Knowledge graph driven power market clearing combination scenario generation method and system

Abstract

The application belongs to the technical field of power systems, and provides a power market clearing combination scenario generation method and system driven by a knowledge graph, comprising: acquiring real-time data of power market clearing and a power market knowledge graph; mapping the acquired real-time data to entities in the knowledge graph to obtain a power market entity set; performing risk mode identification of the knowledge graph and the entity set based on subgraph matching to obtain a power market knowledge subgraph set; calculating a triggering probability of a current real-time data risk mode in the obtained power market knowledge subgraph set to screen a generated scenario of the power market; constructing a target function of scenario parameters according to the screened generated scenario, optimizing the constructed target function, and completing generation of a power market clearing combination scenario.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, specifically relating to a method and system for generating power market clearing combination scenarios based on knowledge graphs. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Market clearing is a core decision-making process in power system operation, and the robustness of its outcome directly affects the safety and stability of the power grid and the economic efficiency of the market. Currently, clearing models based on Security-Constrained Unit Combination (SCUC) and Security-Constrained Economic Dispatch (SCED) are widely adopted. However, in practical applications, setting up comprehensive and effective test scenarios to verify the robustness of the clearing results remains a significant challenge.

[0004] Scenario construction relies on human experience and lacks comprehensive coverage: Currently, test scenarios (such as N-1 fault sets, load fluctuation ranges, etc.) are mostly manually set by dispatchers based on their personal experience. This method is not only labor-intensive, but more seriously, it is difficult to avoid subjectivity and limitations, and it is very easy to miss extremely complex scenario combinations with "low probability and high loss", forming safety hazards. As the proportion of new energy sources in the new power system continues to increase, the randomness and volatility on both the source and load sides are intensified, and it is difficult for humans to exhaustively cover the massive number of potential risk combinations.

[0005] Tacit knowledge cannot be effectively utilized, and decision-making lacks a basis: Over the years, the power system has accumulated a large amount of historical operating data, accident reports, anti-accident drill plans and dispatching procedures, which contain valuable expert experience and decision-making knowledge; however, this knowledge is mostly scattered in different systems in the form of unstructured text, and cannot be directly called and applied by the clearing algorithm, forming "knowledge silos".

[0006] Static pre-defined scenarios cannot adapt to real-time changes, resulting in delayed responses: Existing clearing systems typically use pre-set static fault sets and scenario libraries, which cannot be dynamically adjusted based on the latest weather forecasts, load predictions, equipment status, and other boundary conditions. When emergencies occur (such as extreme weather warnings or temporary shutdowns of critical equipment), the system cannot automatically identify risks and generate corresponding test scenarios, lacking real-time awareness and proactive defense capabilities.

[0007] The poor interpretability of generated scenarios and the difficulty of human-machine collaboration: Even if some scenarios are generated through optimized algorithms, their decision-making logic remains a "black box" for schedulers. Operators struggle to quickly understand why a specific scenario needs to be calculated and cannot trace its generation basis, thus reducing human trust in machine decision-making and hindering effective human-machine collaboration.

[0008] Scenarios can be generated by introducing optimization algorithms or random sampling, but this is limited to mathematical completeness and lacks guidance from business knowledge. It is easy to generate a large number of scenarios that are mathematically valid but irrelevant to business, resulting in low computational efficiency.

[0009] Therefore, there is an urgent need for an intelligent scenario generation method that can transform historical experience into knowledge, structure business requirements, and respond to changes in power grid status in real time, so as to achieve a leap from "manually setting scenarios" to "intelligent scenario perception and generation". Summary of the Invention

[0010] To address the aforementioned issues, this invention proposes a knowledge graph-driven method and system for generating combined scenarios of electricity market clearing. This method leverages a knowledge graph based on a large language model to fully realize human-computer interaction and knowledge-driven processes, thereby enabling the generation of intelligent scenarios.

[0011] According to some embodiments, the first solution of the present invention provides a method for generating a combined scenario of electricity market clearing based on knowledge graphs, which adopts the following technical solution:

[0012] A knowledge graph-driven method for generating combined scenarios of electricity market clearing includes:

[0013] Obtain real-time data on electricity market clearing and an electricity market knowledge graph;

[0014] The acquired real-time data is mapped to entities in the knowledge graph to obtain a set of entities in the electricity market.

[0015] Risk pattern recognition of knowledge graphs and entity sets is performed based on subgraph matching to obtain a set of knowledge subgraphs for the electricity market.

[0016] Calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs, and filter the generation scenarios of the electricity market;

[0017] Based on the selected generation scenarios, construct the objective function for the scenario parameters, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario.

[0018] As a further technical limitation, the constructed scene parameter objective function for ;in, The system severity function representing the scenario is, i.e. ;in, The sub-objective function representing the severity of the load is... The weighting coefficients represent the sub-objective function of load severity; The sub-objective function represents the severity of new energy fluctuations. The weighting coefficients of the sub-objective function representing the severity of new energy fluctuations; The sub-objective function representing the severity of topological impact is... The weighting coefficients represent the sub-objective function for the severity of topological shock; This represents the sub-objective function representing the expected severity of default; These are the weighting coefficients of the sub-objective function for the expected severity of default.

[0019] As a further technical limitation, the constraints of the constructed scene parameter objective function include at least physical constraints, data perturbation constraints, temporal correlation constraints, and topological rationality constraints.

[0020] As a further technical limitation, the selected generated scenarios are transformed into test scenarios, wherein the transformed test scenarios are composed of adjustable boundary condition parameters. Definition, that is ;in, This represents the load curve for the scenario to be generated; This indicates the available output of conventional units in the scenario to be generated; This indicates the predicted output of new energy sources in the scenario to be generated; This represents the power grid topology in the scenario to be generated; This indicates the set of additional security constraints that need to be added.

[0021] As a further technical limitation, in the process of risk pattern recognition of knowledge graphs and entity sets based on subgraph matching, the comprehensive risk confidence of candidate knowledge subgraphs is calculated based on graph theory algorithms. The conditional probability is calculated based on the obtained comprehensive risk confidence. Candidate knowledge subgraphs with conditional probabilities exceeding the risk threshold are screened. In the power market knowledge graph, the candidate knowledge subgraphs that are most relevant to the obtained power market entities and have the highest risk are found to complete the risk pattern recognition and obtain the power market knowledge subgraph set.

[0022] As a further technical limitation, the trigger probability is a function based on the degree of matching between real-time data and pattern conditions, i.e., the trigger probability. for ];in, It indicates the degree of agreement between real-time data and entity conditions in the model; Representation pattern The frequency of occurrence; This represents the base rate.

[0023] As a further technical constraint, the generated electricity market clearing combination scenario is a complete generation chain, including at least source data, knowledge path, risk contribution, parameter comparison, and source tracing.

[0024] As a further technical limitation, the mapping rules in the entity mapping process include at least threshold rules, pattern matching rules, topology analysis rules, and weather rules.

[0025] According to some embodiments, the second aspect of the present invention provides a knowledge graph-driven power market clearing combination scenario generation system, which adopts the following technical solution:

[0026] A knowledge graph-driven power market clearing scenario generation system includes:

[0027] The acquisition module is configured to acquire real-time data on electricity market clearing and an electricity market knowledge graph.

[0028] The mapping module is configured to map the acquired real-time data to entities in the knowledge graph to obtain a set of electricity market entities.

[0029] The identification module is configured to perform risk pattern recognition of knowledge graphs and entity sets based on subgraph matching, thereby obtaining a set of knowledge subgraphs for the electricity market.

[0030] The filtering module is configured to calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs and filter the generation scenarios of the electricity market.

[0031] The generation module is configured to construct an objective function for scenario parameters based on the selected generation scenarios, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario.

[0032] According to some embodiments, the third aspect of the present invention provides a computer program product, which adopts the following technical solution:

[0033] A computer program product includes software code, wherein the program in the software code performs the steps of the knowledge graph-driven power market clearing combination scenario generation method as described in the first aspect of the present invention.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This invention solidifies the implicit expert experience accumulated in historical reports and procedures into computable structured knowledge through knowledge graphs; each generated scenario originates from historically real cases or explicitly defined rules, ensuring the rationality of the scenario and its high relevance to business, and eliminating invalid calculations.

[0036] This invention can receive the latest forecasts, weather, topology and other data in real time, and automatically map and activate high-risk patterns in the knowledge graph, realizing "data-driven" risk perception and scenario generation. This enables the system to make pre-decision for the most pressing risk points and truly achieve proactive defense.

[0037] Based on the reasoning ability of graphs, this invention can automatically and intelligently combine risk factors from multiple dimensions such as time (e.g., evening rush hour), weather (e.g., sudden drop in photovoltaic power), equipment (e.g., line breakage), and network (e.g., cross-section exceeding limits) to generate "extreme yet reasonable" test scenarios, greatly improving the ability to discover complex cascading risks.

[0038] This invention's automated process significantly improves scenario management efficiency. At the same time, since the generated scenarios are all high-risk and critical, limited computing power is concentrated on the "cutting edge," achieving the greatest security gain with the least amount of computation. This avoids wasting computing power on irrelevant scenarios, thereby improving clearing calculation efficiency, reducing operating costs, and enhancing power grid reliability. Attached Figure Description

[0039] The accompanying drawings, which form part of this embodiment, are used to provide a further understanding of this embodiment. The illustrative embodiments and their descriptions are used to explain this embodiment and do not constitute an improper limitation of this embodiment.

[0040] Figure 1 This is a flowchart of the knowledge graph-driven method for generating combined electricity market clearing scenarios in Embodiment 1 of the present invention;

[0041] Figure 2 This is a structural block diagram of the knowledge graph-driven power market clearing combination scenario generation system in Embodiment 2 of the present invention. Detailed Implementation

[0042] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0043] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0044] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0045] In this invention, terms such as "upper," "lower," "left," "right," "front," "back," "vertical," "horizontal," "side," and "bottom" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used only to facilitate the description of the structural relationships of the various components or elements of this invention and do not specifically refer to any component or element in this invention. They should not be construed as limiting the invention.

[0046] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0047] Example 1

[0048] Embodiment 1 of this invention introduces a method for generating a combined scenario of power market clearing based on knowledge graph.

[0049] like Figure 1 The method shown is a knowledge graph-driven approach for generating a combination of electricity market clearing scenarios, including:

[0050] Obtain real-time data on electricity market clearing and an electricity market knowledge graph;

[0051] The acquired real-time data is mapped to entities in the knowledge graph to obtain a set of entities in the electricity market.

[0052] Risk pattern recognition of knowledge graphs and entity sets is performed based on subgraph matching to obtain a set of knowledge subgraphs for the electricity market.

[0053] Calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs, and filter the generation scenarios of the electricity market;

[0054] Based on the selected generation scenarios, construct the objective function for the scenario parameters, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario.

[0055] As one or more implementation methods, this embodiment aims to achieve intelligent scene generation based on knowledge graphs. This requires the real-time inflow of diverse data from the real world to be mapped into standardized entities that the knowledge graph can "understand" and process, providing "trigger points" for subsequent graph reasoning.

[0056] Assume that an online knowledge graph has been formed by training an open-source large language model (LLM). ,Right now Where V represents the set of relationships between nodes in the knowledge graph; It represents the set of all nodes in a knowledge graph.

[0057] The real-time data stream of key elements for electricity market clearing is now defined. It is a heterogeneous tuple, specifically in the form of ;in, This represents the load forecast curve for a future period of time (such as 96 points in a day), and is a time-power series. The representative new energy forecast curve is also a time-power series. The power grid topology for time period t can be represented by a graph. (bus, branch line, transformer, etc.) indicates that some of the lines or transformers are disconnected. This indicates equipment status information, including a list of planned maintenance. Unplanned service suspension list ; This represents weather forecast and warning information, such as wind speed, irradiance, typhoon path, and rainstorm warning level.

[0058] This embodiment will use real-time data streams. Mapping data to entities in a knowledge graph is essentially a function that maps real-time data. Elements in the graph are mapped to the knowledge graph. Entity set in Above, its form is ;in, Represents a mapping function; Represents a set of mapping rules. It is a single rule within it; It is a professional LLM that has already been used for the map. Extracted single entity; Represents real-time data stream The components in.

[0059] To ensure that the above mapping has business significance for power clearing, this embodiment provides the following example of mapping rules:

[0060] (1) Threshold rule

[0061] If the load forecast value Exceeding historical threshold (e.g., 95th percentile) then maps to an entity with "extremely high load," specifically... ;

[0062] (2) Pattern matching rules

[0063] If the photovoltaic prediction curve During the evening period The rate of decline exceeds the threshold This is then mapped to the entity "sudden drop in photovoltaic power," specifically as follows: ;

[0064] (3) Topology analysis rules

[0065] Analyze the current topology Identify all N-1 components (lines, transformers) that are in an open state. For each such component... This is mapped to the entity "[Component Name] N-1 Interrupt" (e.g., "500kV Mulan Line N-1 Interrupt"), as detailed below. ;in, This represents the N-1 element entities that are in a disconnected state when mapped to the map.

[0066] (4) Weather rules

[0067] Typhoon warning level A yellow alert or higher is issued, and the predicted path affects an area of ​​[missing information]. With photovoltaic power station If there is an overlap, it maps to the entities "typhoon weather" and "extreme weather in photovoltaic power generation areas," as detailed below: ;in, This represents the logical operation "AND". This represents the empty set. and These are two entities that are listed side by side.

[0068] Based on the above rules and steps, the mapping function provides the following set of real-time entities: Each entity They are all knowledge graphs A node that already exists in the array; m is... The total number of entities perceived in the process; This represents all known risk factors or critical states that were "activated" in the previous system state.

[0069] This embodiment will utilize graph theory algorithms in knowledge graphs. Searching for and real-time entity sets The most relevant and highest-risk knowledge subgraph (i.e., historical experience patterns).

[0070] First, this embodiment will Considered as a set of "seed" nodes, in the knowledge graph Find the neighboring nodes directly connected to these seed nodes and their relationships, forming an initial, small connected subgraph. It can be regarded as an "anchor point" for the current real-time situation in the ocean of knowledge.

[0071] Using the graph query language Cypher, this embodiment does not directly search. The search algorithm for finding the neighbors of [the target] is given below through multi-hop traversal:

[0072] ① Define a traversal path An alternating sequence of nodes and edges: ;in, Representing a connected subgraph middle All nodes on; This represents the edge between the aforementioned nodes; k is the search step size.

[0073] ②From Starting from the beginning, all nodes and edges that satisfy the following conditions constitute a candidate subgraph set. ;in, Represents the set of candidate subgraphs; The initial node represents the candidate subgraph; Represents the end node of the candidate subgraph; Indicates the path length; This indicates the maximum number of hops.

[0074] Given a set of candidate subgraphs This does not mean that all subgraphs are equally important to the risk decision in this embodiment. This embodiment needs to screen and rank them. To achieve this goal, this embodiment uses risk confidence to construct the following indicators. ;in, express Subgraphs in; This indicates its overall risk confidence level; Represents conditional probability, measuring the probability of observing real-time conditions. Under this risk model The likelihood of it happening; express The degree of impact measures the severity of the consequences should the risk pattern occur. express The connectivity strength of the subgraph measures its connection to the real-time entity set. The degree of closeness of the connection; , , These are the weighting coefficients of these three items.

[0075] Conditional probability A simplified calculation method is to use the geometric mean or minimum of the confidence scores of the relationships in the subgraph, i.e. ;in, The triples (subject, relation, object) in the knowledge graph are extracted from the text by LLM; The confidence level of the triple is given by the trained LLM.

[0076] And the degree of impact Consequences can be obtained from the attributes of the "consequences" entity nodes associated in the knowledge graph, or quantified from loss data in historical events. ;in, Loss estimates representing the consequences of an event can be obtained from accident data in a knowledge graph; This refers to the record marked as "consequence" in the diagram.

[0077] You can use all nodes in the subgraph to It is measured by the reciprocal of the average shortest path to any node. The shorter the path, the stronger the connectivity. ;in, express Middle node to The path length of entity e in the middle; express The set of edges; express The average side length.

[0078] After completing the above assessment process, this embodiment can set a risk threshold. And select the K subgraphs with the highest scores to form a high-risk pattern set, i.e. .

[0079] This embodiment can obtain a set of high-risk knowledge subgraphs sorted by risk. A subgraph Both represent a condition that is currently in effect. The historical experience or risk rules that are triggered and have high confidence and high impact will directly drive the next step—the generation of combined scenarios.

[0080] This embodiment yields a set of high-risk knowledge subgraphs ranked by risk. Now, in this embodiment, we need to consider each high-risk subgraph. Calculate a comparable, comprehensive risk score to determine the priority of scenario generation.

[0081] This embodiment requires calculating the trigger probability, that is, based on the current real-time data. Below, risk model What is the likelihood of it actually occurring? The trigger probability is a function based on the degree of matching between real-time data and pattern conditions, as detailed below. ];in, This indicates the degree of agreement between real-time data and entity conditions in the model. For example, if the model includes "extremely high load," then the current load forecast is calculated. Compared with historical extremes Normalized differences; This indicates the pattern. The frequency of occurrence refers to how often a pattern (or a similar pattern) has occurred in the past, statistically analyzed from historical data. A higher frequency generally indicates a greater prior probability of its current occurrence. The base rate represents the inherent probability of certain events (such as equipment failure rate), which can be obtained from an equipment reliability database.

[0082] Regarding the trigger probability, this embodiment needs to measure the severity of its corresponding consequences, and therefore constructs the following indicators. ;in, For historical impact, actual loss data, such as load shedding (MWh), economic loss (ten thousand yuan), and social impact level, can be obtained from the historical event nodes associated with the knowledge graph; This represents the potential cost of the risk model, indicating the cost that the event might cause if it occurred in the current system state.

[0083] For example, shadow prices from clearing models can be used or estimated through rapid simulations. This indicates the safety severity of the risk mode. Based on power system safety and stability standards, it determines whether the mode will lead to system over-limits, stability disruptions, etc. A severity level can be defined (e.g., 0-Normal, 1-Warning, 2-Danger, 3-Crisis).

[0084] Based on the two indicators mentioned above, the following comprehensive risk score can be given: .

[0085] Similarly, this embodiment follows the method for identifying high-risk pattern sets, and... The elements in are arranged according to The scores are sorted from highest to lowest, and a threshold is set. The top M scenes with scores greater than the threshold are selected as the generated scenes. ;in, This represents the overall risk score threshold for scenario selection.

[0086] This embodiment requires generating a scene. The abstract risk is transformed into a specific, quantifiable test scenario that the clearing software can execute. This is essentially a constrained optimization problem (COP).

[0087] This embodiment sets up a scenario. Consists of a set of adjustable boundary condition parameters The definition, in the form of ;in, This represents the load curve for the scenario to be generated; This indicates the available output of conventional units in the scenario to be generated; This indicates the predicted output of new energy sources in the scenario to be generated; Indicates the power grid topology in the scenario to be generated (e.g., which components are disconnected); This indicates the set of additional safety constraints that need to be added (such as line power flow limits and backup requirements).

[0088] The goal of generating scenarios is to reproduce risk patterns to the greatest extent possible within a reasonable scope. To fully test the robustness of the system under the described severe conditions, the following objective function is constructed. ;in, Represents the optimal parameters for the scenario; The function representing the severity of a scenario can be designed as follows: ;in, , , , For sub-objective functions; - These are the weighting coefficients of the sub-objective function; The severity of the load can be defined as the relative deviation between the peak load in the scenario and the predicted peak load. ;in, Indicates the baseline load; The severity of fluctuations in renewable energy output can be defined as the maximum value of the rate of change in renewable energy output within a given scenario. ; The severity of a topology impact can be defined as its importance level based on topology calculations (such as the number of disconnected lines or their total rated capacity), for example... ;in, Represents the set of interrupted lines; This indicates the power capacity of the interrupted line; This indicates the expected severity of the default; it predicts the extent to which the system will violate safety constraints under this scenario. This can be estimated using a fast, simplified power flow calculation model, such as... ;in, It could be a function that calculates the power flow margin at the critical section, or a trained neural network, with scene parameters as input. Output a score representing the stress level of the system.

[0089] As with other power sector decisions, scenario generation still needs to adhere to constraints to ensure that the generated scenarios are "extreme but reasonable." These constraints include...

[0090] (1) Physical constraints

[0091] The generated parameters must conform to basic physical laws, such as power balance as a fundamental constraint, i.e. ;in, This represents the cumulative line loss under a given scenario topology.

[0092] (2) Data disturbance constraints

[0093] Scenario parameters must be within a reasonable range of perturbation of the baseline values ​​and must not deviate from reality. For example, ;in, This represents the base load during time period t; Represents the scenario load curve The generated value during time period t; This represents the baseline value of new energy power output during time period t; This represents the generated value of the renewable energy output sequence in time period t. and This represents the maximum allowable load disturbance and the allowable disturbance of new energy output estimated based on historical data.

[0094] (3) Temporal correlation constraints

[0095] The adjusted curve must maintain the time-series characteristics of the original data (such as smoothness). For example, the rate of load change between adjacent time points must be within a reasonable range. ;in, This indicates the maximum rate of change of the load.

[0096] (4) Topological rationality constraints

[0097] It must be a feasible subset of the power grid topology; for example, two double-circuit lines on the same tower cannot be disconnected simultaneously unless permitted by the rules.

[0098] This optimization problem is typically complex and can be solved using evolutionary algorithms (such as genetic algorithms) or sequential decision optimization. The solver outputs the optimal scenario parameters. Thus, the scene is fully defined. .

[0099] This embodiment will use the scene generated in the previous part. and their corresponding scene parameters Serialize to a standard format (such as JSON, XML, or a specific database format), and automatically push and integrate into the day-ahead market clearing platform via application programming interfaces (APIs) or message queues (such as Kafka, RabbitMQ) as an additional mandatory computing scenario.

[0100] It should be noted that this embodiment is based on a visual interface, and provides information for each generated scene. This demonstrates the complete generation process, achieving transparency in decision-making; that is...

[0101] (1) Source data, i.e., which real-time data is highlighted. This scene was triggered;

[0102] (2) Knowledge path, that is, to present the knowledge path in a graphical way from To the final risk model The reasoning path. For example: 'load forecast' -> (exceeds) -> 'extremely high load' -> (cooccursWith) -> 'sudden drop in photovoltaic load' -> (causes) -> cross-section exceeding limits';

[0103] (3) Risk contribution, which is displayed in a bar chart or radar chart. , , The numerical value and its composition;

[0104] (4) Parameter comparison: This provides a curve comparison chart, comparing scene parameters. By comparing the load curve with the baseline parameters, the differences can be clearly shown (such as where the load curve is raised and where the renewable energy is reduced).

[0105] (5) Source citation: that is, directly linking back to the original document fragment from which the knowledge was initially extracted, which greatly enhances the credibility and interpretability of the system.

[0106] This embodiment solidifies the implicit expert experience accumulated in historical reports and procedures into computable structured knowledge through knowledge graphs; each generated scenario originates from historically real cases or explicitly defined rules, ensuring the rationality of the scenario and its high relevance to business, and eliminating invalid calculations.

[0107] This embodiment can receive the latest forecasts, weather, topology and other data in real time, and automatically map and activate high-risk patterns in the knowledge graph, realizing "data-driven" risk perception and scenario generation. This enables the system to make pre-decision for the most pressing risk points and truly achieve proactive defense.

[0108] Based on the reasoning capabilities of the graph, this embodiment can automatically and intelligently combine risk factors from multiple dimensions, such as time (e.g., evening rush hour), weather (e.g., sudden drop in photovoltaic power), equipment (e.g., line breakage), and network (e.g., cross-section exceeding limits), to generate "extreme yet reasonable" test scenarios, greatly improving the ability to discover complex cascading risks.

[0109] This embodiment's automated process significantly improves scenario management efficiency. At the same time, since the generated scenarios are all high-risk and critical, the limited computing power is concentrated on the "cutting edge," achieving the greatest security gain with the least amount of computation. This avoids wasting computing power on irrelevant scenarios, thereby improving clearing calculation efficiency, reducing operating costs, and enhancing power grid reliability.

[0110] Example 2

[0111] Embodiment 2 of this invention introduces a knowledge graph-driven power market clearing combination scenario generation system.

[0112] like Figure 2 The system shown is a knowledge graph-driven power market clearing combination scenario generation system, comprising:

[0113] The acquisition module is configured to acquire real-time data on electricity market clearing and an electricity market knowledge graph.

[0114] The mapping module is configured to map the acquired real-time data to entities in the knowledge graph to obtain a set of electricity market entities.

[0115] The identification module is configured to perform risk pattern recognition of knowledge graphs and entity sets based on subgraph matching, thereby obtaining a set of knowledge subgraphs for the electricity market.

[0116] The filtering module is configured to calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs and filter the generation scenarios of the electricity market.

[0117] The generation module is configured to construct an objective function for scenario parameters based on the selected generation scenarios, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario.

[0118] The detailed steps are the same as those provided in Example 1 for generating a combined electricity market clearing scenario based on knowledge graphs, and will not be repeated here.

[0119] Example 3

[0120] Embodiment 3 of the present invention provides a computer program product.

[0121] A computer program product includes software code, wherein the program in the software code performs the steps of the knowledge graph-driven power market clearing combination scenario generation method as described in Embodiment 1 of the present invention.

[0122] The detailed steps are the same as those provided in Example 1 for generating a combined electricity market clearing scenario based on knowledge graphs, and will not be repeated here.

[0123] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0124] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0125] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0126] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0127] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0128] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0129] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A method for generating a combined scenario of electricity market clearing based on knowledge graphs, characterized in that, include: Obtain real-time data on electricity market clearing and an electricity market knowledge graph; The acquired real-time data is mapped to entities in the knowledge graph to obtain a set of entities in the electricity market. Risk pattern recognition of knowledge graphs and entity sets is performed based on subgraph matching to obtain a set of knowledge subgraphs for the electricity market. Calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs, and filter the generation scenarios of the electricity market; Based on the selected generation scenarios, construct the objective function of the scenario parameters, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario; The constructed scene parameter objective function for ;in, The system severity function representing the scenario is, i.e. ;in, The sub-objective function representing the severity of the load is... The weighting coefficients represent the sub-objective function of load severity; The sub-objective function represents the severity of new energy fluctuations. The weighting coefficients of the sub-objective function representing the severity of new energy fluctuations; The sub-objective function representing the severity of topological impact is... The weighting coefficients represent the sub-objective function for the severity of topological shock; This represents the sub-objective function representing the expected severity of default; These are the weighting coefficients for the sub-objective function of expected default severity; The selected generated scenarios are transformed into test scenarios, wherein the transformed test scenarios are composed of adjustable boundary condition parameters. Definition, that is ;in, This represents the load curve for the scenario to be generated; This indicates the available output of conventional units in the scenario to be generated; This indicates the predicted output of new energy sources in the scenario to be generated; This represents the power grid topology in the scenario to be generated; This indicates the set of additional security constraints that need to be added. In the process of risk pattern recognition of knowledge graphs and entity sets based on subgraph matching, the comprehensive risk confidence of candidate knowledge subgraphs is calculated based on graph theory algorithms. The conditional probability is calculated based on the obtained comprehensive risk confidence. Candidate knowledge subgraphs with conditional probabilities exceeding the risk threshold are screened. In the power market knowledge graph, the candidate knowledge subgraphs that are most relevant to the obtained power market entities and have the highest risk are found to complete the risk pattern recognition and obtain the power market knowledge subgraph set. The trigger probability is a function based on the degree of matching between real-time data and pattern conditions, i.e., the trigger probability. for ];in, It indicates the degree of agreement between real-time data and entity conditions in the model; Representation pattern The frequency of occurrence; Indicates the base rate; The generated electricity market clearing scenario is a complete generation chain, including at least source data, knowledge path, risk contribution, parameter comparison, and source tracing. In the process of entity mapping, the mapping rules include at least threshold rules, pattern matching rules, topology analysis rules, and weather rules.

2. The method for generating a combined electricity market clearing scenario based on knowledge graphs as described in claim 1, characterized in that, The constraints of the constructed objective function for scene parameters include at least physical constraints, data perturbation constraints, temporal correlation constraints, and topological rationality constraints.

3. A knowledge graph-driven power market clearing combination scenario generation system, employing the knowledge graph-driven power market clearing combination scenario generation method as described in any one of claims 1-2, characterized in that, include: The acquisition module is configured to acquire real-time data on electricity market clearing and an electricity market knowledge graph. The mapping module is configured to map the acquired real-time data to entities in the knowledge graph to obtain a set of electricity market entities. The identification module is configured to perform risk pattern recognition of knowledge graphs and entity sets based on subgraph matching, thereby obtaining a set of knowledge subgraphs for the electricity market. The filtering module is configured to calculate the trigger probability of the current real-time data risk pattern in the obtained set of electricity market knowledge subgraphs and filter the generation scenarios of the electricity market. The generation module is configured to construct an objective function for scenario parameters based on the selected generation scenarios, optimize the constructed objective function, and complete the generation of the electricity market clearing combination scenario.

4. A computer program product, comprising software code, characterized in that, The program in the software code executes the steps of the knowledge graph-driven power market clearing combination scenario generation method as described in any one of claims 1-2.

Images