An event identification method and system based on a global risk spectrum of a power system

Abstract

This invention belongs to the field of power system risk identification technology, and discloses an event identification method and system based on a power system-wide risk spectrum. The method includes: acquiring a target event for risk identification; inputting the target event into a risk identification model based on the power system-wide risk spectrum for risk identification, obtaining a risk identification result; and outputting the risk identification result corresponding to the target event. The risk identification result is a three-dimensional risk event coordinate and a three-dimensional risk event location map based on risk elements, cycle stages, and risk loss targets. This invention automatically analyzes target events using a risk identification model based on the power system-wide risk spectrum, outputting specific risk identification results. This overcomes the problems of incomplete element coverage, missing cycle dimensions, and incompatibility with insurance semantics in existing risk event identification technologies, providing support for cross-dimensional risk assessment, source tracing analysis, and adaptation to insurance business in the power system.

Description

Technical Field

[0001] This invention relates to the field of power system risk identification, and in particular to an event identification method and system based on the power system's global risk spectrum. Background Technology

[0002] With the high proportion of new energy access, the improvement of power electronics level, the enhancement of source-grid-load-storage synergy, the deep embedding of carbon constraints and the continuous empowerment of digitalization, the system risks of new power have gradually evolved from discrete disturbances around single equipment, single links and local scenarios under traditional conditions to complex systemic risks that are transmitted across subjects, links and cycles.

[0003] Accurately identifying the types of risks in the power system, clarifying the boundaries of responsibility, and defining the forms of loss are the foundation of power system risk governance.

[0004] However, existing financial and insurance businesses still have significant limitations in identifying and analyzing power system risk events: First, risk identification is limited to a single dimension. Existing research mainly identifies risks from the four physical links of source, grid, load, and storage, failing to consider the impact of other factors such as carbon constraints on system risks, resulting in the omission of a large number of non-physical risk events. Second, risk analysis is static. Existing technologies usually perform risk analysis on a specific stage of a certain link, ignoring the dynamic tracking of the evolution and transmission patterns of risk events over time, and failing to predict the diffusion path of risks at different stages. Third, there is a disconnect between engineering and insurance terminology. Risk classification in the power industry revolves around engineering and technical dimensions such as personal safety, equipment safety, and grid safety. The classification system lacks a mapping relationship with the insured objects in the financial and insurance fields, resulting in the identified risk events not being able to directly support the design of insurance products and the determination of claims liability.

[0005] In the existing technology, the identification and analysis of power system risk events by financial and insurance businesses have limitations, therefore, the existing technology needs to be improved. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide an event identification method based on the full-domain risk spectrum of the power system, in order to address the limitations of existing financial and insurance businesses in identifying and analyzing power system risk events.

[0007] The technical solution adopted by this invention to solve the technical problem is as follows: In a first aspect, the present invention provides an event identification method based on the global risk spectrum of a power system, comprising: Acquire the target events for the risks to be identified; The target event is input into a risk identification model based on the power system global risk spectrum for risk identification, and the risk identification result is obtained. Output the risk identification results corresponding to the target event.

[0008] In one implementation, the method for constructing the risk identification model based on the power system's global risk spectrum includes the following steps: Determine a large language model for risk identification; Obtain a set of sample events, and train the large language model using the set of sample events to obtain a risk identification model; wherein, the set of sample events contains sample events, and the sample events are events for which risk identification results exist; Based on the sample event set, a global risk spectrum of the power system is constructed; The risk identification model is optimized based on the constructed power system global risk spectrum to obtain a risk identification model based on the power system global risk spectrum.

[0009] In one implementation, constructing a power system-wide risk spectrum based on the sample event set includes the following steps: The sample events in the sample event set are preprocessed; the preprocessing steps include data cleaning and normalization. Feature extraction is performed on the preprocessed sample events to obtain the element association features, cycle stage features and loss morphology features corresponding to each sample event; Based on the sample events, determine the correlation between the element association features, the periodic stage features, and the loss morphology features; A global risk spectrum for the power system is constructed based on the aforementioned relationships.

[0010] In one implementation, determining the correlation between the element association features, the periodic stage features, and the loss morphology features based on the sample events includes: Obtain the risk element candidate set corresponding to the element association feature, the cycle stage candidate set corresponding to the cycle stage feature, and the risk target candidate set corresponding to the loss pattern feature; Based on the sample events, determine the first association between the candidate set of risk factors and the candidate set of cycle stages; Based on the sample events, a second association relationship is determined between the candidate set of risk elements, the candidate set of cycle stages, and the candidate set of risk targets.

[0011] In one implementation, constructing a power system-wide risk spectrum based on the correlation includes: The risk elements in the candidate set of risk elements are divided into traditional elements and new elements; Based on the first association relationship, the cycle stages of the candidate set of cycle stages are classified to obtain the full life cycle dimension corresponding to the traditional elements and the development cycle dimension corresponding to the new elements. Obtain the types of risk targets from the candidate set of risk targets; Based on the second correlation, the traditional elements, the new elements, the full life cycle dimension, the development cycle dimension, and the type of risk target, a power system full-domain risk spectrum is constructed.

[0012] In one implementation, the step of inputting the target event into a risk identification model based on the power system's global risk spectrum for risk identification, and obtaining the risk identification result, includes: The target event is preprocessed; the preprocessing steps include data cleaning and normalization. Feature extraction is performed on the preprocessed target event to obtain the element association features, periodic stage features and loss morphology features corresponding to the target event; The element association characteristics, cycle stage characteristics, and loss pattern characteristics of the target event are matched with the risk spectrum of the entire power system to obtain the risk elements, cycle stages, and risk loss targets of the target event. Generate a three-dimensional risk event coordinate and a three-dimensional risk event location map based on the risk elements, cycle stages, and risk loss targets.

[0013] In one implementation, the event identification method based on the power system's global risk spectrum further includes: Obtain the coordinates of all target events and their corresponding 3D risk events in the target area; Based on the acquired target events and their corresponding three-dimensional risk event coordinates, a Sankey diagram of the global risk spectrum of the target location is obtained.

[0014] Secondly, the present invention provides an event identification system based on the global risk spectrum of a power system, comprising: The data acquisition module is used to acquire target events for the risks to be identified. The risk identification module is used to input the target event into a risk identification model based on the power system global risk spectrum for risk identification and to obtain the risk identification result. The result output module is used to output the risk identification result corresponding to the target event.

[0015] Thirdly, the present invention provides a terminal, comprising: a processor and a memory, wherein the memory stores an event recognition program based on a power system global risk spectrum, and the event recognition program based on the power system global risk spectrum, when executed by the processor, is used to implement the operation of the event recognition method based on the power system global risk spectrum as described in the first aspect.

[0016] Fourthly, the present invention also provides a computer-readable storage medium storing an event recognition program based on a power system global risk spectrum, wherein the event recognition program based on a power system global risk spectrum, when executed by a processor, is used to implement the operation of the event recognition method based on a power system global risk spectrum as described in the first aspect.

[0017] The present invention, by employing the above technical solution, has the following effects: This invention discloses an event identification method and system based on a power system-wide risk spectrum, comprising: acquiring a target event for risk identification; inputting the target event into a risk identification model based on the power system-wide risk spectrum for risk identification, obtaining a risk identification result; and outputting the risk identification result corresponding to the target event. The risk identification result is a three-dimensional risk event coordinate and a three-dimensional risk event location map based on risk elements, cycle stages, and risk loss targets. This invention automatically analyzes target events using a risk identification model based on the power system-wide risk spectrum, outputting specific risk identification results. This overcomes the problems of incomplete element coverage, missing cycle dimensions, and incompatibility with insurance semantics in existing risk event identification technologies, providing support for cross-dimensional risk assessment, source tracing analysis, and adaptation to insurance business in the power system. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the event identification method based on the global risk spectrum of the power system in this invention.

[0020] Figure 2 This is a flowchart of an event identification method based on the global risk spectrum of the power system in another implementation of the present invention.

[0021] Figure 3 This is a Sankey diagram of the global risk spectrum of the target region in one implementation of the present invention.

[0022] Figure 4 This is a three-dimensional risk event location map corresponding to the target event in one implementation of the present invention.

[0023] Figure 5 This is a three-dimensional risk event location map corresponding to target event two in one implementation of the present invention.

[0024] Figure 6 This is a schematic diagram of the event recognition system structure based on the global risk spectrum of the power system in one implementation of the present invention.

[0025] Figure 7 This is a functional schematic diagram of the terminal in one implementation of the present invention.

[0026] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0028] Exemplary methods Accurately identifying the types of risks in the power system, clarifying the boundaries of responsibility, and defining the forms of loss are fundamental to power system risk governance. However, existing financial and insurance businesses still have significant limitations in identifying and analyzing power system risk events: First, the risk identification dimension is too narrow. Existing studies mostly identify risks from four physical links: power source side, grid side, load side, and energy storage side. They fail to consider the impact of factors such as carbon constraints, digitalization, and market mechanisms in the construction of new power systems on system risks, resulting in the omission of a large number of non-physical risk events. Second, the risk analysis is static. Existing technologies usually conduct risk analysis for a specific stage of a certain process. However, risks are not static. They ignore the dynamic tracking of the evolution and transmission patterns of risk events over time and cannot predict the spread path of risks at different stages. Third, there is a disconnect between engineering and insurance terminology. Risk classification in the power industry revolves around engineering and technical dimensions such as personal safety, equipment safety, and power grid safety. The classification system lacks a mapping relationship with the insured objects in the financial and insurance fields, resulting in the identification of risk events failing to directly support the design of insurance products and the determination of claims liability.

[0029] In the existing technology, the identification and analysis of power system risk events by financial and insurance businesses have limitations, therefore, the existing technology needs to be improved.

[0030] To address the above technical problems, this invention provides an event identification method based on a power system-wide risk spectrum, comprising: inputting the target event into a risk identification model based on the power system-wide risk spectrum for risk identification, obtaining a risk identification result; and outputting the risk identification result corresponding to the target event. The risk identification result is a three-dimensional risk event coordinate and a three-dimensional risk event location map based on risk elements, cycle stages, and risk loss targets. This invention, through an automated analysis of target events using a risk identification model based on the power system-wide risk spectrum, outputs specific risk identification results. This overcomes the problems of incomplete element coverage, missing cycle dimensions, and incompatibility with insurance semantics in existing risk event identification technologies, providing support for cross-dimensional risk assessment, source tracing analysis, and adaptation to insurance business in the power system.

[0031] like Figure 1 As shown, this embodiment of the invention provides an event identification method based on the global risk spectrum of a power system, including the following steps: Step S100: Obtain the target event for the risk to be identified.

[0032] In this example, the target event for identifying the risk is defined as an event with clear spatiotemporal, equipment, or behavioral attributes that is planned, occurs, or detected throughout the entire process of power system operation, equipment maintenance, engineering work, or grid dispatching, and that has the potential to trigger any one of the following risks: property risk, personal injury risk, liability risk, or credit risk. For example, events such as charging pile installation plans, wind turbine blade transportation, and electricity price curve push notifications can all serve as target events for identifying risks.

[0033] like Figure 1 As shown, this embodiment of the invention provides an event identification method based on the global risk spectrum of a power system, including the following steps: Step S200: Input the target event into the risk identification model based on the power system global risk spectrum for risk identification, and obtain the risk identification result.

[0034] In this embodiment, before inputting the acquired target event into the risk identification model based on the power system global risk spectrum for risk identification, it is necessary to first obtain the risk identification model based on the power system global risk spectrum. The method for constructing the risk identification model based on the power system global risk spectrum includes the following steps: Step a, determine the large language model for risk identification.

[0035] In this embodiment, a large language model is determined for risk identification. For example, open-source large language models such as DeepSeek and GPT can be used as the preset large language model. Through the feature extraction capabilities of the large language model and natural language processing technology, the acquired events can be analyzed, the various features contained in the events can be extracted and analyzed, and risk identification can be completed.

[0036] Step b: Obtain a set of sample events, and train the large language model using the set of sample events to obtain a risk identification model; wherein, the set of sample events contains sample events, and the sample events are events for which risk identification results exist.

[0037] In this embodiment, a sample event set is obtained, which contains sample events. The sample events are events with risk identification results, that is, the sample events have completed risk identification, and the risk identification result of the sample events is any one of property risk, personal risk, liability risk or credit risk.

[0038] In this embodiment, the risk identification model is obtained by training the large language model with the sample event set. The sample events in the sample event set are input into the large language model. The large language model extracts and analyzes the features of the sample events to identify the risk target category corresponding to each sample event in the sample event set. The large language model is then optimized based on the risk identification results of the sample events to obtain the risk identification model.

[0039] Step c: Construct a global risk spectrum for the power system based on the sample event set.

[0040] In this embodiment, a power system-wide risk spectrum is constructed to overcome the problems of incomplete element coverage, missing cycle dimension, and incompatibility of insurance semantics in existing event identification methods. Therefore, the constructed power system-wide risk spectrum includes at least: a full-element dimension risk spectrum, a full-cycle dimension risk spectrum, and a full-target dimension risk spectrum.

[0041] In this embodiment, the main content of the all-element dimension risk spectrum is all-element dimension risk. All-element dimension risk refers to a risk analysis perspective that systematically identifies the sources, carriers, and transmission mechanisms of risk, starting from the underlying components of the new power system. Unlike traditional power systems where risk identification mainly revolves around single equipment, single links, and localized scenarios, the risks of the new power system are deeply embedded in multiple elements such as power generation, transmission and distribution, consumption, energy storage, carbon constraints, digital empowerment, price transmission, computing power support, technological innovation, and market allocation, exhibiting significant cross-element coupling characteristics. Essentially, the all-element dimension reveals the structural foundation of risk: "where it originates, what carrier it relies on, and through what mechanism it is triggered and spreads."

[0042] In this embodiment, the main content of the full-cycle risk spectrum is full-cycle risk. Full-cycle risk refers to a risk analysis perspective that characterizes the generation, accumulation, amplification, and externalization of risks by starting from the dynamic process of the evolution, functional migration, and state transformation of various elements in the new power system over time. Risks in the new power system are not static but evolve continuously with different elements in a periodic manner. Their manifestations, scope of impact, and modes of damage are constantly reconstructed with the changes in stages. Therefore, the full-cycle dimension focuses not only on the risk state at a certain point in time, but also on the temporal logic of the continuous evolution of risks from innate, primary, and cumulative to derivative in different stages.

[0043] In this embodiment, the main content of the full-scope risk spectrum is full-scope risk. Full-scope risk is an analytical perspective that starts from the loss outcome and categorizes the risks of new power systems into specific harmful objects. System engineering technology risks are generally classified only according to the cause of loss, such as equipment failure risk, natural disaster risk, and human operation risk, which cannot be directly matched with the type of insured object. Regardless of the origin of the risk or the stage at which it occurs, it will ultimately be transmitted and realized as a specific loss consequence, manifested as damage to asset safety, personnel safety, liability, and the reputation of the entity, and can be matched with the target of the insured object. The full-scope risk dimension reveals the result-oriented logic of "where the risk ultimately acts, what form of loss it forms, and what type of insurance object it corresponds to," thereby realizing the effective transformation of system engineering risks into financial insurance objects.

[0044] In one implementation of this embodiment, the method for constructing a global risk spectrum of the power system based on the sample event set includes the following steps: Step c1 involves preprocessing the sample events in the sample event set; the preprocessing steps include data cleaning and normalization.

[0045] In this embodiment, the sample events in the sample event set are preprocessed. The preprocessing steps include data cleaning and normalization. The sample events in the sample event set are usually text descriptions. The directly obtained text descriptions are long and may contain other content unrelated to risk identification. Data cleaning of the sample events involves deleting erroneous, invalid, and irrelevant descriptions from the sample event descriptions. Normalization involves converting the event descriptions into a preset text format, thereby accurately extracting the features of the sample events through the relevant descriptions.

[0046] Step c2 involves extracting features from the preprocessed sample events to obtain the element association features, periodic stage features, and loss morphology features corresponding to each sample event.

[0047] In this embodiment, feature extraction is performed on the preprocessed sample events to obtain the element association features, periodic stage features and loss morphology features corresponding to each sample event.

[0048] Among them, the element association features are obtained by extracting keywords such as entities and rules involved in the event, such as "charging pile", "electricity price prediction" and "carbon quota", thereby obtaining a candidate set of risk elements associated with the event; Cycle phase characteristics are obtained by extracting time-series features such as event timestamps, the operational duration of associated assets, and the effective time of rule versions, thereby obtaining a candidate set of cycle phases in which the event is located; Loss pattern characteristics are obtained by extracting descriptions of the consequences of the event, such as "equipment damage", "personal injury or death", "third-party claims", "contract breach", etc., thereby obtaining a candidate set of insurable objects corresponding to the event.

[0049] Step c3: Based on the sample events, determine the correlation between the element association features, the periodic stage features, and the loss morphology features.

[0050] In this embodiment, determining the correlation between the element association features, the periodic stage features, and the loss morphology features based on the sample events includes the following steps: Obtain the risk element candidate set corresponding to the element association feature, the cycle stage candidate set corresponding to the cycle stage feature, and the risk target candidate set corresponding to the loss pattern feature; Based on the sample events, determine the first association between the candidate set of risk factors and the candidate set of cycle stages; Based on the sample events, a second association relationship is determined between the candidate set of risk elements, the candidate set of cycle stages, and the candidate set of risk targets.

[0051] In this embodiment, a candidate set of risk elements corresponding to the associated features of the elements is obtained, wherein the risk elements include at least: source elements, network elements, load elements, storage elements, carbon elements, data elements, price elements, computing elements, innovation elements, and market elements. The acquired candidate set of cycle stages includes at least the following cycle stages: planning stage, construction stage, operation stage, decommissioning stage, start-up stage, development stage, maturity stage, and replacement stage. The acquired candidate set of risk targets includes at least: property risk, personal risk, liability risk, and credit risk.

[0052] In this embodiment, based on the sample events, the first association relationship between the candidate set of risk elements and the candidate set of cycle stages is determined. Specifically, for the same sample event, the risk element corresponding to the sample event is associated with the cycle stage, further obtaining the first association relationship between the risk elements of all sample events in the sample event set and the cycle stages. For example, for the risk element of a source element, which belongs to a physical entity element, the corresponding cycle stage can be any one of the four stages: planning stage, construction stage, operation stage, and decommissioning stage.

[0053] In this embodiment, based on the sample events, a second association relationship is determined between the candidate set of risk elements, the candidate set of cycle stages, and the candidate set of risk targets. Specifically, for the same sample event, the risk element, cycle stage, and risk target corresponding to that sample event are associated to further obtain the second association relationship between the risk element, cycle stage, and risk target for all sample events in the sample event set. For example, for the risk element of a network, property risk needs to be considered during the planning stage, personal risk during the construction stage, or liability risk at any stage.

[0054] Step c4: Construct a global risk spectrum of the power system based on the aforementioned relationships.

[0055] In this embodiment, constructing a power system-wide risk spectrum based on the aforementioned correlation includes the following steps: Step c41: Divide the risk elements in the candidate set of risk elements into traditional elements and new elements.

[0056] In this embodiment, the risk elements in the candidate set of risk elements are divided into traditional elements and new elements. Among them, the four physical elements of "power source, grid, load, and energy storage" are physical entities that directly participate in the conversion and transmission of electricity in the new power system. They are the carriers that mainly rely on maintaining system balance. The risk forms are mainly related to the physical links such as power production, transmission, and consumption. Therefore, the traditional elements are obtained as source elements, grid elements, load elements, and storage elements, as follows: Source elements, on the power supply side, include not only traditional thermal power, hydropower, and nuclear power, but also renewable energy sources such as wind power, photovoltaic power, and biomass energy; Network elements, on the power grid side, are the networks for the transmission, distribution, and dispatch of electrical energy, including ultra-high voltage power grids, regional transmission networks, medium and low voltage distribution networks, and intelligent dispatch systems at all levels; Load elements, on the load side, include all electricity consumers, including industrial load, commercial and public building load, residential load, and emerging loads; Energy storage, or energy storage side, is a technology system that stores energy through a medium or device and releases it when needed. It includes forms such as pumped hydro storage, electrochemical energy storage, hydrogen energy storage, and flywheel energy storage.

[0057] In this embodiment, with the continuous advancement of the "dual carbon" target and the rapid development of the digital economy, the new power system has evolved from a relatively simple physical system into a complex mega-system encompassing multiple dimensions such as physical, environmental, economic, and digital. Information, rules, and technological elements, embodied in carbon indicators, data, electricity prices, computing power, technological innovation, and the electricity market, do not directly participate in energy conversion, but profoundly influence the system's operation through the regulation, constraint, pricing, and empowerment of traditional physical elements. Unlike the physical risks of traditional elements, the risks of new elements mainly manifest as information risks and mechanism risks. They do not directly burn out a piece of equipment or break a wire, but rather indirectly transmit to the physical system through misleading decision-making information and altering constraint boundaries, ultimately causing equipment damage, operational instability, or economic losses. Therefore, the risk elements, based on traditional physical elements, are expanded to include six new elements: carbon elements, data elements, price elements, computing elements, innovation elements, and market elements, as detailed below: Carbon elements refer to the carbon quantification indicators, management mechanisms, and related assets in new power systems to achieve the goals of "carbon peaking and carbon neutrality," including carbon emission trajectories, carbon quotas, and carbon assets. Data elements refer to power data, which is a digital mapping of the operating status of a new type of power system and a carrier for the exchange of status information, including operating data, environmental data, and user data; Price factors, referring to electricity prices, are core price signals and value adjustment levers that reflect the supply, demand and cost of electricity, including government-approved electricity prices and market-clearing electricity prices; Computing elements refer to computing power, which is the computing capability that supports decision optimization, including computing resources, algorithm models, AI decision-making systems, etc. Innovation elements refer to technological innovation, which encompasses comprehensive breakthroughs in innovation from fundamental key materials and high-end equipment manufacturing to complex system control algorithms, continuously providing inexhaustible power for new power systems to cope with unknown risks and achieve iterative upgrades; The market element refers to the electricity market, which is the decisive mechanism for achieving large-scale optimal allocation of resources, including the spot market, ancillary services market, and capacity market.

[0058] In this embodiment, the risk elements in the candidate set of risk elements are divided into traditional elements and new elements. The risks of the new power system are hierarchically classified and systematically characterized through ten core elements. For any target event, its risk element can be obtained through feature extraction and analysis.

[0059] Step c41: Based on the first association relationship, classify the cycle stages of the candidate set of cycle stages to obtain the full life cycle dimension corresponding to the traditional elements and the development cycle dimension corresponding to the new elements.

[0060] In this embodiment, based on the first association relationship, the cycle stages of the candidate set of cycle stages are classified to obtain the full life cycle dimensions of the traditional elements "source elements, network elements, load elements, and storage elements" as "planning stage, construction stage, operation stage, and decommissioning stage", and the development cycle dimensions of the new elements "carbon elements, number elements, price elements, computing elements, innovation elements, and market elements" as "start-up stage, development stage, maturity stage, and replacement stage".

[0061] The full life-cycle dimension is a risk analysis from the perspective of the temporal evolution of traditional physical entities—"source elements, grid elements, load elements, and storage elements"—in new power systems. From design blueprints to implementation and operation, and ultimately to their eventual demise, the risk characteristics of physical assets dynamically evolve as the project progresses. This dimension combines the linear management model of traditional infrastructure with the complex operational constraints of new power systems, systematically analyzing various risks existing in the four stages of planning, construction, operation, and decommissioning of traditional elements. These risks include those that exist from the outset, are system-rooted, have accumulated over a long period, and are caused by other factors.

[0062] From a development cycle perspective, the focus is on six emerging factors: carbon, data, price, computing, innovation, and market. Risks are identified by tracing their evolution from initiation to development, maturity, and replacement. Unlike physical equipment, which typically follows a "construction-operation-retirement" lifecycle, emerging factors are characterized by the continuous evolution of systems, platforms, rules, algorithms, computing power infrastructure, and new business models. Their risks do not necessarily arise with the formation of the equipment, nor do they necessarily end with its retirement. Instead, they evolve with application penetration, rule improvement, market expansion, and technological iteration. Typical risks in this dimension include four sub-dimensions: initiation, development, maturity, and replacement.

[0063] In this embodiment, risks in the new power system evolve continuously throughout the entire lifecycle of traditional elements, from planning, construction, operation to decommissioning, and throughout the entire development cycle of new elements, from initiation, development, maturity to replacement. Therefore, the full-cycle risk spectrum identifies risks within the framework of the entire process of evolution, in order to reveal the dynamic patterns of their emergence, latency, diffusion, and outbreak over time.

[0064] Step c42: Obtain the types of risk targets in the candidate set of risk targets.

[0065] In this embodiment, starting from the loss outcome, the risks of the new power system are categorized into specific harmful objects, thus obtaining the following types of risk objects in the candidate set of risk objects: property risk, personal risk, liability risk, and credit risk, as detailed below: Property risk, in the context of new power systems, refers to the risk that various tangible physical facilities, intangible digital assets, and expected economic benefits within the system will be damaged, lost, or depreciated due to natural disasters, accidents, or systemic failures. In new power systems, personal risk refers to the risk of life injury, health damage, or occupational exposure to workers, maintenance personnel, and the general public caused by the interaction of human factors, material factors, environmental factors, and management factors during the planning, construction, operation, maintenance, emergency repair, and emergency response of the system. In new power systems, liability risk refers to the economic compensation and joint legal liability that participating entities should bear in accordance with the law when, during the planning, construction, operation, or decommissioning of the system, due to physical equipment failure, intangible rule system failure, or management oversight, third parties (such as the general public, downstream enterprises, and the ecological environment) suffer property losses or personal injuries. In new power systems, credit risk refers to the risk that participating entities may fail to fulfill their obligations to make economic payments, deliver rights, or respond to regulatory instructions during power trading, environmental rights transfer, ancillary service provision, and project financing due to deteriorating financial conditions, physical operational deviations, or subjective intent to default.

[0066] Step c43: Based on the second correlation, the traditional elements, the new elements, the full life cycle dimension, the development cycle dimension, and the type of risk target, construct a power system global risk spectrum.

[0067] In this embodiment, a power system-wide risk spectrum is constructed based on the second correlation, the traditional elements, the new elements, the full life cycle dimension, the development cycle dimension, and the type of risk target.

[0068] In this embodiment, the power system global risk spectrum includes: A comprehensive risk spectrum encompassing all elements: source elements, network elements, load elements, storage elements, carbon elements, data elements, price elements, computing elements, innovation elements, and market elements. A full-cycle risk spectrum, encompassing the planning stage, construction stage, operation stage, decommissioning stage, initial stage, development stage, maturity stage, and transition stage; The entire risk spectrum includes property risk, personal risk, liability risk, and credit risk.

[0069] In one implementation of this embodiment, based on the classification of risk factors, the power system-wide risk spectrum is divided into a traditional factor power system-wide risk spectrum and a new factor power system-wide risk spectrum, as detailed below: The traditional elements of the power system encompass a full-domain risk spectrum, including four categories of traditional elements: source elements, grid elements, load elements, and storage elements; the entire life cycle dimension, including the planning stage, construction stage, operation stage, and decommissioning stage; and a full-target dimension risk spectrum, including property risk, personal risk, liability risk, and credit risk.

[0070] The new element power system's comprehensive risk spectrum includes six new elements: carbon, data, price, computing, innovation, and market; a development cycle dimension including the initial stage, development stage, maturity stage, and transition stage; and a full-target dimension risk spectrum including property risk, personal risk, liability risk, and credit risk.

[0071] Step d: Optimize the risk identification model based on the constructed power system global risk spectrum to obtain a risk identification model based on the power system global risk spectrum.

[0072] In this embodiment, the risk identification model is optimized based on the constructed power system global risk spectrum. Specifically, the output of the risk identification model is transformed into a risk identification result based on the power system global risk spectrum. The risk identification result needs to include the risk elements, cycle stages, and risk targets corresponding to the input events.

[0073] It should be noted that the risk identification results output by the risk identification model based on the power system's global risk spectrum must correspond to the rules of the power system's global risk spectrum. For example, when the risk element in the risk identification result of an event is a traditional element, its corresponding development stage can only be any stage in the full life cycle dimension.

[0074] In one implementation of this embodiment, the risk identification model based on the power system global risk spectrum constructed based on the above steps includes the following steps in step S200: Step S201: Preprocess the target event; the preprocessing steps include data cleaning and normalization.

[0075] In this embodiment, the target event is preprocessed. The preprocessing steps include data cleaning and normalization. Data cleaning of the target event involves deleting erroneous, invalid, and irrelevant descriptions in the description of the target event that are not related to the risk spectrum of the entire power system. Normalization involves converting the event description into a preset text format, thereby accurately extracting the features of the target event through relevant descriptions.

[0076] Step S202: Extract features from the preprocessed target event to obtain the element association features, periodic stage features and loss morphology features corresponding to the target event.

[0077] In this embodiment, feature extraction is performed on the preprocessed target event to obtain the element association features, periodic stage features and loss morphology features corresponding to the target event.

[0078] Among them, the element association characteristics are obtained by extracting keywords such as entities and rules involved in the event, such as "charging pile", "electricity price prediction" and "carbon quota". Cycle phase characteristics are obtained by extracting time-series features such as event timestamps, the operational duration of associated assets, and the effective time of rule versions; Loss pattern characteristics are obtained by extracting descriptions of the consequences of the event, such as "equipment damage", "personal injury or death", "third-party claims", and "contract breach".

[0079] Step S203: The element association characteristics, cycle stage characteristics and loss pattern characteristics of the target event are matched with the risk spectrum of the entire power system to obtain the risk elements, cycle stages and risk loss targets of the target event.

[0080] In this embodiment, the element association features, periodic stage features, and loss pattern features of the target event are matched with the power system global risk spectrum to obtain the specific risk elements, periodic stages, and risk loss targets corresponding to the target event in the power system global risk spectrum.

[0081] In this embodiment, the extracted event features are matched with the constructed power system global risk spectrum. The specific matching rules are as follows: For risk factor matching, if the event involves physical equipment failure or abnormal energy transmission, the traditional elements of "source, grid, load, and storage" are prioritized; if the event involves policy compliance, data processing, price signals, algorithmic decision-making, technological innovation, or market transactions, the new elements of "carbon, data, price, computing, innovation, and market" are matched; if there is multi-element coupling, the direct causal element of the event is used as the main dimension, and the rest are labeled as related elements. Cycle phase matching: If it has been matched as a traditional element, it will be matched with four phases: planning, construction, operation, and decommissioning, based on the project progress and commissioning status of the related assets; if it has been matched as a new element, it will be matched with four phases: start-up, development, maturity, and iteration, based on the popularization rate of the rules, the maturity of the technology, and the transaction scale of the market. Risk targets are matched as follows: if the consequence of the event is damage to equipment or assets, it is matched as "property risk"; if the consequence of the event is personal injury or occupational disease, it is matched as "personal risk"; if the event triggers third-party claims or environmental penalties, it is matched as "liability risk"; if the event is a transaction default or non-performance, it is matched as "credit risk".

[0082] It should be noted that the features listed in the above matching rules are only illustrative and used to explain the principle of the rules. In actual application, the range of features that can be used for matching is not limited to these. All feature extraction and matching logic that conforms to the core idea of ​​this application is within the scope of protection.

[0083] Step S204: Generate three-dimensional risk event coordinates and a three-dimensional risk event location map based on the risk elements, cycle stages, and risk loss targets.

[0084] In this embodiment, the risk identification model based on the power system global risk spectrum uses risk elements as the Y-axis, cycle stages as the X-axis, and risk targets as the Z-axis. Through spatial scatter projection, the new power system safety risks are anchored to precise three-dimensional coordinates, generating three-dimensional risk event coordinates and three-dimensional risk event location maps based on the risk elements, cycle stages, and risk loss targets, thus achieving intuitive and accurate positioning of risk scenarios.

[0085] Specifically, based on the matching result of step S203, a unique three-dimensional risk event coordinate R(x,y,z) is output. When the risk element of the target event is a traditional element, its three-dimensional risk coordinate mapping relationship is shown in Table 1 below; when the risk element of the target event is a new element, its three-dimensional risk coordinate mapping relationship is shown in Table 2 below.

[0086] Table 1. Three-dimensional risk coordinate mapping table for traditional elements.

[0087]

[0088] Table 2. Three-dimensional risk coordinate mapping table for new elements.

[0089]

[0090] Furthermore, for any target event input into the risk identification model based on the global risk spectrum of the power system, a corresponding three-dimensional risk event location map can be generated based on its three-dimensional risk event coordinates R(x,y,z).

[0091] like Figure 1 As shown, this embodiment of the invention provides an event identification method based on the global risk spectrum of a power system, including the following steps: Step S300: Output the risk identification result corresponding to the target event.

[0092] In this embodiment, the risk identification result corresponding to the target event is output, that is, the three-dimensional risk event coordinates and three-dimensional risk event location map of the target event based on the power system global risk spectrum are output.

[0093] like Figure 2 As shown, another embodiment of the present invention also provides an event identification method based on the global risk spectrum of a power system. Building upon existing event identification methods based on the global risk spectrum of a power system, this method introduces a Sankey diagram of the global risk spectrum to achieve risk identification and analysis of all target events within a target region. Specifically, it includes the following steps: Step S100: Obtain the target event for the risk to be identified; Step S200: Input the target event into the risk identification model based on the power system global risk spectrum for risk identification, and obtain the risk identification result; Step S300: Output the risk identification result corresponding to the target event; Step S400: Obtain all target events and their corresponding three-dimensional risk event coordinates in the target area. Based on the obtained target events and their corresponding three-dimensional risk event coordinates, obtain the Sankey map of the global risk spectrum of the target area.

[0094] In this embodiment, when a comprehensive power system risk assessment of the entire target region is required, all target events, including data on new power system-related accidents and cases, are obtained from data sources such as monitoring systems and business platforms in the target region. These target events are then input into a risk identification model based on a comprehensive power system risk spectrum. The model analyzes the target events to obtain their corresponding three-dimensional risk event coordinates, such as R1(x1,y1,z1), R2(x2,y2,z2), R3(x3,y3,z3), ..., Rn(xn,yn,zn). This allows for the creation of a comprehensive risk spectrum Sankey diagram, using a "risk element—cycle stage—target type" framework as the main line and employing differentiated color coding to distinguish the risk level or frequency density of different coordinate combinations.

[0095] like Figure 3 The diagram shown illustrates a Sankey diagram of the overall risk spectrum of the target region in one implementation of this embodiment. This Sankey diagram visually presents how numerous potential risks evolve gradually from risk-causing factors at different stages of development, ultimately concentrating on four main risk targets throughout the entire process. It breaks down the silos of traditional risk data, making risk correlations clear and traceable. This not only supports the prediction of high-frequency risk scenarios but also provides data-driven decision-making basis for adjusting risk management mechanisms and optimizing insurance service strategies.

[0096] Furthermore, in this embodiment of the application, the above-mentioned event identification method based on the global risk spectrum of the power system is described based on a specific application scenario.

[0097] In some application scenarios, the target event is described as follows: A V2G charging pile deployed in the underground parking garage of a commercial complex experiences thermal runaway and explosion during the charging of electric vehicles. The fire spreads to surrounding third-party vehicles and building structures, resulting in joint liability for third-party property compensation and personal injury.

[0098] The aforementioned target event is input into a risk identification model based on the power system's global risk spectrum for risk identification. The model analyzes the target event, recognizing that the V2G charging pile belongs to the load element within traditional elements, and since the event occurred during the operational phase and caused losses to a third party, it falls under the category of liability risk. Therefore, the output of the three-dimensional risk event coordinates based on the power system's global risk spectrum is R1 (operation, load, liability risk), and the output is as follows: Figure 4 The three-dimensional risk event location map corresponding to the target event shown.

[0099] In other application scenarios, the second target event is described as follows: A virtual power plant platform pushes the next day's electricity price curve to its contracted users, suggesting that users reduce electricity consumption at a certain time (marked as "high-price period"). In reality, the system is in a low-price period at that time, and the users stop production due to misjudgment, resulting in business losses. It was later found that the platform's electricity price prediction was incorrect, and the virtual power plant platform bears responsibility for the user's economic losses due to misleading advice.

[0100] The aforementioned target event is input into a risk identification model based on the power system's global risk spectrum for risk identification. The model analyzes the target event, classifying it as a price element risk of a new type of factor, occurring at a relatively mature stage of the price element, and falling under the liability risk of the virtual power plant platform. Therefore, the output three-dimensional risk event coordinates based on the power system's global risk spectrum are R2 (maturity, price, liability risk), and the output is as follows: Figure 5 The three-dimensional risk event location map corresponding to target event two is shown.

[0101] This embodiment achieves the following technical effects through the above technical solution: With comprehensive coverage, this embodiment innovatively expands the risk identification elements of the new power system from the four traditional elements of "source, grid, load and storage" to ten core elements including new elements of "carbon, data, price, calculation, innovation and market". It effectively incorporates consideration of new risk factors such as carbon constraints, digitalization and market mechanisms, enhances the ability to identify new risk factors, and makes up for the blind spots of the traditional element perspective. With strong dynamic tracking capabilities, this embodiment extends the risk analysis perspective from static assessment to the entire life cycle of traditional elements from planning, construction, operation to decommissioning, and the entire development cycle of new elements from initiation, development, maturity to replacement. It solves the problem of unclear understanding of the evolution and transmission patterns of risks at each stage caused by a static, single-stage risk perspective, and provides a basis for phased prediction and prevention for risk management. The insurance industry has high adaptability. This embodiment maps the engineering and technical risks of new power systems to four core targets in the financial and insurance field: property risk, personal risk, liability risk and credit risk. This solves the problem of insufficient insurance coverage and limited adaptability caused by the mismatch between the language of insurance protection risks and engineering and technical risks.

[0102] It is highly interpretable and scalable, with clear 3D coordinate positioning logic, which facilitates consensus within and outside the industry. It also supports subsequent model iteration by adding new elements, refining cycles, and expanding targets, adapting to new risk scenarios brought about by the continuous development of new power systems.

[0103] Exemplary device Based on the above embodiments, the present invention also provides an event identification system based on the global risk spectrum of a power system, such as... Figure 6 As shown, the event identification system based on the power system's global risk spectrum includes: Data acquisition module 61 is used to acquire target events for risks to be identified; Risk identification module 62 is used to input the target event into a risk identification model based on the power system global risk spectrum for risk identification and obtain risk identification results; The result output module 63 is used to output the risk identification result corresponding to the target event.

[0104] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 7 As shown.

[0105] The terminal includes: a processor, a memory, an interface, a display screen, and a communication module connected via a system bus; wherein, the processor of the terminal provides computing and control capabilities; the memory of the terminal includes a computer-readable storage medium and internal memory; the computer-readable storage medium stores an operating system and computer programs; the internal memory provides an environment for the operation of the operating system and computer programs in the computer-readable storage medium; the interface is used to connect to external devices; the display screen is used to display relevant information; and the communication module is used to communicate with a cloud server or other devices.

[0106] When executed by a processor, this computer program is used to implement an event identification method based on the global risk spectrum of the power system.

[0107] It will be understood by those skilled in the art that Figure 7 The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0108] In one embodiment, a terminal is provided, comprising: a processor and a memory, the memory storing an event identification program based on a power system global risk spectrum, the event identification program based on a power system global risk spectrum being executed by the processor to implement the operation of the event identification method based on a power system global risk spectrum as described above.

[0109] In one embodiment, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores an event recognition program based on a power system global risk spectrum, which, when executed by a processor, is used to implement the operation of the event recognition method based on the power system global risk spectrum as described above.

[0110] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, database, or other media used in the embodiments provided by this invention can include both non-volatile and volatile memory.

[0111] In summary, this invention provides an event identification method and system based on a power system-wide risk spectrum, comprising: inputting the target event into a risk identification model based on the power system-wide risk spectrum for risk identification, obtaining a risk identification result; and outputting the risk identification result corresponding to the target event. The risk identification result is a three-dimensional risk event coordinate and a three-dimensional risk event location map based on risk elements, cycle stages, and risk loss targets. This invention, through an automated analysis of target events using a risk identification model based on the power system-wide risk spectrum, outputs specific risk identification results, overcoming the problems of incomplete element coverage, missing cycle dimensions, and incompatibility with insurance semantics in existing risk event identification technologies. It provides support for cross-dimensional risk assessment, source tracing analysis, and adaptation to insurance business in the power system.

[0112] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. An event identification method based on the global risk spectrum of a power system, characterized in that, include: Acquire the target events for the risks to be identified; The target event is input into a risk identification model based on the power system global risk spectrum for risk identification, and the risk identification result is obtained. Output the risk identification results corresponding to the target event.

2. The event identification method based on the global risk spectrum of the power system according to claim 1, characterized in that, The method for constructing the risk identification model based on the global risk spectrum of the power system includes the following steps: Determine a large language model for risk identification; Obtain a set of sample events, and train the large language model using the set of sample events to obtain a risk identification model; wherein, the set of sample events contains sample events, and the sample events are events for which risk identification results exist; Based on the sample event set, a global risk spectrum of the power system is constructed; The risk identification model is optimized based on the constructed power system global risk spectrum to obtain a risk identification model based on the power system global risk spectrum.

3. The event identification method based on the global risk spectrum of the power system according to claim 2, characterized in that, The step of constructing a global risk spectrum of the power system based on the sample event set includes the following steps: The sample events in the sample event set are preprocessed; the preprocessing steps include data cleaning and normalization. Feature extraction is performed on the preprocessed sample events to obtain the element association features, cycle stage features and loss morphology features corresponding to each sample event; Based on the sample events, determine the correlation between the element association features, the periodic stage features, and the loss morphology features; A global risk spectrum for the power system is constructed based on the aforementioned relationships.

4. The event identification method based on the global risk spectrum of the power system according to claim 3, characterized in that, Based on the sample events, the correlation relationships between the element association features, the periodic stage features, and the loss morphology features are determined, including: Obtain the risk element candidate set corresponding to the element association feature, the cycle stage candidate set corresponding to the cycle stage feature, and the risk target candidate set corresponding to the loss pattern feature; Based on the sample events, determine the first association between the candidate set of risk factors and the candidate set of cycle stages; Based on the sample events, a second association relationship is determined between the candidate set of risk elements, the candidate set of cycle stages, and the candidate set of risk targets.

5. The event identification method based on the global risk spectrum of the power system according to claim 4, characterized in that, The construction of a power system-wide risk spectrum based on the aforementioned correlation includes: The risk elements in the candidate set of risk elements are divided into traditional elements and new elements; Based on the first association relationship, the cycle stages of the candidate set of cycle stages are classified to obtain the full life cycle dimension corresponding to the traditional elements and the development cycle dimension corresponding to the new elements. Obtain the types of risk targets from the candidate set of risk targets; Based on the second correlation, the traditional elements, the new elements, the full life cycle dimension, the development cycle dimension, and the type of risk target, a power system full-domain risk spectrum is constructed.

6. The event identification method based on the global risk spectrum of the power system according to claim 1, characterized in that, The step of inputting the target event into a risk identification model based on the power system's global risk spectrum for risk identification, and obtaining the risk identification result, includes: The target event is preprocessed; the preprocessing steps include data cleaning and normalization. Feature extraction is performed on the preprocessed target event to obtain the element association features, periodic stage features and loss morphology features corresponding to the target event; The element association characteristics, cycle stage characteristics, and loss pattern characteristics of the target event are matched with the risk spectrum of the entire power system to obtain the risk elements, cycle stages, and risk loss targets of the target event. Generate a three-dimensional risk event coordinate and a three-dimensional risk event location map based on the risk elements, cycle stages, and risk loss targets.

7. The event identification method based on the global risk spectrum of the power system according to claim 6, characterized in that, The event identification method based on the power system global risk spectrum also includes: Obtain the coordinates of all target events and their corresponding 3D risk events in the target area; Based on the acquired target events and their corresponding three-dimensional risk event coordinates, a Sankey diagram of the global risk spectrum of the target location is obtained.

8. An event identification system based on a power system global risk spectrum, used to implement the event identification method based on a power system global risk spectrum as described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to acquire target events for the risks to be identified. The risk identification module is used to input the target event into a risk identification model based on the power system global risk spectrum for risk identification and to obtain the risk identification result. The result output module is used to output the risk identification result corresponding to the target event.

9. A terminal, characterized in that, include: The processor and memory, wherein the memory stores an event recognition program based on the power system global risk spectrum, and the event recognition program based on the power system global risk spectrum, when executed by the processor, is used to implement the operation of the event recognition method based on the power system global risk spectrum as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an event recognition program based on a power system global risk spectrum, which, when executed by a processor, is used to implement the operation of the event recognition method based on a power system global risk spectrum as described in any one of claims 1-7.

Images