal relationships to obtain an operation instruction set; when extracting the transfer path in the power grid topology structure, performing de-loop calculation on the ring network structure in the transfer path to obtain a single path set; performing data association operations on the state change sequence, operation instruction set, single path set and fault feature vector to generate a plan knowledge base.

[0009] In the above embodiment, the intelligent decision support system reorders the key nodes in the fault handling process, reorganizes the operating instructions according to electrical causal relationships, and de-loops the ring network structure to obtain a single path set; this makes the handling plans stored in the plan knowledge base more standardized and efficient, facilitates subsequent retrieval and application, and improves the quality of plan generation.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the step of extracting and expanding the target feature vector in the fault scenario and generating a plan framework in combination with the plan knowledge base specifically includes: extracting electrical topology features, equipment health features and load distribution features from the fault scenario to generate a target feature vector; calculating feature similarity based on topological distance, load similarity and equipment type matching, and retrieving similar scenarios from the plan knowledge base; extracting reconstruction sequence, transfer path and equipment constraints from the plan template of similar scenarios to generate a basic plan framework; fusing and supplementing the basic plan framework based on multiple similar scenarios retrieved to generate a plan framework.

[0011] In the above embodiment, the intelligent decision support system extracts multi-dimensional features from the fault scenario, retrieves similar scenarios by calculating feature similarity, and fuses the plan templates of multiple similar scenarios to generate a plan framework; it makes full use of existing experience and improves the integrity and reliability of the plan through multi-scenario fusion.

[0012] In combination with some embodiments of the first aspect, in some embodiments, before performing load flow calculation on the power transfer plan in the plan framework and calculating the target plan according to the equipment load constraint, the method also includes: reading the conductor reactance resistance, transformer ratio and switch device status from the primary equipment parameter library of the power grid to construct a branch electrical parameter table; reading the voltage limit, current limit and power limit from the secondary equipment parameter library of the power grid to construct the equipment load boundary; according to the branch electrical parameter table and the equipment load boundary, determining the associated equation group containing the node injection power and the branch flow, and generating a load flow calculation formula.

[0013] In the above embodiment, the intelligent decision support system constructs an electrical parameter table and equipment load boundary from the equipment parameter library, and establishes a set of associated equations including node power and branch flow; it realizes strict verification of the feasibility of the plan and ensures that the generated plan meets the various constraints for the safe operation of the power grid.

[0014] In combination with some embodiments of the first aspect, in some embodiments, the step of determining a group of associated equations including node injection power and branch flow based on the branch electrical parameter table and the equipment load boundary, and generating a load flow calculation formula, specifically includes: substituting the conductor reactance resistance and transformer ratio in the branch electrical parameter table into the node admittance equation to generate an initial relationship between node voltage and injection power; correcting the initial relationship according to the state of the switching device, adding topological constraints of the circuit breaker and the disconnector, and obtaining a corrected relationship; converting the equipment load boundary into voltage amplitude constraints and phase angle difference constraints, and supplementing them to the corrected relationship to obtain a complete relationship; performing an algebraic transformation on the complete relationship to obtain a standard form of the load flow calculation formula.

[0015] In the above embodiment, the intelligent decision support system obtains a standard form of load flow calculation formula by modifying and transforming the equation, taking into account the constraints of the switching device status and the equipment load boundary; making the load flow calculation more accurate and comprehensive, and improving the reliability of the plan verification.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of performing simulation combination based on the fault data set to generate an expanded fault scenario, the method also includes: extracting the combination relationship of the fault equipment from the expanded fault scenario to construct fault propagation link data; classifying and organizing the fault scenarios according to the fault propagation link to generate a scenario feature group; extracting common data in the scenario feature group to generate a scenario evolution rule set.

[0017] In the above embodiment, the intelligent decision support system extracts combination relationships from fault scenarios to construct propagation links and generates a set of scenario evolution rules; it can predict the development trend of faults and improve the adaptability of the plan to complex fault scenarios.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of classifying and organizing the fault scenarios according to the fault propagation link and generating a scenario feature group, the method also includes: extracting the fault impact range data from the scenario feature group and marking the boundary points affected by the fault; dividing the fault isolation area according to the boundary points and calculating the electrical connection data between each fault isolation area; mapping the electrical connection data to the fault propagation link to obtain a new fault path; and performing electrical rule verification on the new fault path to obtain new scenario data.

[0019] In the above embodiment, the intelligent decision support system generates new fault paths and performs rule verification by dividing fault isolation areas and analyzing electrical connections; this expands the system's understanding of fault scenarios and improves the plan's ability to handle new types of faults.

[0020] In a second aspect, an embodiment of the present application provides an intelligent decision support system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the intelligent decision support system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when the above-mentioned computer program product is run on an intelligent decision support system, enables the above-mentioned intelligent decision support system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium comprising instructions. When the instructions are executed on an intelligent decision support system, the intelligent decision support system executes the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understood that the intelligent decision support system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects that can be achieved can be referenced to the beneficial effects of the corresponding methods and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. Due to the adoption of a method for feature extraction and plan knowledge base construction based on historical fault data, the expansion of fault scenarios through simulation expansion, and the technical solution of plan verification using load flow calculation, the system can systematically organize and expand the application of scattered historical experience, and ensure the feasibility of the plan through strict technical verification, effectively solving the problems in the existing technology that the plan formulation process relies on manual experience, covers limited scenarios, and has insufficient reliability verification, thereby achieving standardization, automation and reliability improvement of the plan generation process.

[0026] 2. Due to the adoption of a technical solution that reads electrical parameters and equipment constraints from the equipment parameter library and constructs a standardized load flow calculation formula, the system can conduct comprehensive technical verification of the emergency plan to ensure that all constraints on equipment operation are met. This effectively solves the problem of insufficient strictness in emergency plan verification and possible technical risks in existing technologies, thereby realizing the technical feasibility verification of the emergency plan and ensuring the safety and reliability of the plan execution.

[0027] 3. Due to the adoption of a technical solution based on fault propagation link analysis and scenario evolution rule extraction, the system can deeply analyze the fault development laws and predict possible fault evolution paths, effectively solving the problems of insufficient prediction of fault development trends and lack of targeted plans in existing technologies, thereby achieving accurate prediction and effective response to complex fault scenarios, and improving the practicality and foresight of the plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of a method for compiling a scheduling accident plan based on anti-accident simulation in an embodiment of the present application;

[0029] Figure 2 This is another flowchart of the method for compiling a scheduling accident plan based on anti-accident simulation in an embodiment of the present application;

[0030] Figure 3 It is a schematic diagram of the physical device structure of the intelligent decision support system in the embodiment of the present application. DETAILED DESCRIPTION

[0031] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations of one or more of the listed items.

[0032] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0033] For ease of understanding, the application scenarios of the embodiments of the present application are introduced below.

[0034] A large-scale power outage occurred in a provincial power grid, affecting power supply areas in multiple cities. The initial fault was a lightning strike on a 500kV transmission line. Due to flaws in the protection configuration, the fault escalated into a chain reaction. Dispatchers needed to quickly develop a plan to restore power, but faced challenges such as the large system scale, complex fault chains, and multiple power transfer paths. Traditional manual planning methods struggled to quickly analyze and make decisions, and were prone to missing key steps or violating technical constraints. This incident exposed the lack of intelligent planning capabilities in power grid fault response plans, particularly in complex fault scenarios, where efficient plan generation and optimization methods were lacking.

[0035] In related technologies, fault contingency plan retrieval and generation can be achieved through rule-based template matching and simple similarity calculations. This approach primarily relies on a preset rule base and fixed contingency plan templates, and cannot effectively handle complex fault scenarios and dynamically changing system states. The following describes a scenario using the related art method for writing dispatch contingency plans based on reverse accident simulation.

[0036] A local power supply company used a traditional emergency plan management system to handle substation equipment failures. This system primarily relied on historical emergency plan templates, searching for similar cases through keyword matching. When a 220kV transformer failure occurred, the system retrieved multiple historical emergency plans. However, these plans were designed for single equipment failures and were unable to effectively address cascading failures such as malfunctioning busbar sectionalizers caused by the transformer failure. Furthermore, the system failed to dynamically assess the feasibility of emergency plans and failed to account for changes in the current network structure and load distribution. This resulted in overloads on some lines during the actual implementation of the power transfer plan.

[0037] The method for writing a dispatch accident plan based on counter-accident simulation, as described in the embodiments of this application, achieves intelligent generation and optimization of fault plans through multi-dimensional feature vector extraction, dynamic scenario analysis, and load flow constraint verification. This not only accurately identifies fault propagation links but also adaptively adjusts the plan framework to ensure the technical feasibility of the plan. The following describes scenarios in which the method for writing a dispatch accident plan based on counter-accident simulation, as described in this application, is used.

[0038] A regional power grid implemented an intelligent decision support system to handle 110kV line faults. The system first extracted topology and operating parameters from the fault data to construct a standardized feature vector. Feature matching then retrieved similar cases from a knowledge base and optimized the emergency plan based on the current network status. When a line tripped and triggered the disconnection of distributed generation (DGs), the system quickly analyzed the fault propagation chain, identified key nodes, and automatically generated a complete emergency plan encompassing fault isolation, load transfer, and power restoration. The system also performed load flow verification on the emergency plan to ensure it met equipment capacity and voltage quality requirements, effectively mitigating the risk of secondary faults.

[0039] It can be seen that the scheduling accident plan writing method based on anti-accident simulation in the embodiment of the present application can not only achieve rapid plan formulation, but also effectively solve the problems of poor plan adaptability and insufficient constraint verification in traditional methods, thereby realizing the intelligence and reliability of plan generation.

[0040] For ease of understanding, the following describes the process of the method provided by this implementation in combination with the above scenario. Figure 1 , which is a flow chart of a method for writing a scheduling accident plan based on anti-accident simulation in an embodiment of the present application.

[0041] S101. Extract the fault type, faulty equipment, impact range, and handling process of historical power grid fault data to generate a standard fault data set.

[0042] Among them, historical power grid fault data represents recorded information on various types of fault events that have occurred during the operation of the power system. Fault type refers to the classification of the causes of power grid abnormalities, including equipment failure, natural disasters, and human operational errors. Faulty equipment refers to the specific power equipment that has failed, including transformers, circuit breakers, transmission lines, etc. The impact range is used to indicate the power outage area caused by the fault, the scope of equipment tripping, and the degree of impact on system stability. The handling process refers to the complete operation and control process from fault occurrence to power restoration. The standard fault dataset represents a collection of historical fault cases that have been standardized and formatted.

[0043] Before building a contingency plan knowledge base, the intelligent decision support system must first acquire and process historical fault data. Specifically, the intelligent decision support system first extracts fault records from the dispatch automation system database, including raw data such as equipment action information recorded by the SCADA system, protection device action reports, and dispatcher operation records. This data is then cleaned and standardized to unify the data format, remove duplicate and erroneous information, and supplement missing fields. The fault type is then labeled according to predefined classification standards, determining the specific type and location of the faulty equipment. The scope of the fault is also analyzed, including the direct power outage area and the scope of the chain reaction. Finally, the fault handling process is standardized into a sequence of standard operating steps to form a complete standard fault data set.

[0044] In some embodiments, the extraction and standardization of historical fault data can be achieved in a variety of ways: optionally, a rule-based data cleaning method is used to filter and convert the original data through preset data quality rules, including data format unification, outlier detection and correction, missing value completion and other processing; optionally, a knowledge graph-based data association method is used to establish a relationship network between entities such as equipment, faults, and operations, and the fault propagation path and impact range are extracted through graph algorithms. It is understandable that other data processing and feature extraction methods can also be used to achieve the standardization of historical fault data, which is not limited here. It should be noted that the key features of the original data should be retained during the data cleaning process to avoid information loss due to oversimplification.

[0045] In actual applications, we may encounter problems such as uneven quality and inconsistent formats of historical fault data. To this end, the intelligent decision support system adopts a multi-level data quality assessment and processing solution: first, a data quality scoring system is established to score the data from the dimensions of completeness, accuracy, consistency, etc.; then, the data is graded according to the scoring results, and high-quality data is directly stored, medium-quality data is corrected and stored, and low-quality data is temporarily stored for processing. For the problem of inconsistent formats, the system establishes a unified data template, which contains required fields and optional fields, and converts data in different formats into a standard format through mapping rules. For example, a circuit breaker refusal to operate failure occurred in a substation. The original records may be scattered in multiple systems, and the recording methods are inconsistent. The system integrates them into standard fault records through data association and conversion.

[0046] S102: Extract the power grid topology, equipment operating status, and load distribution parameters from the fault data set to generate a fault feature vector.

[0047] The grid topology represents the physical and electrical connections between devices in the power system. Equipment operating status refers to the operating parameters of power equipment, including voltage, current, active power, reactive power, and other operational parameters. Load distribution parameters represent the power consumption characteristics of each load node in the system, including load size, type, and importance. A fault feature vector is a mathematically expressed set of fault scenario characteristics.

[0048] After acquiring a standard fault dataset, the intelligent decision support system needs to extract key information that reflects the characteristics of the fault scenario. Specifically, the intelligent decision support system first establishes a grid topology model, including a connection matrix for busbars, transformers, lines, and other equipment. It then retrieves the operating data of each device at the time of the fault, recording state changes before, during, and after the fault. It also analyzes the load distribution, including the spatial distribution and temporal variation characteristics of each load type. Finally, these features are organized according to a predefined vector structure to form a standardized fault feature vector.

[0049] In some embodiments, the extraction and vectorization of fault features can be achieved in a variety of ways: optionally, graph theory methods are used to analyze the topology of the power grid, network characteristics are described through indicators such as node degree and centrality, and the electrical distance and correlation strength between devices are calculated; optionally, time series analysis methods are used to process equipment operating status data to extract state change trends, mutation characteristics, and correlation characteristics. It is understandable that other feature engineering methods can also be used to achieve feature extraction and representation of fault scenarios, which are not limited here. It should be noted that the dimensional design of the feature vector should consider the balance between computational efficiency and feature expression capabilities.

[0050] In practical applications, problems such as incomplete feature extraction or feature redundancy may arise. Intelligent decision support systems employ feature importance assessment methods for optimization. First, the discriminatory power of each feature is evaluated using metrics such as information gain and mutual information. Dimensionality reduction methods such as principal component analysis are then used to eliminate redundant features. Finally, cross-validation is used to verify the effectiveness of feature selection. For example, for a transmission line trip, the system not only extracts the features of the fault point but also considers related features such as the status of adjacent equipment and the likelihood of load transfer.

[0051] S103: Extract key nodes, operation sequences, and transfer paths in the fault data set during the handling process, associate the fault feature vector with the handling process, and generate a plan knowledge base.

[0052] Key nodes represent important time points and decision points in the fault handling process. An operation sequence is an ordered set of specific control and operation steps in the handling process. A transfer path represents the backup power supply channel used to restore power. Feature association establishes a mapping relationship between fault features and handling solutions. The contingency plan knowledge base represents a structured database containing the correspondence between fault features and handling solutions.

[0053] After extracting features, the intelligent decision support system needs to analyze historical response experience and build a knowledge base. Specifically, the intelligent decision support system first performs a time-series analysis of the response process, identifying key decision points and state transitions. It then normalizes the operational steps into a standard instruction sequence, including specific actions such as on / off switching and parameter adjustments. It also analyzes the available transfer paths in the system, assessing the available capacity and switching time for each path. Finally, it establishes a mapping from feature vectors to response plans, forming a structured knowledge base for emergency plans.

[0054] In some embodiments, the extraction of treatment experience and the construction of a knowledge base can be achieved through a variety of methods: optionally, process mining methods can be used to analyze operation sequences and identify key paths and branching conditions in the treatment process; optionally, knowledge graph methods can be used to construct an association network between features and solutions, supporting similarity-based solution retrieval and recommendation. It is understood that other knowledge engineering methods can also be used to represent and organize treatment experience, which are not limited here. It should be noted that the organizational structure of the knowledge base should support rapid retrieval and dynamic updating.

[0055] In practical applications, inconsistent handling experience or knowledge conflicts may arise. The intelligent decision support system uses a knowledge consistency check method: first, a handling rule base is established, including electrical regulations and safety constraints. Then, rules are verified against historical handling plans to identify potential conflicts. Finally, conflicts are resolved through expert review to ensure the consistency of the knowledge base. For example, for a line fault, there may be multiple power transfer options. The system needs to evaluate the feasibility and priority of each option based on the rules.

[0056] S104: Perform simulation combination based on the fault data set to generate an extended fault scenario.

[0057] Simulation combination refers to the reorganization and expansion of historical faults through computer simulation. Extended fault scenarios refer to new possible fault conditions generated through simulation. Combination rules represent the constraints and logical relationships that control the fault combination process. Scenario validity refers to the rationality and feasibility of the generated scenarios.

[0058] After establishing a basic knowledge base, the intelligent decision support system needs to expand fault scenarios to improve the coverage of emergency plans. Specifically, the intelligent decision support system first analyzes the combination patterns of historical faults, including device correlations, fault propagation patterns, and chain reactions. It then builds a fault simulation model, including an electrical characteristics model and a device state transition model. Next, it combines faults according to pre-set rules, including single-device fault expansion, multi-device fault superposition, and fault propagation chain construction. Finally, the generated scenarios are validated to ensure compliance with physical laws and engineering practices.

[0059] In some embodiments, fault scenarios can be simulated and expanded using a variety of methods: Optionally, Monte Carlo methods can be used to randomly generate fault combinations, controlling the temporal and spatial distribution of fault occurrences through probability distribution; Optionally, rule-based reasoning methods can be used to deduce possible fault propagation paths based on the physical and electrical relationships between devices. It is understood that other simulation methods can also be used to expand fault scenarios, and these are not limited here. It should be noted that the scenario generation process should consider the actual operational constraints of the system.

[0060] In some embodiments, an intelligent decision support system can employ a fault combination generation algorithm. First, based on the grid topology, the electrical correlation and geographic proximity between devices are calculated to construct a device correlation matrix. Then, using a layered fault injection approach, starting with the initial fault, the fault impact range is gradually expanded based on electrical distance, while also considering the operating characteristics of protection devices and system operational constraints. Within each fault expansion layer, the faulty device combination is determined through Monte Carlo sampling, and a Markov chain is used to simulate the fault propagation process. For each generated fault scenario, the system performs dynamic simulation verification, analyzing the system's transient response and steady-state characteristics, and eliminating scenarios that do not conform to physical laws. Expert rules are also introduced to evaluate the scenarios, scoring them based on typicality, severity, and probability of occurrence, to select a set of valuable scenarios. The entire process utilizes a parallel computing architecture, with dynamic load balancing improving computational efficiency. The resulting expanded scenarios ensure both physical feasibility and the complexities likely to be encountered in actual engineering projects.

[0061] In practical applications, scenarios may explode in number or be of low quality. The intelligent decision support system employs a scenario screening and optimization approach: first, a scenario evaluation index system is established, including likelihood, severity, and typicality; then, the generated scenarios are scored and ranked; and finally, a representative set of scenarios is selected. For example, for a substation fault scenario, the system needs to consider the electrical and geographical distances between equipment to avoid generating unreasonable fault combinations.

[0062] S105: Extract target feature vectors from the expanded fault scenario and generate a contingency plan framework in combination with the contingency plan knowledge base.

[0063] The target feature vector represents the characteristic description of the fault scenario to be handled. The contingency plan framework refers to the overall structure of the fault handling plan. Feature similarity indicates the degree of similarity between different fault scenarios. The contingency plan template represents the standardized format of the handling plan.

[0064] After generating expanded scenarios, the intelligent decision support system needs to develop corresponding response plans for each scenario. Specifically, the intelligent decision support system first extracts standardized feature vectors from the newly generated fault scenario. It then searches the response plan knowledge base for similar scenarios and calculates the distance or similarity between the feature vectors. It then extracts available operational steps and control strategies from the response plans for similar scenarios. Finally, it combines the response experience from multiple similar scenarios to generate a response plan framework for the new scenario.

[0065] In some embodiments, the generation of the emergency plan framework can be achieved through a variety of methods: optionally, a case-based reasoning method based on similarity is used to retrieve and adapt applicable disposal plans from historical emergency plans; optionally, a template filling method is used to automatically select and assemble standard modules in the emergency plan template based on scenario characteristics. It is understood that other methods can also be used to achieve the construction of the emergency plan framework, which is not limited here. It should be noted that the emergency plan framework should be flexible enough to adapt to the specific execution environment.

[0066] In actual applications, you may encounter issues such as incomplete emergency plan frameworks or poor execution. The intelligent decision support system utilizes an emergency plan optimization approach: first, a plan evaluation model is established, taking into account factors such as operational feasibility, time efficiency, and resource consumption; then, simulations are used to verify the effectiveness of the plan execution; and finally, the plan framework is adjusted and refined based on the verification results. For example, in a complex cascading failure scenario, the system needs to balance power restoration speed with operational safety and rationally arrange the sequence of operational steps.

[0067] S106. Perform load flow calculation on the power transfer plan in the plan framework, and obtain the target plan according to the equipment load constraint.

[0068] The power transfer plan represents the specific path and operational plan for restoring power. Load flow calculation, also known as power flow calculation, is a numerical method for analyzing the voltage, current, and power distribution in a power system. Equipment load constraints represent the electrical and physical limits that various types of equipment must meet during operation. A target plan represents a complete, verified, and practically executable response plan.

[0069] After generating the emergency plan framework, the intelligent decision support system needs to conduct detailed technical feasibility verification. Specifically, the intelligent decision support system first establishes a detailed power grid calculation model, including the node admittance matrix and branch parameters. Then, for each power transfer scenario in the emergency plan framework, the system calculates the power flow distribution after power is restored. It then checks whether all equipment operating parameters meet the constraints, including voltage qualification, line current carrying capacity, and transformer capacity. Finally, based on the calculation results, the emergency plan is optimized and adjusted until a feasible solution that meets all constraints is obtained.

[0070] It should be noted that the power grid calculation model is a static characteristic model constructed around the node admittance matrix. The network impedance characteristics are expressed in complex form, and the admittance matrix elements Yij = Gij + jBij reflect the electrical connection between nodes i and j. The power balance equation is established based on the node voltage method. The injected power at any node i can be expressed as Si = Pi + jQi = Vi∑(VjYij). During model training, historical measurement data is first used for parameter identification, and least squares estimation is used to obtain the optimal values ​​for line and transformer parameters. State estimation techniques are then used to modify the model parameters based on current measurement data to ensure that the model accurately reflects the system's operating status. In practical applications, real-time switch status, power generation, and load data are input, and the model outputs state variables such as node voltage and branch power, which are used to verify the technical feasibility of emergency plans. For example, if fault N-1 occurs at a 220kV substation, the model can calculate the operating parameters of each device under the power transfer plan and determine whether there is an overload risk.

[0071] In some embodiments, the technical verification of the plan can be achieved in a variety of ways: optionally, the improved Newton-Raphson method is used to solve the load flow equation, considering the voltage and reactive power optimization and network loss minimization objectives; optionally, a sensitivity analysis method is used to evaluate the impact of system parameter changes on the feasibility of the plan, thereby improving the robustness of the plan. It is understandable that other power system analysis methods can also be used to achieve plan verification, which is not limited here. It should be noted that the calculation process should take into account the dynamic characteristics of the system operation.

[0072] In practical applications, load flow calculations may not converge or constraints may be difficult to satisfy. Intelligent decision support systems employ a hierarchical approach: first, the network is partitioned, breaking down the large-scale system into several subsystems; then, detailed calculations are performed within the subsystems; and finally, the coupling relationships between subsystems are considered to coordinate the overall solution. For example, when considering a power transfer plan for a specific regional power grid, the system must simultaneously consider multiple constraints, including voltage distribution, power transmission, and equipment load.

[0073] S107. After verifying that the plan execution data of the target plan meets the expected execution result, the target plan is stored in the plan knowledge base.

[0074] Plan execution data represents the process and results of executing a plan in a simulation environment. Expected execution results refer to the target states and performance indicators predetermined during plan design. Plan storage involves saving verified plans to a knowledge base in a standard format. Plan evaluation metrics represent a quantitative indicator system for measuring plan execution effectiveness.

[0075] After completing emergency plan verification, the intelligent decision support system needs to evaluate its execution and update its knowledge base. Specifically, the intelligent decision support system first executes the complete emergency plan steps in a simulation environment, recording system state changes and operational responses. The results are then compared with expected targets, evaluating metrics such as power restoration time, load transfer efficiency, and system stability margin. Plans that achieve the expected results are then standardized, extracting key features and implementation points. Finally, qualified plans are stored in the knowledge base and associated with the fault characteristics.

[0076] In some embodiments, the evaluation and storage of emergency plans can be achieved through a variety of methods: optionally, a multi-objective evaluation method is used to comprehensively consider the technical feasibility, economic efficiency, and operability of the emergency plan; optionally, a knowledge graph method is used to construct a semantic association network between the emergency plan and the fault scenario to support intelligent retrieval and recommendation. It is understood that other methods can also be used to achieve emergency plan evaluation and knowledge base updates, which are not limited here. It should be noted that the knowledge base update process should ensure the consistency and integrity of the data.

[0077] In practical applications, problems may arise such as inconsistent emergency plan evaluation standards or inefficient knowledge accumulation. The intelligent decision support system adopts a standardized management approach: first, a unified indicator system for emergency plan evaluation is established, including mandatory and optimized indicators; then, emergency plan templates and storage structures are designed to standardize the management of emergency plan knowledge; and finally, a knowledge update mechanism is established to support the continuous optimization and improvement of emergency plans. For example, for emergency plans for a typical type of fault, the system needs to comprehensively consider its universality and specificity, and rationally organize and store relevant knowledge.

[0078] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , which is another flow chart of the method for writing a scheduling accident plan based on anti-accident simulation in an embodiment of the present application.

[0079] S201. Extract the fault type, faulty equipment, impact range, and handling process from historical fault data of the power grid to generate a standard fault data set.

[0080] Referring to step S101 , the intelligent decision support system collects and organizes power grid fault information from historical fault records to form a standardized fault data set.

[0081] S202: Extract the power grid topology, equipment operating status, and load distribution parameters from the fault data set to generate a fault feature vector.

[0082] Referring to step S102 , the intelligent decision support system extracts system structure, operation and load characteristics from the standard fault data set and constructs a fault feature vector.

[0083] S203: Extract key nodes, operation sequences, and transfer paths in the fault data set during the handling process, associate the fault feature vector with the handling process, and generate a plan knowledge base.

[0084] Referring to step S103 , the intelligent decision support system analyzes the key links and operation steps in the fault handling process and establishes a correlation between features and handling.

[0085] In some embodiments, the intelligent decision support system will conduct a systematic analysis of the fault handling process, that is, the intelligent decision support system will read the equipment state change points from the handling process of the fault data set, reorder the key nodes according to the power grid topology, and obtain a corrected state change sequence; reorganize the operation instructions according to the electrical causal relationship to obtain an operation instruction set; when extracting the transfer path in the power grid topology, the ring network structure in the transfer path is de-looped to obtain a single path set; the state change sequence, operation instruction set, single path set and fault feature vector are subjected to data association operations to generate a plan knowledge base.

[0086] Among them, a device state change point represents the moment when the device's operating state changes during a fault process. A key node refers to the control and measurement points that have a significant impact on system state transitions. A state change sequence represents a record of device state transitions sorted by time and topology. Electrical causality refers to the logical dependencies between device operations. An operation instruction set represents a standardized sequence of control instructions. A ring network structure refers to a network structure with multiple parallel power supply paths within the power grid. A single path set represents the set of basic power supply paths after redundancy is removed. Data association operations are the mathematical processing used to establish mapping relationships between data.

[0087] Before building a contingency plan knowledge base, the intelligent decision support system needs to systematically analyze and reconstruct the historical handling process. Specifically, the intelligent decision support system first extracts records of equipment status changes from historical data, including information such as protection actions, switch position changes, and measurement value mutations. It then analyzes the propagation path of state changes based on the grid topology and reorders key nodes according to physical connection relationships and operation sequence. It then analyzes the dependencies between operation instructions and combines related instructions into a complete operation sequence. When processing the transfer path, the ring network structure is identified through a graph theory algorithm, and the minimum spanning tree method is used to obtain the basic power supply path. Finally, a multidimensional association matrix is ​​established to map the state sequence, operation instructions, and path information with fault characteristics to form a structured contingency plan knowledge base.

[0088] In some embodiments, the analysis and reconstruction of the historical disposal process can be achieved in a variety of ways: optionally, a method based on time series analysis is adopted, firstly a time window is established to identify the state change point, then the sliding correlation analysis is used to determine the correlation of the state change, and finally the triggering condition of the state transition is determined by causal reasoning; optionally, a method based on graph theory is adopted, first a network topology diagram is constructed and the device status is marked, then the shortest path algorithm is applied to calculate the electrical distance between key nodes, and finally the necessary operation sequence is determined by cut set analysis. It is understandable that other data analysis methods can also be used to achieve the reconstruction of the disposal process, which is not limited here. It should be noted that the safety constraints and timing requirements of the equipment operation should be considered during the processing.

[0089] In practical applications, problems may arise such as incomplete state change records or unclear causal relationships. The intelligent decision support system employs a data completion and verification approach: first, a device state transition model is established, including the normal operating sequence and protection action rules. The model is then used to infer missing state change points. Finally, rule verification ensures the rationality of the state sequence. For example, if a circuit breaker trip record is missing, the system can infer the approximate tripping moment based on the state changes and protection configuration of the relevant equipment and verify that this inference complies with the system's operating procedures.

[0090] S204: Perform simulation combination based on the fault data set to generate an extended fault scenario.

[0091] Referring to step S104 , the intelligent decision support system uses historical fault data to perform combined simulations to expand possible fault scenarios.

[0092] S205: Extract the combination relationship of the faulty devices from the expanded fault scenario and construct fault propagation link data.

[0093] The faulty device combination relationship represents the electrical and physical connections between multiple faulty devices. Fault propagation link data describes the propagation path of a fault from its initial point to other devices. The device association degree indicates the likelihood of fault propagation between devices. The propagation sequence indicates the chronological order in which a fault propagates between different devices.

[0094] After generating expanded fault scenarios, the intelligent decision support system needs to analyze the fault propagation characteristics. Specifically, the intelligent decision support system first identifies the set of faulty devices involved in each scenario and establishes a correlation matrix between the devices. It then calculates the probability of fault propagation between devices based on electrical connections and physical distances. It then analyzes the temporal characteristics of fault occurrence and constructs a directed graph model of fault propagation. Finally, the propagation link data is standardized and stored for subsequent scenario analysis.

[0095] In some embodiments, fault propagation links can be constructed using a variety of methods: Optionally, complex network analysis methods can be used to identify key propagation nodes and primary propagation paths through node centrality and path analysis; alternatively, temporal correlation analysis methods can be used to derive causal relationships of fault propagation based on device tripping sequences. It is understood that other network analysis methods can also be used to extract fault propagation characteristics, and these are not limited here. It should be noted that propagation link analysis should take into account the device's protection configuration and control strategy.

[0096] In actual applications, problems such as incomplete propagation links or false associations may arise. The intelligent decision support system employs a link verification approach: first, a device protection configuration model is established to verify whether the propagation link conforms to the protection logic; then, expert knowledge rules are used to filter out unreasonable propagation paths; and finally, the reliability of the propagation link is verified through historical case studies. For example, if a line fault causes a cascading trip, the system needs to analyze the protection action sequence to confirm the actual fault propagation path.

[0097] S206: Classify and organize the fault scenarios according to the fault propagation links to generate scenario feature groups.

[0098] Scenario classification refers to categorizing scenarios based on fault propagation characteristics. A feature group is a collection of scenarios with similar propagation characteristics. Classification criteria refers to the criteria used to categorize scenarios. Feature similarity indicates the degree of similarity between the propagation characteristics of different scenarios.

[0099] After obtaining the fault propagation link, the intelligent decision support system needs to systematically classify the scenarios. Specifically, the system first defines a multidimensional feature space for scenario classification, including dimensions such as propagation range, propagation speed, and impact. It then calculates feature distances between scenarios and uses a clustering algorithm to group them. It then analyzes the typical characteristics of each feature group and extracts the common attributes of the scenarios within the group. Finally, it establishes a hierarchical structure for feature groups to support multi-level scenario management.

[0100] In some embodiments, fault scenarios can be classified using a variety of methods: Optionally, a hierarchical clustering approach can be employed to construct a hierarchical classification system for scenarios through a bottom-up merging strategy; Alternatively, a fuzzy clustering approach can be employed to allow scenarios to belong to multiple feature groups, increasing classification flexibility. It is understood that other classification methods can also be employed to organize and manage scenarios, and these are not limited here. It should be noted that the classification results should have clear physical meaning and engineering value.

[0101] In practical applications, problems may arise with fuzzy classification boundaries or difficulty determining the number of categories. Intelligent decision support systems employ an adaptive classification approach: first, they evaluate the effectiveness of different classification schemes using metrics such as the silhouette coefficient; then, they determine an appropriate number of categories based on expert experience; and finally, they maintain classification effectiveness through regular evaluation and dynamic adjustments. For example, for a regional power grid fault scenario, the system needs to balance the level of classification sophistication with practicality.

[0102] In some embodiments, the intelligent decision support system will perform a fault propagation impact analysis, that is, the intelligent decision support system will extract the fault range data from the scenario feature group and mark the boundary points of the fault impact; divide the fault isolation area according to the boundary points, and calculate the electrical connection data between each fault isolation area; map the electrical connection data to the fault propagation link to obtain a new fault path; perform electrical rule verification on the new fault path to obtain new scenario data.

[0103] Among them, the fault impact range represents the physical extent of the fault's spread. Boundary points refer to critical locations within the fault-affected area. Fault isolation areas represent network subareas that can be isolated from each other through switching operations. Electrical connection data is a quantitative indicator describing the degree of electrical connectivity between areas. Emerging fault paths represent predicted possible fault propagation paths. Electrical rules refer to the physical laws and technical specifications that power system operation must comply with. Emerging scenario data represents extended fault scenarios generated through analysis.

[0104] When expanding fault scenarios, the intelligent decision support system needs to deduce possible propagation scenarios based on known fault characteristics. Specifically, the intelligent decision support system first analyzes the impact range of historical faults and determines the boundary characteristics of the fault's impact. It then divides the network into sub-regions based on switchgear locations and establishes a connectivity matrix between these regions. It then calculates the electrical coupling strength between these regions, including factors such as power flow distribution and impedance characteristics. During the mapping process, constraints such as device protection configuration and system operation mode are considered. Finally, the rationality of the newly generated scenario is verified using electrical rules, including technical requirements such as power balance and voltage stability.

[0105] In some embodiments, fault scenario expansion can be achieved in a variety of ways: optionally, using a network propagation model, first establishing a state transition matrix for fault diffusion, then simulating the propagation process of the fault in the network, and finally evaluating the probability of each node being affected; optionally, using an expert system approach, first establishing a fault rule base to describe typical propagation patterns, then predicting possible propagation paths through rule reasoning, and finally verifying the technical feasibility of the scenario. It is understandable that other scenario generation methods can also be used to implement fault propagation analysis, which is not limited here. It should be noted that the scenario expansion process should take into account the dynamic characteristics and uncertainty of the system.

[0106] In actual applications, the problem of irrational generated scenarios may arise. The intelligent decision support system employs a multi-layered scenario verification approach: first, a scenario evaluation index system is established, including physical feasibility, operational feasibility, and impact severity; then, simulation tools are used to verify the dynamic characteristics of the scenarios; and finally, expert knowledge is used to screen valuable typical scenarios. For example, if a generated fault propagation path violates the protection configuration logic, the system automatically identifies and eliminates such irrational scenarios.

[0107] S207: Extract common data from the scene feature group to generate a scene evolution rule set.

[0108] Common data represents the characteristic patterns shared by scenarios within a feature group. The evolution rule set is a set of rules that describe the laws governing fault development. Feature patterns represent typical feature combinations during scenario development. Rule credibility indicates the reliability of the rules.

[0109] After completing scenario classification, the intelligent decision support system needs to summarize fault evolution patterns. Specifically, the system first performs a time series analysis of the scenarios within each feature group to extract key stages and turning points in fault development. It then identifies characteristic patterns at each stage, including triggering conditions, evolutionary processes, and outcome characteristics. It then uses a frequent pattern mining algorithm to discover statistically significant evolutionary rules. Finally, the rules are evaluated for credibility to form a hierarchical rule set.

[0110] It should be noted that fault evolution patterns can be calculated based on a fault evolution model; this fault evolution model utilizes time series data mining to construct a fault propagation prediction model. First, the fault process is represented as a state sequence {St}, where each state consists of a device state vector and a system measurement vector. A sequential pattern mining algorithm is used to extract frequent subsequence patterns, and the transition probability matrix P(St+1|St) between subsequences is calculated. Association rule learning is also used to discover the triggering conditions for state transitions, such as If(condition) Then P(event)=p. Model training utilizes maximum likelihood estimation, with the optimization objective being to maximize the probability of occurrence of historical fault sequences. Training data includes time series records such as protection action sequences and device tripping sequences. When the model is used, the current system state S0 is input and the probability distribution P(St|S0) of subsequent states is predicted, guiding emergency plan development. For example, after a line short circuit occurs, the model can predict the potential cascading tripping paths caused by the protection configuration.

[0111] In practical applications, rule conflicts or insufficient rule generalization may occur. Intelligent decision support systems employ a rule optimization approach: first, a rule evaluation system is established, including metrics such as support and confidence. Redundancies and inconsistencies are then eliminated through rule pruning and merging. Finally, the generalization of the rules is ensured through verification in new scenarios. For example, for the evolution of a particular type of equipment failure, the system needs to balance the accuracy and generalizability of the rules.

[0112] S208: Extract target feature vectors in the expanded fault scenario and generate a contingency plan framework in combination with the contingency plan knowledge base.

[0113] Referring to step S105 , the intelligent decision support system analyzes the newly generated fault scenario characteristics and generates a disposal plan framework based on existing plan knowledge.

[0114] In some embodiments, the intelligent decision support system will perform feature extraction and similar scenario retrieval, that is, the intelligent decision support system will extract electrical topology features, equipment health features and load distribution features from the fault scenario to generate a target feature vector; calculate feature similarity based on topological distance, load similarity and equipment type matching, and retrieve similar scenarios from the plan knowledge base; extract reconstruction sequence, transfer path and equipment constraints from the plan template of similar scenarios to generate a basic plan framework; integrate and supplement the basic plan framework based on multiple similar scenarios retrieved to generate a plan framework.

[0115] Among them, electrical topology characteristics represent the connection relationship and electrical parameter characteristics of the power grid structure. Equipment health characteristics refer to comprehensive indicators reflecting the operating status and reliability of equipment. Load distribution characteristics represent the spatial distribution and power consumption characteristics of the system load. The target feature vector refers to the standardized feature description of the fault scenario to be handled. Topological distance represents the length of the electrical connection path between devices. Load similarity refers to the degree of matching of load distribution patterns. Equipment type matching indicates the degree of similarity of equipment characteristics. The plan template refers to the standardized disposal plan framework. The reconstruction sequence represents the order of operation steps in the plan. Equipment constraints refer to the equipment operation restrictions that need to be followed during the execution of the plan.

[0116] Before generating a contingency plan framework, the intelligent decision support system needs to extract scenario features and retrieve similar scenarios. Specifically, the intelligent decision support system first constructs a multidimensional feature space to extract the structural features of the electrical topology diagram, the statistical features of the equipment operating parameters, and the spatial features of the load distribution from the fault scenario. It then calculates the similarity between the feature vectors, including the edit distance of the topology diagram, the Euclidean distance of the equipment parameters, and the KL divergence of the load distribution. The historical scenario with the highest similarity is then retrieved from the knowledge base to obtain its contingency plan template. During the template reconstruction process, the portability of the operation sequence is analyzed, the connection relationship of the transfer path is adjusted, and the applicability of the equipment constraints is verified. Finally, the contingency plan templates of multiple similar scenarios are weighted and fused to supplement the proprietary operation steps for specific scenarios.

[0117] It should be noted that the generation of contingency plans can adopt a contingency plan generation model; this contingency plan generation model is a contingency plan framework generation model based on knowledge graph and case reasoning. In the knowledge graph G=(V, E), the vertex set V contains entities such as equipment, operation and status, and the edge set E represents the relationship between entities. Entities and relationships are mapped to a low-dimensional vector space through graph embedding technology, and the entity vector vi and the relationship vector r satisfy vi'=vj+r. The case reasoning part adopts the attention mechanism to calculate the similarity weight αi=softmax(f(vcurrent,vi)) between the current scenario and the historical case. The model training uses a contrastive learning method to minimize the vector distance of positive sample pairs and maximize the vector distance of negative sample pairs. When the fault scenario feature vector is input, the model generates a contingency plan framework that meets the electrical rules by weighted combination of the operation sequences of similar cases, while considering the graph structure constraints. For example, when faced with a new complex fault scenario, the model can extract and integrate effective handling steps from multiple similar cases.

[0118] In actual applications, we may encounter incomplete scenario features or inappropriate contingency plan templates. The intelligent decision support system utilizes feature completion and template optimization methods: first, a correlation model of scenario features is established, using known features to infer missing features; then, the retrieved contingency plan templates are evaluated for their adaptability; and finally, local adjustments are made to optimize the contingency plan framework. For example, if a new fault scenario lacks some device parameters, the system can estimate these parameters based on historical data from similar devices and expert rules, and adjust the operation steps in the contingency plan template accordingly.

[0119] S209: Read the conductor reactance resistance, transformer ratio, and switchgear status from the power grid primary equipment parameter library to construct a branch electrical parameter table.

[0120] The primary equipment parameter library represents a database storing the electrical parameters of the main power system equipment. Conductor reactance and resistance refer to the electrical characteristic parameters of the transmission line. Transformer ratio represents the ratio of the transformer's primary to secondary voltages. Switchgear status represents the on / off status of devices like circuit breakers and disconnectors. The branch electrical parameter table is a collection of data describing the electrical characteristics of each power grid branch.

[0121] Before performing load flow calculations, the intelligent decision support system must establish a complete electrical parameter model. Specifically, the system first connects to the equipment parameter library to read the resistance, reactance, and capacitance values ​​of all transmission lines. It then obtains the transformer's rated capacity, turns ratio, and impedance parameters. It then reads the real-time status of the switchgear to determine the actual network topology. Finally, these parameters are organized into a standard format to generate a branch parameter table for power flow calculations.

[0122] In some embodiments, electrical parameter extraction and organization can be achieved through a variety of methods: optionally, a distributed parameter model can be used to describe the electrical characteristics of long-distance transmission lines, taking into account electromagnetic transient effects; optionally, an equivalent circuit model can be used to represent the electrical characteristics of the transformer, including the excitation branch and leakage reactance. It is understood that other modeling methods can also be used to express electrical characteristics, which are not limited here. It should be noted that the parameter model should take into account the operating temperature and environmental factors of the equipment.

[0123] In practical applications, problems may arise with parameter uncertainty or dynamic parameter changes. Intelligent decision support systems employ a parameter processing strategy: first, a parameter measurement error model is established to assess parameter uncertainty; then, key parameters are corrected using state estimation techniques; and finally, a dynamic parameter update mechanism is established to track changes in equipment characteristics. For example, for a transmission line, the system needs to consider the impact of conductor temperature on resistance.

[0124] S210: Read voltage limit, current limit, and power limit from the grid secondary equipment parameter library to establish the equipment load boundary.

[0125] The secondary device parameter library represents a database storing the setting values ​​of protection and measurement and control devices. The voltage limit represents the voltage range within which the device is permitted to operate. The current limit represents the maximum current allowed to flow through the device. The power limit represents the transmission capacity constraint of the device. The device load limit represents the set of constraints for the safe operation of the device.

[0126] After determining the electrical parameters, the intelligent decision support system needs to establish a complete operational constraint system. Specifically, the intelligent decision support system first reads the operating limits for each voltage level, including normal and allowable deviation ranges. It then obtains the rated current and overload capacity of the line and transformer. It then reads power-related constraints, including active power limits and reactive power ranges. Finally, these limits are converted into standard mathematical constraints.

[0127] In some embodiments, load constraints can be modeled in a variety of ways: optionally, a time-segmented constraint model can be used to account for changes in the load capacity of the device over different time periods; optionally, a fuzzy constraint model can be used to describe the soft characteristics of the constraint through a membership function. It is understood that other constraint modeling methods can also be used to express device limits, and these are not limited here. It should be noted that the constraint model should consider the dynamic characteristics of the device.

[0128] In practical applications, constraints may conflict or be overly strict. Intelligent decision support systems employ a constraint coordination approach: first, a constraint priority system is established to distinguish between hard and soft constraints; then, a multi-objective optimization approach is used to coordinate the different constraints; and finally, constraint margins are adjusted based on system operating experience. For example, for a transformer, the system needs to balance overload capacity with equipment lifespan.

[0129] S211. According to the branch electrical parameter table and the equipment load boundary, determine the associated equation group including the node injection power and the branch power flow, and generate the load flow calculation formula.

[0130] Node injection power represents the power generation and load at each node. Branch power flow refers to the power flow on each branch. The associated equations represent the set of mathematical equations that describe the steady-state operation of the power system. The load flow calculation formula is a standard mathematical model for determining the system's operating state.

[0131] After obtaining electrical parameters and constraints, the intelligent decision support system needs to establish a complete calculation model. Specifically, the intelligent decision support system first creates a node admittance matrix to describe the electrical connections of the network. It then establishes a power balance equation based on Kirchhoff's laws. It then converts device constraints into inequality constraints. Finally, it constructs a standard form load flow calculation model.

[0132] In some embodiments, the load flow model can be established in a variety of ways: optionally, the rectangular coordinate method can be used to construct the equation system, which facilitates constrained linearization processing; optionally, the polar coordinate method can be used to construct the equation system, which facilitates the understanding of the physical meaning. It is understood that other modeling methods can also be used to implement load flow calculations, which are not limited here. It should be noted that the calculation model should consider the efficiency and stability of the numerical solution.

[0133] In practical applications, equations may be difficult to solve or computationally inefficient. Intelligent decision support systems employ a model optimization approach: first, sparse matrix technology is used to optimize the storage structure; then, decoupling algorithms are employed to simplify the computational process; and finally, fast algorithms are employed to improve solution efficiency. For example, for large-scale power grid systems, a partitioned solution strategy is required to increase computational speed.

[0134] It should be noted that the core formula system for load flow calculation is based on the node power equation. For any node i, its injected power can be expressed as:

[0135] Pi=Vi 2 Gii+Vi∑(Vj|Yij|cos(θij+δj-δi));

[0136] and, Qi=-Vi 2 Bii+Vi∑(Vj|Yij|sin(θij+δj-δi));

[0137] Where Vi and δi represent the node voltage amplitude and phase angle respectively, and Yij=Gij+jBij is the node admittance matrix element.

[0138] To solve this set of nonlinear equations, the Newton-Raphson iteration method is used to establish the iterative equation:

[0139] [ΔP / V]=[HN][Δδ] and [ΔQ / V]=[JL][ΔV / V],

[0140] The Jacobian matrix elements include:

[0141] Hij=ViVj|Yij|sin(θij+δj-δi),

[0142] Nij=ViVj|Yij|cos(θij+δj-δi),

[0143] Jij=-ViVj|Yij|sin(θij+δj-δi),

[0144] And Lij=ViVj|Yij|cos(θij+δj-δi), etc., by calculating the correction amount:

[0145] Δδi=ΔPi / (dPi / dδi) and ΔVi / Vi=ΔQi / (dQi / dVi) are iteratively updated until the power imbalance is less than the preset convergence accuracy, and finally the steady-state operation solution of the system is obtained.

[0146] In some embodiments, the intelligent decision support system will construct a standardized load flow calculation model, that is, the intelligent decision support system will substitute the conductor reactance resistance and transformer ratio in the branch electrical parameter table into the node admittance equation to generate an initial relationship between the node voltage and the injected power; the initial relationship will be modified according to the state of the switching device, and the topological constraints of the circuit breaker and the disconnector will be added to obtain a modified relationship; the equipment load boundary will be converted into a voltage amplitude constraint and a phase angle difference constraint, and added to the modified relationship to obtain a complete relationship; the complete relationship will be algebraically transformed to obtain a standard form of the load flow calculation formula.

[0147] The node admittance equation is a mathematical equation describing the relationship between grid node voltage and injected power. The initial equation refers to the basic power flow equation without considering the switching state. Topological constraints describe the impact of the on / off state of switching devices on the network structure. The modified equation refers to the power flow equation after considering the switching state. The voltage amplitude constraint specifies the allowable range of node voltages. The phase angle difference constraint specifies the maximum difference in voltage phase angles between adjacent nodes. The complete equation refers to the power flow equation system that includes all constraints. The standard form is a standardized mathematical expression suitable for numerical solution.

[0148] Before performing load flow calculations, the intelligent decision support system must establish a complete mathematical model. Specifically, the system first substitutes branch parameters into the node admittance equation to establish a basic relationship between node voltage and injected power. The network topology is then modified based on the switch device status, and switch on / off constraints are introduced into the admittance matrix. The device operating limits are then converted into mathematical constraints, including voltage per unit range and phase angle stability margin. The complete relationship also needs to consider power balance constraints and branch flow constraints. Finally, through matrix transformation and equation normalization, a standard calculation formula is derived that is easy to solve.

[0149] In some embodiments, the establishment of the load flow calculation model can be achieved in a variety of ways: optionally, the node voltage method is used for modeling, first constructing a node admittance matrix to represent the network structure, then introducing a power balance equation to describe the node characteristics, and finally solving the nonlinear equations through Newton iteration; optionally, the branch current method is used for modeling, first establishing a loop voltage equation group, then introducing Kirchhoff's law constraints, and finally solving the linear equation group through matrix decomposition. It is understandable that other modeling methods can also be used to achieve load flow calculation, which is not limited here. It should be noted that the model establishment process should take into account computational efficiency and numerical stability.

[0150] In practical applications, poor convergence of equation solutions may be encountered. The intelligent decision support system employs an improved solution strategy: First, a reasonable estimate of initial values ​​is made, using historical calculation results or typical operating modes to set the starting point. Then, an improved Newton method is used for iterative solution, incorporating step-size control and damping factors. Finally, the reliability of the results is verified using convergence criteria. For example, when the system load level is high, traditional methods may encounter convergence difficulties. In this case, the continuous power flow method can be used to gradually increase the load to ensure the stability of the calculation process.

[0151] S212. Perform load flow calculation on the power transfer plan in the plan framework, and obtain the target plan according to the equipment load constraints.

[0152] Referring to step S106 , the intelligent decision support system performs load calculation and generates a target plan.

[0153] S213. After verifying that the plan execution data of the target plan meets the expected execution result, the target plan is stored in the plan knowledge base.

[0154] Referring to step S107, the intelligent decision support system will verify the execution effect of the plan and store the effective plan in the knowledge base.

[0155] In the embodiments of this application, due to the use of innovative technologies such as feature vector-based scenario expression, dynamic knowledge base construction, and multi-constraint plan optimization, it is possible to achieve accurate description of fault scenarios, intelligent retrieval of similar cases, and automatic generation of plan frameworks. This effectively solves the problems existing in traditional plan management systems, such as template rigidity, insufficient constraint verification, and poor scenario adaptability, and thus realizes the intelligent generation of power grid fault handling plans. Specifically, through the standardized extraction and correlation analysis of fault feature vectors, the expressiveness of scenario features is improved; through the construction and updating of a dynamic knowledge base, the reuse efficiency of plan knowledge is enhanced; through the optimization of plans under multi-constraint conditions, the technical feasibility of the scheme is guaranteed; and through scenario expansion and simulation verification, the coverage and reliability of the plans are improved.

[0156] The following describes the intelligent decision support system in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , which is a schematic diagram of the physical device structure of the intelligent decision support system in an embodiment of the present application.

[0157] It should be noted that Figure 3 The structure of the intelligent decision support system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0158] like Figure 3As shown, the intelligent decision support system includes a CPU 301, which can perform various appropriate actions and processes according to the programs stored in the ROM 302 or the programs loaded from the storage unit 308 into the RAM 303, such as executing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An I / O interface 305 is also connected to the bus 304.

[0159] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, push button switches, and the like; an output section 307 including a liquid crystal display (LCD), an audio output device, indicator lights, and the like; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 310 as needed, so that computer programs read from the removable media can be installed in the storage section 308 as needed.

[0160] In particular, according to an embodiment of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present invention includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 309 and / or installed from the removable medium 311. When the computer program is executed by the CPU 301, the various functions defined in the present invention are performed.

[0161] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings.

[0162] Specifically, the intelligent decision support system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the scheduling accident plan writing method based on anti-accident simulation provided by the above embodiment is implemented.

[0163] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the intelligent decision support system described in the above embodiments, or may exist independently and not be incorporated into the intelligent decision support system. The storage medium carries one or more computer programs, which, when executed by a processor of the intelligent decision support system, enable the intelligent decision support system to implement the method for writing a dispatch accident plan based on anti-accident simulation as provided in the above embodiments.

[0164] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0165] As used in the above embodiments, the term “when” may be interpreted to mean “if” or “after” or “in response to determining that” or “in response to detecting that”, depending on the context. Similarly, the phrases “upon determining that” or “if (stated condition or event) is detected” may be interpreted to mean “if determining that” or “in response to determining that” or “upon detecting (stated condition or event)” or “in response to detecting (stated condition or event)”, depending on the context.

Claims

1. A method for compiling a dispatch accident plan based on anti-accident simulation, characterized in that: Applied to an intelligent decision support system, the method comprises: Extract the fault type, faulty equipment, impact scope, and handling process from historical power grid fault data to generate a standard fault data set; Extracting the power grid topology, equipment operating status, and load distribution parameters from the fault data set to generate a fault feature vector; Extracting the key nodes, operation sequence and transfer path of the fault data set in the handling process, feature-associate the fault feature vector with the handling process, and generate a plan knowledge base; the steps of extracting the key nodes, operation sequence and transfer path of the fault data set in the handling process, feature-associate the fault feature vector with the handling process, and generate a plan knowledge base specifically include: reading the equipment state change points from the handling process of the fault data set, reordering the key nodes according to the power grid topology structure, and obtaining a corrected state change sequence; reorganizing the operation instructions according to electrical causal relationships to obtain an operation instruction set; when extracting the transfer path in the power grid topology structure, performing de-loop calculation on the ring network structure in the transfer path to obtain a single path set; performing data association operation on the state change sequence, the operation instruction set, the single path set and the fault feature vector to generate a plan knowledge base; Perform simulation combinations based on fault data sets to generate extended fault scenarios; Extracting target feature vectors from the expanded fault scenario and generating a contingency plan framework in combination with the contingency plan knowledge base; Perform load flow calculation on the power transfer scheme in the plan framework, and obtain the target plan according to the equipment load constraint; After verifying that the plan execution data of the target plan meets the expected execution result, the target plan is stored in the plan knowledge base.

2. The method according to claim 1, characterized in that The step of extracting the target feature vector from the expanded fault scenario and generating a contingency plan framework in combination with the contingency plan knowledge base specifically includes: Extract electrical topology features, equipment health features, and load distribution features from the fault scenario to generate a target feature vector; Calculate feature similarity based on topological distance, load similarity, and equipment type matching, and retrieve similar scenarios from the plan knowledge base; Extracting reconstruction sequences, transfer paths, and equipment constraints from the emergency plan templates of similar scenarios to generate a basic emergency plan framework; The basic emergency plan framework is integrated and supplemented according to the retrieved multiple similar scenarios to generate an emergency plan framework.

3. The method according to claim 1, characterized in that Before the step of performing load flow calculation on the power transfer plan in the plan framework and obtaining a target plan according to equipment load constraints, the method further includes: Read the conductor reactance, transformer ratio, and switchgear status from the primary equipment parameter library of the power grid to construct a branch electrical parameter table; Read voltage limit, current limit and power limit from the grid secondary equipment parameter library to build the equipment load boundary; According to the branch electrical parameter table and the equipment load boundary, a group of associated equations including node injection power and branch power flow is determined to generate a load flow calculation formula.

4. The method according to claim 3, characterized in that The step of determining a set of associated equations including node injection power and branch power flow based on the branch electrical parameter table and the equipment load boundary, and generating a load flow calculation formula specifically includes: Substituting the conductor reactance resistance and transformer ratio in the branch electrical parameter table into the node admittance equation to generate an initial relationship between the node voltage and the injected power; Modifying the initial relational expression according to the state of the switchgear, adding topological constraints of the circuit breaker and the disconnector, and obtaining a modified relational expression; Converting the equipment load boundary into a voltage amplitude constraint and a phase angle difference constraint, and adding them to the corrected relational expression to obtain a complete relational expression; Performing algebraic transformation on the complete relational expression, a standard form of load flow calculation formula is obtained.

5. The method according to claim 1, wherein After the step of performing simulation combination based on the fault data set to generate an extended fault scenario, the method further includes: Extracting the combination relationship of faulty devices from the expanded fault scenario and constructing fault propagation link data; Classify and organize the fault scenarios according to the fault propagation link to generate a scenario feature group; Common data in the scene feature group is extracted to generate a scene evolution rule set.

6. The method according to claim 5, characterized in that After the step of classifying and arranging the fault scenarios according to the fault propagation links to generate a scenario feature group, the method further includes: Extracting fault impact range data from the scenario feature group and marking the boundary points affected by the fault; Divide the fault isolation areas according to the boundary points, and calculate the electrical connection data between the fault isolation areas; Mapping the electrical connection data to a fault propagation link to obtain a new fault path; An electrical rule check is performed on the newly generated fault path to obtain newly generated scenario data.

7. An intelligent decision support system, characterized in that: The intelligent decision support system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the intelligent decision support system to execute the method described in any one of claims 1-6.

8. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on an intelligent decision support system, the intelligent decision support system is caused to execute the method according to any one of claims 1 to 6.

9. A computer program product, characterized in that When the computer program product is run on an intelligent decision support system, the intelligent decision support system is enabled to perform the method according to any one of claims 1 to 6.

Images