Risk management method and system for intelligent decision support system for normal / abnormal operation of nuclear power plants

The risk management system for nuclear power plant decision support systems addresses risks from open-source software and adversarial learning by testing and switching to a safe mode, ensuring safety and security.

KR103046826B1Active Publication Date: 2026-11-11KOREA ATOMIC ENERGY RES INST
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
KR1020240137601
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2024-10-10
Publication Date
2026-11-11
Estimated Expiration
2044-10-10

AI Technical Summary

Technical Problem

Existing intelligent decision support systems for nuclear power plants face risks from open-source software and adversarial machine learning, which can lead to critical errors and safety threats.

Method used

A risk management system that includes an input unit, control unit, and output unit to test the decision support system, generate test data, compare with correct answers, and switch functions to a safe mode when risks are detected, using methods like logic checks and performance tests.

Benefits of technology

The system effectively manages risks by ensuring the decision support system enters a safe mode when threats occur, mitigating license and cyber security risks from open-source software and adversarial machine learning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112024109971951-PAT00004_ABST
    Figure 112024109971951-PAT00004_ABST
Patent Text Reader

Abstract

The present invention provides a risk management method comprising receiving user input for performing a test of a decision support system through an input module provided in an operator indicator of a nuclear power plant, performing a test of the decision support system based on the user input, generating test data for the decision support system based on the result of the test, comparing the test data with predetermined correct answer data for the decision support system, switching a function corresponding to the input module based on the comparison result, and controlling the decision support system to a safe mode based on user input for the switched input module.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present invention relates to a risk management method and system for an intelligent decision support system for normal / abnormal operation of a nuclear power plant. Background Technology

[0002] Intelligent Decision Support Systems (IDSS) for nuclear power plant operators are designed to identify essential safety information regarding the plant's operating status and in the event of an accident, and to provide operators with efficient response guidelines; research and development are currently underway worldwide.

[0003] In particular, with the recent advancements in computer hardware and artificial intelligence technology, the feasibility and effectiveness of AI-based decision support systems are increasing. Furthermore, from the perspective of nuclear safety, there is a growing demand for effective decision support systems to reduce human error, which is a major cause of accidents during plant operations.

[0004] However, in the case of artificial intelligence technology, risks arising from the use of open-source software and adversarial machine learning exist, potentially leading to critical errors in the system; this is cited as a major factor that must be considered when developing and operating decision support systems in nuclear power plant operations where safety is critical.

[0005] Therefore, risk management measures are required to ensure the safety of nuclear power plants during the development and operation of decision support systems. The problem to be solved

[0006] The present invention relates to a risk management method and system for an intelligent decision support system for normal / abnormal operation of a nuclear power plant, capable of preparing for license and cyber security threats resulting from the use of open source software for the nuclear power plant's decision support system and adversarial machine learning threats resulting from the use of artificial intelligence technology. means of solving the problem

[0007] To solve the problem described above, the risk management method according to the present invention may include, in a risk management method using a risk management system, the steps of: receiving user input to perform a test of an Intelligent Decision Support System (IDSS) through an input module provided in an operator indicator of a nuclear power plant; performing a test of the Intelligent Decision Support System based on the user input and generating test data for the Intelligent Decision Support System based on the result of the test; comparing the test data with predetermined correct answer data for the Intelligent Decision Support System and switching a function corresponding to the input module based on the comparison result; and controlling the Intelligent Decision Support System to a safe mode based on user input for the switched input module.

[0008] In addition, the risk management system according to the present invention includes: an input unit that receives user input for performing a test of a decision support system through an input module provided in an operator indicator of a nuclear power plant; and a control unit that performs a test of the decision support system based on the user input. The control unit generates test data for the decision support system based on the result of performing the test, compares the test data with predetermined correct answer data for the decision support system, switches a function corresponding to the input module based on the comparison result, and controls the decision support system to a safe mode based on user input for the switched input module.

[0009] In addition, a program stored on a computer-readable recording medium according to the present invention is executed by one or more processes in an electronic device and is a program stored on a computer-readable recording medium, wherein the program may include instructions for performing: receiving user input to perform a test of an Intelligent Decision Support System (IDSS) through an input module provided in an operator indicator of a nuclear power plant; performing a test of the Intelligent Decision Support System based on the user input and generating test data for the Intelligent Decision Support System based on the result of the test; comparing the test data with predetermined correct answer data for the Intelligent Decision Support System and switching a function corresponding to the input module based on the comparison result; and controlling the Intelligent Decision Support System to a safe mode based on user input for the switched input module. Effects of the invention

[0010] According to various embodiments of the present invention, a risk management method and system for an intelligent decision support system for normal / abnormal operation of a nuclear power plant manage risks to the decision support system based on various risk management measures and are implemented such that the decision support system enters a safe mode when a risk occurs, thereby enabling preparation against license and cyber security threats resulting from the use of open source software for the nuclear power plant's decision support system and adversarial machine learning threats resulting from the use of artificial intelligence technology. Brief explanation of the drawing

[0011] FIG. 1 illustrates an embodiment of a system risk management framework of a nuclear power plant decision support system that includes operation by a risk management system according to the present invention. FIG. 2 illustrates an example of the system risk management planning method of FIG. 1. FIG. 3 illustrates an example of a method for administrative control and response to system risk of FIG. 1. FIG. 4 illustrates a risk management system according to the present invention. FIG. 5 illustrates an embodiment in which a function corresponding to the input module of the driver indicator is switched. FIG. 6 is a flowchart illustrating a risk management method according to the present invention. FIG. 7 is a flowchart illustrating one example of the step of generating test data. FIG. 8 is a flowchart illustrating one embodiment of a step for switching functions. FIG. 9 is a flowchart illustrating an example of a step for controlling in a safe mode. Specific details for implementing the invention

[0012] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are assigned the same reference number regardless of the drawing symbols, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the present invention.

[0013] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0014] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0015] A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0016] In this application, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0017] FIG. 1 illustrates an embodiment of a system risk management framework of a nuclear power plant decision support system that includes an operation by a risk management system according to the present invention. FIG. 2 illustrates an embodiment of a system risk management planning method of FIG. 1. FIG. 3 illustrates an embodiment of a method for administrative control and response to system risk of FIG. 1. FIG. 4 illustrates a risk management system according to the present invention. FIG. 5 illustrates an embodiment in which a function corresponding to an input module of an operator indicator is switched.

[0018] The risk management system (100) according to the present invention can be implemented to control the decision support system in a safe mode by performing a test on the decision support system (IDSS, Intelligent Decision Support System) of a nuclear power plant (i.e., a nuclear power plant), generating test data for the decision support system based on the results of the test, and switching the function corresponding to the input module equipped in the operator indicator of the nuclear power plant by comparing the test data with the predetermined correct answer data.

[0019] Here, the decision support system may be implemented to support decision-making for the operation of the nuclear power plant by collecting data measured and generated through sensors or various modules during the operation of the nuclear power plant, and by performing mathematical models or simulations related to the operation of the nuclear power plant based on the collected data to provide information such as the status of the nuclear power plant to the user.

[0020] Therefore, testing of the decision support system may include tests for logic checks and performance tests of the decision support system.

[0021] Here, the logic check can be performed by inputting predetermined input data related to the nuclear power plant into a decision support system and comparing the data output from the decision support system with the correct answer data output from the decision support system during normal operation corresponding to the input data.

[0022] In addition, the performance test may be conducted by measuring the response time from when the decision support system processes predetermined input data until it outputs the data, and by comparing the measured response time with the maximum response time processed by the decision support system during normal operation in response to the corresponding input data.

[0023] Accordingly, the test data may include data output from the decision support system in response to predetermined input data, and may include the response time until the data corresponding to the input data is output from the decision support system.

[0024] In this case, the correct answer data may include data output in response to predetermined input data when the decision support system is operating normally, and the maximum response time.

[0025] Meanwhile, the safety mode can be implemented to monitor the operating status of the nuclear power plant and prevent accidents through a decision support system. To this end, the safety mode can be implemented to provide various data measured or generated at the nuclear power plant in real time, output a warning notification if an anomaly is detected among the plant's data, and perform predetermined inspection procedures.

[0026] In addition, the operator indicator may be provided so that the operator (or manager) of the nuclear power plant can check information generated by the decision support system through the output module, and the input module provided in the operator indicator may be provided to generate control commands for the decision support system according to user input by the operator of the nuclear power plant.

[0027] The input module of such an operator indicator may be equipped with a plurality of buttons to detect user input from each button and transmit a predetermined control command to a decision support system, and at this time, the risk management system (100) may transition (or switch) a function (or control command) corresponding to any one of the plurality of buttons provided in the input module of the operator indicator.

[0028] Accordingly, the risk management system (100) can perform a test on the decision support system based on user input through the input module and output information related to the test results through the operator indicator.

[0029] At this time, the risk management system (100) can be implemented so that when a risk occurs in the decision support system according to the test results, it outputs a warning notification about the risk and switches the function corresponding to the input module, thereby allowing the decision support system to enter a safe mode according to user input.

[0030] Here, risks related to the decision support system may include risks related to the logic of the decision support system and risks related to performance. Risks related to logic are risks arising from logic checks of the decision support system, and may correspond to cases where the data output from the decision support system based on predetermined input data differs from the correct answer data output from the decision support system during normal operation corresponding to said input data. Risks related to performance are risks arising from performance tests of the decision support system, and may correspond to cases where the response time measured by the decision support system based on predetermined input data is longer than the maximum response time processed by the decision support system during normal operation corresponding to said input data.

[0031] Meanwhile, the risk management system (100) can identify risks according to predetermined categories. For example, the predetermined categories may include data, algorithms, models, systems, and human-AI interactions.

[0032] Here, data-related risks may include risks to data used in a decision support system, such as risks to data quality representing errors in the decision support system caused by inaccurate, incomplete, and biased data, and risks to data security representing damages resulting from data leakage, corruption, and misuse.

[0033] In addition, risks associated with algorithms are risks to various algorithms implemented to process specific tasks in a decision support system, and may include risks of algorithm bias that generate biased results for specific situations, risks of algorithm transparency where the intuitiveness (or understanding) of how the algorithm operates is lower than a certain level, and risks of algorithm errors that indicate malfunctions caused by algorithm design or implementation errors.

[0034] Risks related to the model are risks associated with various models used in decision support systems, such as numerical analysis models and learning models, and may include risks of model errors resulting in inaccurate results due to errors in the model learning process, risks of difficulties in interpreting the model that are not intuitive during the interpretation of model results and decision-making processes, and risks of model abuse resulting in malicious outcomes due to the misuse of the model.

[0035] Risks related to the system are risks that may arise from the structure of the decision support system or the process of connecting external devices to the decision support system, and may include risks to system stability representing damage caused by system errors and hacking, risks to system security representing damage caused by the exploitation of system vulnerabilities, and risks to system compatibility representing compatibility issues with other systems.

[0036] In this case, the risk to system security may include damage caused by the exploitation of vulnerabilities in open source or adversarial machine learning models used in the decision support system.

[0037] In addition, risks related to human-AI interaction are risks that may arise during the interaction process between an operator (or manager) and an AI (Artificial Intelligence) system equipped in a decision support system, and may include risks of a lack of human-AI trust resulting from the operator's lack of trust in the AI ​​system, risks of difficulties in human-AI collaboration due to a lack of understanding or proficiency regarding the AI ​​system, and risks of human error representing system errors caused by the operator's incorrect input or misoperation.

[0038] In this regard, the risk management system (100) may be able to identify various risks described above based on a test of the decision support system, or identify the results of the test of the decision support system as a predetermined risk based on the risks described above.

[0039] Accordingly, the driver (or manager) can utilize various risk management measures together with the risk management system (100) to identify various risks as described above.

[0040] For example, data risks related to data quality and security can be minimized by standardizing data management through the creation of data definitions and data quality verification guidelines for training, testing, and verification data, and verifying the data based on these.

[0041] Furthermore, regarding risks associated with algorithms and models, transparency can be ensured during the development process by creating documentation for AI models. By providing information about the models and explanations for result derivation, the impact of algorithm and model risks can be minimized. Additionally, the impact of risks can be minimized by establishing confidence levels and acceptance criteria for output results and applying models with enhanced reliability, thereby strengthening the models' robustness against adversarial machine learning threats. Moreover, the impact of risks can be minimized by developing a model validation plan and performing independent validation for each model based on it.

[0042] Furthermore, system-related risks may include damage caused by system errors and hacking, damage resulting from the exploitation of system vulnerabilities, and damage caused by compatibility issues with other systems; therefore, risk management through open source software utilized in the process of implementing AI technology-based systems is required.

[0043] To this end, a Software Bill of Material (SBOM) created during the development process of a decision support system may be utilized. Such an SBOM is a specification containing information about the software from the initial development stage to the final distribution and utilization stage, and may be created for the purpose of software supply chain management. Such an SBOM may include packaged software, operating systems, frameworks used in software development, and libraries.

[0044] In this regard, in Korea, the information and communications industry standards TTAK.KO-11.0309 (December 2022), "SBOM Attribute Specification for Open Source Software Supply Chain Management," and TTAK.KO-11.0322 (December 2023), "Guidelines for Open Source SBOM Governance Management" have been established based on cooperation between the Ministry of Science and ICT and the Software Policy Research Institute.

[0045] Therefore, in the nuclear power sector as well, it is possible to establish processes suitable for the development and operation of decision support systems based on the standard attribute specifications described in TTAK.KO-11.0309 and the management procedures described in TTAK.KO-11.0322. Accordingly, by continuously verifying the list and versions of open source software at each development and operation stage of the nuclear power plant's decision support system, and by examining operational and security risk factors, the impact of system-related risks can be minimized.

[0046] In addition, regarding system risks caused by adversarial machine learning, decision support systems can be recognized as important cyber security targets for nuclear power plants, and security objectives for data confidentiality, integrity, and availability, as well as a system capable of defending against threats such as data theft and falsification by arbitrary users, may be required. Therefore, the cyber security of decision support systems in nuclear power plants can be evaluated for differential applicability of security measures based on the licensing guidelines for nuclear power plant cyber security, security measure requirements KINAC / RS-015 (December 2016), "Security of Computer and Information Systems in Nuclear Facilities," and the security measure evaluation methodology, NEI 13-10, rev.6, "Cyber ​​Security Control Assessments," and based on this, cyber security evaluation guidelines (or system risk management plans) for nuclear power plant decision support systems can be prepared.

[0047] Accordingly, system risk can be managed by implementing activities at each stage of the Software Development Life Cycle (SDLC) for the nuclear power plant's decision support system in accordance with the contents of the system risk management plan, which includes cybersecurity guidelines.

[0048] According to an embodiment, system risk management for a decision support system may include a system risk management planning stage, a managerial control and response stage for system risk, and a technical control and response stage for system risk, as shown in FIG. 1.

[0049] In such cases, the system risk management planning phase is a phase for planning the security development and operating environment of the decision support system, and may be designed to identify system risks of the decision support system and establish administrative / technical control guidelines and implementation plans to control the identified risks.

[0050] Furthermore, the administrative control and response phase for system risks may be a stage of implementing security development environment management according to the SDLC stages of the decision support system, in accordance with the detailed guidelines and implementation plans defined in the management planning stage. Accordingly, the administrative control and response phase can be handled in a direction that minimizes the impact of system risks by analyzing and evaluating the potential impact of risks while performing risk-specific management controls according to the SDLC stages. Additionally, the administrative control and response phase may include the process of establishing a plan to improve risk management control methods based on the results of the risk impact analysis and evaluation, and implementing each item.

[0051] In addition, the technical control and response phase for system risks is a phase for maintaining the secure operating environment of the decision support system. If problems such as failure, performance degradation, or security threats of the decision support system are identified despite the administrative control and response to system risks during the operation of the decision support system, it may be configured to suggest entering the decision support system's safe mode to warn the operator of the situation and execute a predetermined safe mode of the decision support system.

[0052] These technical control and response steps can be performed by the risk management system (100) according to the present invention, that is, the risk management system (100) according to the present invention can be implemented to enable technical control and response to system risks based on the system risk management planning step and the administrative control and response step of system risks described above.

[0053] Meanwhile, referring to Fig. 2, the system risk management planning stage can be carried out according to the software environment analysis and system risk identification stage during the development and operation phases, the stage of defining managerial and technical control items for system risk, and the stage of establishing implementation guidelines for each managerial and technical control item for system risk.

[0054] Here, the software environment analysis and system risk identification phases for the development and operation stages may be configured to identify threats from adversarial machine learning resulting from the use of open source software and the application of artificial intelligence technology as system risks of the decision support system. Additionally, the development and operation environment of the decision support system may be analyzed to create a preliminary SBOM list of open sources used in the development and operation of the decision support system in accordance with the TTA.KO-11.0309 standard. Furthermore, the differential applicability of security measures for the decision support system may be evaluated based on the Security Measure Evaluation Methodology (NEI 13-10).

[0055] Furthermore, regarding the stage of defining managerial and technical control items for system risk, the decision support system is a system that supports decision-making by nuclear power plant operators in normal / abnormal operating conditions based on the results of the differential applicability assessment of security measures; even if the decision support system fails to perform its pre-implemented functions normally due to adversarial learning threats arising from the use of open source software, it does not adversely affect the safety or safety functions of the nuclear power plant, and since the operator is the ultimate decision-maker regarding nuclear power plant operation, it can be considered that means exist to replace the functions of the decision support system.

[0056] Therefore, a decision support system can be defined as an indirect critical digital asset, and according to the Security Measures Assessment Methodology (NEI 13-10), only the Baseline security control items among the security measures of KINAC / RS-015 apply to indirect critical digital assets, and these Baseline security control items can be configured to be established as administrative control items for system risk regarding the decision support system. For example, the Baseline security controls of a decision support system may consist of physical controls, wireless controls, network controls, media controls, configuration management and evaluation, periodic function checks, and security measure monitoring items.

[0057] In addition, the stage of defining administrative and technical control items for system risk can be configured to evaluate whether licenses and security vulnerabilities of open source software can be addressed by Baseline security control management during the development and operation of the decision support system based on the preliminary SBOM list created during the software environment analysis and system risk identification stages of the development and operation phases, and to finalize the SBOM list for the development and operation of the decision support system.

[0058] The stage of establishing implementation guidelines for each managerial and technical control item of system risk can be structured to prepare a system risk management plan for the decision support system by establishing practical implementation guidelines for the managerial and technical security control items of risk for the decision support system defined in the stage of defining managerial and technical control items of system risk. At this time, documentation such as preparing implementation result reports by continuously implementing the managerial and technical control items of system risk according to the guidelines during the SDLC and operational stages of the decision support system may also be included in the management plan.

[0059] Meanwhile, the administrative control and response phase of system risk can be carried out according to the system risk monitoring phase and the system risk management and minimization phase, as shown in Fig. 3.

[0060] Here, the system risk monitoring phase may be configured to perform SBOM management, Baseline security control management, and risk impact analysis and evaluation to perform risk-specific management controls according to the SDLC phase.

[0061] For example, SBOM management can be configured to continuously implement SBOM management and generate supporting documentation during the SDLC and operational phases of the decision support system in accordance with the implementation guidelines for each administrative control item of the risk management plan for the decision support system. That is, based on the SBOM list finalized during the system risk management plan establishment phase of the decision support system, configuration management of SBOMs is performed by comparing the SBOM list, licenses, and security vulnerabilities at each stage of the SDLC of the decision support system, and implementation data is generated through ledger management for each management task.

[0062] In addition, Baseline security control management can be configured to continuously implement Baseline security control item management and generate evidence during the SDLC and operational phases of the decision support system in the following ways, in accordance with the implementation guidelines for each administrative control item in the system risk management plan for the decision support system.

[0063] At this time, in Baseline security control management, physical control can be configured to set the space for the development and operation of the decision support system as a physical protection zone so that only authorized members can enter, and to record the entry details in an entry log when an unauthorized person enters the control zone.

[0064] In addition, in Baseline security control management, wireless control may be configured to prohibit, in principle, the use of wireless internet for personal purposes by users in the space for the development and operation of decision support systems.

[0065] In addition, in Baseline security control management, network controls can be configured so that users cannot arbitrarily change the IP address assigned to their PCs, and can be configured so that the Internet cannot be accessed through separate means other than an independent internal communication network for development and operation.

[0066] In addition, in Baseline security control management, media control prohibits the use of removable storage media in principle, can be configured to disable the write function of removable storage media by applying control measures such as media security solutions, and can be configured to manage a log of media usage.

[0067] Furthermore, in Baseline security control management, configuration management and evaluation can be configured so that only authorized users can modify devices under development and in operation, and that change requests and processing history are recorded and managed in the configuration management system. Additionally, these configuration management activities can be configured to be subject to evaluation and management by the configuration management manager.

[0068] In addition, in Baseline security control management, periodic function checks and security measure monitoring can be configured to set security check items to conduct security checks at least once a month, and to perform checks on security activities to manage check results and remediation plans.

[0069] Risk impact analysis and evaluation can be configured to assign risk-specific outcome levels, severity, and frequency of occurrence to the decision support system for each SBOM and Baseline security control item, and to qualitatively determine the results of the risk impact analysis and evaluation into predetermined levels (e.g., high, medium, low) based on the assigned frequency and severity.

[0070] As described above, the risk management system (100) can be implemented to manage risks to the decision support system based on various risk management methods, and to enable the decision support system to enter a safe mode when a risk occurs.

[0071] To this end, referring to FIG. 4, the risk management system (100) may include an input unit (110), a storage unit (120), a control unit (130), and an output unit (140).

[0072] The input unit (110) may receive information necessary for the operation of the risk management system (100) according to the present invention. To this end, the input unit (110) may be connected wirelessly or via a wired network to various modules provided in the nuclear power plant and to the input module (11) of the operator indicator (10).

[0073] Accordingly, the input unit (110) can receive data measured or generated from various modules provided in the nuclear power plant, and can receive user input from the input module (11) of the operator indicator (10).

[0074] Additionally, the storage unit (120) may store commands and information necessary for the operation of the risk management system (100) according to the present invention. For example, the storage unit (120) may store information regarding a plurality of functions (e.g., test function, warning recognition function, screen control function, and safety mode function, etc.) that correspond to the input module (11) of the operator indicator (10), and may store data measured from various modules provided in the nuclear power plant or data generated.

[0075] Additionally, the storage unit (120) may store information for testing the decision support system (20), and may store information generated according to the test of the decision support system (20).

[0076] The control unit (130) can control the overall operation of the risk management system (100) according to the present invention. That is, the control unit (130) can test the decision support system (20) based on user input through the input module (11) provided in the operator indicator (10), and control the decision support system (20) to a safe mode based on the test results.

[0077] Specifically, the control unit (130) can receive user input for performing a test of the decision support system (20) through an input module (11) (or input unit (110)) provided in the operator indicator (10) of the nuclear power plant. At this time, the control unit (130) can receive user input requesting a test for risk management of the decision support system (20) through a specific input module (11) provided in the shape of a button in the operator indicator (10) of the nuclear power plant equipped with the decision support system (20).

[0078] To this end, the control unit (130) can designate a function corresponding to an input module (11) provided in the operator indicator (10) as a test function (31) when the decision support system (20) is in operation, as shown in FIG. 5.

[0079] Additionally, the control unit (130) can perform a test of the decision support system (20) based on user input and generate test data for the decision support system (20) based on the results of the test.

[0080] That is, the control unit (130) executes the test function (31) for the decision support system (20) based on user input for the input module (11) designated as the test function (31), and automatically controls the decision support system (20) based on a predetermined test procedure according to the test function (31) to generate test data according to the results of the test.

[0081] To this end, when the test function (31) is executed, the control unit (130) generates predetermined input data and inputs it into the decision support system (20), and can check result data generated in correspondence with the input data from the decision support system (20).

[0082] At this time, the control unit (130) may measure the response time from the time of inputting input data to the decision support system (20) to the time of checking result data.

[0083] Furthermore, the control unit (130) can compare predetermined correct answer data and test data with respect to the decision support system (20), and switch the function corresponding to the input module (11) based on the comparison result.

[0084] To this end, the control unit (130) compares the correct answer data and the test data to determine whether the confidence level of the decision support system (20) satisfies a predetermined acceptance criterion, and can determine whether the result of the logic check of the decision support system (20) is successful based on the result of the determination.

[0085] Additionally, the control unit (130) compares the response time according to the test data with a predetermined maximum response time to determine whether the performance of the decision support system (20) satisfies a predetermined acceptance standard, and can determine whether the performance test result of the decision support system (20) is successful based on the result of the determination.

[0086] Accordingly, if at least one of the logic check result and the performance test result fails, the control unit (130) outputs a warning notification to the decision support system (20) through the operator indicator (10) and can switch the function corresponding to the input module (11) from the test function (31) to the warning recognition function (33) (or safety mode function (37)).

[0087] Additionally, when the control unit (130) receives user input for the input module (11) that has been switched to the warning recognition function (33), it terminates the previously outputted warning notification and outputs an information screen containing information on at least one of the logic check result and the performance test result through the operator indicator (10), and together with this, the function corresponding to the input module (11) can be switched from the warning recognition function (33) to the screen control function (35).

[0088] At this time, the control unit (130) can switch the function corresponding to the input module (11) from the screen control function (35) to the safety mode function (37) when the last information screen among the plurality of information screens pre-implemented for at least one of the logic check result and the performance test result is output through the operator indicator (10).

[0089] Accordingly, the control unit (130) can control the decision support system (20) into a safe mode based on user input to the switched input module (11). That is, when user input is received through the input module (11) that has been switched to correspond to the safe mode function (37), the control unit (130) can stop the operation of the decision support system (20) and control the decision support system (20) to enter a safe mode.

[0090] Additionally, when user input is received through an input module (11) that has been switched to correspond to a safety mode function and the decision support system (20) enters a safety mode, the control unit (130) can switch the function corresponding to the input module (11) from the safety mode function (37) to the safety mode release function (39).

[0091] Accordingly, when the control unit (130) receives user input through the input module (11) that has been switched to correspond to the safety mode release function (39), it can release the safety mode of the decision support system (20) and operate the decision support system (20) to restore it to a normal state.

[0092] The output unit (140) can output information generated by the operation of the risk management system (100) according to the present invention. To this end, the output unit (140) can be connected wirelessly or via a wired network to an output module (13) provided in the operator indicator (10).

[0093] Accordingly, the output unit (140) can output information generated from the decision support system (20) so that the user can visually or audibly confirm it through the output module (13) provided in the driver indicator (10), and can output warning notifications regarding the test of the decision support system (20), information related to the test results, etc. Additionally, depending on the embodiment, the output unit (140) may be connected to a separate device and transmit information generated from the decision support system (20), warning notifications, and information related to the test results, etc., to the said device.

[0094] Based on the configuration of the risk management system (100) examined above, the risk management method will be explained in more detail below.

[0095] FIG. 6 is a flowchart illustrating a risk management method according to the present invention. FIG. 7 is a flowchart illustrating an example of a step of generating test data. FIG. 8 is a flowchart illustrating an example of a step of switching functions. FIG. 9 is a flowchart illustrating an example of a step of controlling in a safe mode.

[0096] Referring to FIG. 6, the risk management system (100) according to the present invention can receive user input to perform a test of the decision support system through an input module provided in the operator indicator of the nuclear power plant (S100).

[0097] Specifically, the risk management system (100) can receive user input requesting a test for risk management of the decision support system through an input module provided in the form of a button on an operator indicator of a nuclear power plant equipped with a decision support system.

[0098] For example, when the decision support system is operating in a normal state, the risk management system (100) can designate a function corresponding to an input module provided in the operator indicator as a test function.

[0099] Accordingly, when the risk management system (100) receives user input for an input module provided on an operator indicator while the decision support system is in operation, it can execute a test function corresponding to the input module.

[0100] The risk management system (100) according to the present invention can perform a test of the decision support system based on user input and generate test data for the decision support system based on the results of the test (S200).

[0101] Specifically, the risk management system (100) can execute a test function for a decision support system based on user input for an input module designated as a test function, and automatically control the decision support system based on a predetermined test procedure according to the test function to generate test data according to the results of the test execution.

[0102] For example, with reference to FIG. 7, when a test function is executed, the risk management system (100) can generate predetermined input data and input it into a decision support system (S210). Here, the input data may be data determined to perform control of the nuclear power plant through the decision support system, and, depending on the embodiment, may be data determined to execute or stop a specific function related to the nuclear power plant, or may be data determined to correspond to sensing data measured within the nuclear power plant.

[0103] Accordingly, the risk management system (100) can check the result data output from the decision support system in response to the input data (S220). At this time, the risk management system (100) may also measure the response time from the time the input data is entered into the decision support system to the time the result data is output (S230).

[0104] Accordingly, the risk management system (100) may generate the previously generated input data-result data pairs and response times, etc., as test data (S240), and the risk management system (100) may generate various information for testing the decision support system as test data according to the embodiment.

[0105] Referring again to FIG. 6, the risk management system (100) according to the present invention compares predetermined correct answer data and test data with respect to a decision support system, and can switch the function corresponding to the input module based on the comparison result (S300).

[0106] Referring to FIG. 8, specifically, the risk management system (100) compares the correct answer data and the test data to determine whether the confidence level of the decision support system satisfies a predetermined acceptance criterion, and based on the result of the determination, determines whether the result of the logic check of the decision support system is successful (S310).

[0107] For example, the risk management system (100) can compare predetermined correct answer data with predetermined input data and test data output from the decision support system corresponding to the input data, and determine the level of trust in the decision support system based on whether the correct answer data and the test data are identical according to the comparison result.

[0108] At this time, when multiple input data are predetermined, the risk management system (100) can calculate the ratio of the correct answer data and the test data being the same (or different) among the multiple input data as a confidence level.

[0109] Accordingly, the risk management system (100) can determine that the result of the logic check for the decision support system is successful when the confidence level satisfies a predetermined acceptance criterion, and can determine that the result of the logic check for the decision support system is a failure when the confidence level deviates from the predetermined acceptance criterion.

[0110] Furthermore, the risk management system (100) compares the response time according to the test data with a predetermined maximum response time to determine whether the performance of the decision support system satisfies a predetermined acceptance standard, and can determine whether the performance test result of the decision support system is successful based on the result of the determination (S320).

[0111] For example, the risk management system (100) can determine the performance of the decision support system based on whether the determined response time is shorter than the maximum response time determined for the input data, and the time interval from when predetermined input data is entered into the decision support system to when test data corresponding to the input data is output (or verified).

[0112] At this time, when multiple input data are predetermined, the risk management system (100) can calculate the ratio of the response time among the multiple input data that is shorter (or longer) than the maximum response time predetermined for each of the multiple input data as the performance of the decision support system.

[0113] Accordingly, the risk management system (100) can determine that the performance test result for the decision support system is successful if the previously calculated performance satisfies a predetermined acceptance criterion, and can determine that the performance test result for the decision support system is a failure if the performance deviates from the predetermined acceptance criterion.

[0114] Furthermore, if at least one of the logic check result and the performance test result fails, the risk management system (100) outputs a warning notification to the decision support system through the operator indicator and can switch the function corresponding to the input module from the test function to the warning recognition function (or safety mode function) (S330).

[0115] For example, the risk management system (100) may output a warning notification through an output module provided together with an input module when at least one of the logic check result and the performance test result fails. At this time, the output module may include a visual or auditory output module, and depending on the embodiment, may include an LED module, an oscillation module, and a speaker module, etc.

[0116] Additionally, the risk management system (100) can switch the function corresponding to the input module from a test function to a warning recognition function, and when it receives user input for the input module, it can terminate the previously outputted warning notification. At this time, the risk management system (100) can output information regarding at least one of the logic check result and the performance test result through the operator indicator along with the termination of the warning notification according to the user input for the input module.

[0117] In this regard, the risk management system (100) can output at least one of the results of a logic check and at least one of the results of a performance test, and at least one of the reliability level according to the logic check and at least one of the performance according to the performance test.

[0118] Additionally, the risk management system (100) may output information related to one or more failed results among the logic check results and the performance test results, or information related to both results.

[0119] Furthermore, when the risk management system (100) receives user input for an input module that has been switched to a warning recognition function, it terminates the previously outputted warning notification and outputs information on at least one of a logic check result and a performance test result through an operator indicator, and together with this, can switch the function corresponding to the input module from the warning recognition function to a screen control function (or a safety mode function).

[0120] For example, when information regarding at least one of the logic check result and the performance test result is output through the operator indicator, the risk management system (100) can switch a function corresponding to an input module to a screen control function so as to sequentially switch through a plurality of information screens that are pre-implemented to display information regarding at least one of the logic check result and the performance test result.

[0121] Here, each of the plurality of information screens may be configured to provide information related to at least one of the logic check results and the performance test results. In one embodiment, regarding the logic check results, information may be displayed such as a confidence level value, whether the correct answer data and the test data are identical for each of the plurality of input data, and information related to input data where the correct answer data and the test data are different (e.g., information about the function of the decision support system corresponding to the input data or information about the nuclear power module, etc.), or regarding the performance test results, information may be displayed such as a performance value, the response time at which the test data was output (or verified) for each of the plurality of input data, and information related to input data where the response time is longer than a predetermined maximum response time (e.g., information about the function of the decision support system corresponding to the input data or information about the nuclear power module, etc.).

[0122] Additionally, the risk management system (100) can switch the function corresponding to the input module from the screen control function to the safety mode function when the last information screen among a plurality of information screens pre-implemented for at least one of the logic check result and the performance test result is output through the operator indicator.

[0123] In this regard, according to an embodiment, when at least one of the success or failure of a logic check result for a decision support system and the success or failure of a performance test result is determined, the risk management system (100) outputs at least one of the success or failure of at least one of the logic check result and the performance test result and at least one of the reliability level according to the logic check and the performance according to the performance test through an operator indicator along with a warning notification for the decision support system, and may switch the function corresponding to the input module from a test function to a safety mode function.

[0124] Referring again to FIG. 6, the risk management system (100) according to the present invention can control the decision support system in a safe mode based on user input to the switched input module (S400).

[0125] Referring to FIG. 9, specifically, when user input is received through an input module switched to correspond to a safety mode function, the risk management system (100) can stop the operation of the decision support system and control the decision support system to enter a safety mode (S410).

[0126] At this time, the risk management system (100) can provide information related to the risk of the decision support system through an operator indicator as the decision support system enters safety mode. Here, the information related to the risk of the decision support system may include a plurality of information screens pre-implemented to display information on at least one of the results of a logic check and the results of a performance test, or may include information (e.g., inspection procedures, etc.) pre-prepared for risk management of the decision support system or the nuclear power plant.

[0127] Additionally, according to an embodiment, when a warning notification for the decision support system is output through the operator indicator, the risk management system (100) can switch the function corresponding to the input module from a test function to a warning recognition function.

[0128] Accordingly, when user input is received through an input module switched to correspond to a warning recognition function, the risk management system (100) provides whether at least one of the logic check result and the performance test result is successful through an operator indicator, and can switch the function corresponding to the input module from the warning recognition function to the safety mode function.

[0129] Next, the risk management system (100) controls the decision support system to enter safety mode when user input is received through an input module switched to correspond to the safety mode function, provides a plurality of pre-implemented information screens to display information on at least one of the logic check result and the performance test result through an operator indicator, and may switch the function corresponding to the input module from a warning recognition function to a screen control function.

[0130] Accordingly, when the last screen among the multiple pre-implemented information screens is provided, the risk management system (100) can switch the function corresponding to the input module from the screen control function to the safety mode release function described later.

[0131] In this way, the risk management system (100) may have the order of providing information on logic check results, performance test results, and entering safety mode set differently according to the embodiment, and to this end, the functions corresponding to the input module may be switched in a different order.

[0132] Furthermore, when the risk management system (100) receives user input through an input module that has been switched to correspond to a safety mode function and the decision support system enters safety mode, the function corresponding to the input module can be switched from the safety mode function to the safety mode release function (S420).

[0133] Accordingly, when the risk management system (100) receives user input through an input module that has been switched to respond to the safety mode release function, it can release the safety mode of the decision support system and operate the decision support system to restore it to a normal state (S430).

[0134] Through the above configurations, the risk management system (100) according to the present invention manages risks to the decision support system based on various risk management measures and is implemented so that the decision support system enters a safe mode when a risk occurs, thereby enabling preparation against license and cyber security threats resulting from the use of open source software for the nuclear power plant's decision support system and adversarial machine learning threats resulting from the use of artificial intelligence technology.

[0135] Furthermore, the present invention described above can be implemented as a program that is executed by one or more processes in an electronic device and stored on a computer-readable recording medium.

[0136] Accordingly, the present invention can be implemented as computer-readable code or instructions on a medium on which a program is recorded. That is, various control methods according to the present invention can be provided in the form of integrated or individual programs.

[0137] Meanwhile, computer-readable media include all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SSD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0138] Furthermore, the computer-readable medium may be a server or cloud storage that includes a storage and is accessible to an electronic device via communication. In this case, the computer may download the program according to the present invention from the server or cloud storage via wired or wireless communication.

[0139] Furthermore, in the present invention, the computer described above is an electronic device equipped with a processor, namely a CPU (Central Processing Unit), and no special limitations are placed on its type.

[0140] Meanwhile, the above detailed description should not be interpreted restrictively in all respects but should be considered exemplary. The scope of the invention shall be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are included within the scope of the invention.

Claims

Claim 1 A risk management method using a risk management system comprises: receiving user input to perform a test of an Intelligent Decision Support System (IDSS) through an input module provided in an operator indicator of a nuclear power plant; performing a test of the IDSS based on the user input and generating test data for the IDSS based on the result of the test; comparing the test data with predetermined correct answer data for the IDSS and switching a function corresponding to the input module based on the comparison result; and controlling the IDSS to a safe mode based on user input for the switched input module, wherein the step of switching a function corresponding to the input module includes, if a logic check result or a performance test result for the IDSS fails based on the comparison result, outputting a warning notification through the operator indicator and switching a function corresponding to the input module from a test function to a warning recognition function. A risk management method comprising the step of switching a function corresponding to the input module from the warning recognition function to a screen control function based on user input to the input module switched to the warning recognition function, and the step of controlling to a safe mode includes, when the last information screen among a plurality of information screens output based on user input to the input module switched to the screen control function is output, switching a function corresponding to the input module from the screen control function to a safe mode function. Claim 2 A risk management method according to claim 1, wherein the step of receiving the user input is to receive a user input requesting the test for risk management of the decision support system through one of the input modules provided in the form of a button on the operator indicator of the nuclear power plant equipped with the decision support system. Claim 3 A risk management method according to claim 1, further comprising the step of designating a function corresponding to the input module provided in the operator indicator as a test function when the decision support system is operating in a normal state. Claim 4 In claim 3, the step of generating the test data comprises: a step of executing a test function for the decision support system based on the user input for the input module designated as the test function; and a step of automatically controlling the decision support system based on a predetermined test procedure according to the test function to generate the test data according to the result of the test execution. Claim 5 In claim 4, the step of executing the test function involves, when the test function is executed, generating predetermined input data and inputting it into the decision support system, and the step of generating the test data involves confirming result data generated from the decision support system in correspondence with the input data as the test data, a risk management method. Claim 6 A risk management method according to claim 1, wherein the step of switching the function comprises: a step of comparing the correct answer data and the test data to determine whether the confidence level of the decision support system satisfies a predetermined acceptance criterion; and a step of determining whether the result of a logic check of the decision support system is successful based on the result of the determination. Claim 7 A risk management method according to claim 1, wherein the step of switching the function comprises: a step of determining whether the performance of the decision support system satisfies a predetermined acceptance criterion by comparing the response time according to the test data with a predetermined maximum response time; and a step of determining whether the performance test result of the decision support system is successful based on the determination result. Claim 8 delete Claim 9 A risk management method according to claim 7, wherein the step of controlling to the safety mode includes, upon receiving user input for the input module switched to the warning recognition function, terminating the outputted warning notification and outputting information regarding the performance test result through the operator indicator. Claim 10 delete Claim 11 delete Claim 12 A risk management method according to claim 1, wherein the step of controlling to a safety mode further comprises the step of stopping the operation of the decision support system and controlling the decision support system to enter the safety mode when user input is received through the input module switched to correspond to the safety mode function. Claim 13 A risk management method according to claim 12, wherein the step of controlling to the safety mode further comprises the step of switching a function corresponding to the input module from the safety mode function to a safety mode release function when the user input is received through the input module switched to correspond to the safety mode function and the decision support system enters the safety mode. Claim 14 An input unit that receives user input to perform a test of a decision support system through an input module provided in the operator indicator of a nuclear power plant; A risk management system comprising a control unit that performs a test of the decision support system based on the user input, wherein the control unit generates test data for the decision support system based on the result of the test, compares the test data with predetermined correct answer data for the decision support system, switches a function corresponding to the input module based on the comparison result, and controls the decision support system to a safe mode based on user input for the switched input module, wherein the control unit outputs a warning notification through the operator indicator when the logic check result or performance test result for the decision support system fails based on the comparison result, switches the function corresponding to the input module from a test function to a warning recognition function, switches the function corresponding to the input module from the warning recognition function to a screen control function based on user input for the input module switched to the warning recognition function, and when the last information screen among a plurality of information screens output based on user input for the input module switched to the screen control function is output, switches the function corresponding to the input module from the screen control function to a safe mode function. Claim 15 A program that is executed by one or more processes in an electronic device and stored on a recording medium readable by a non-transitory computer, wherein the program comprises: receiving user input to perform a test of an Intelligent Decision Support System (IDSS) through an input module provided in an operator indicator of a nuclear power plant; performing a test of the decision support system based on the user input and generating test data for the decision support system based on the result of the test; comparing the test data with predetermined correct answer data for the decision support system and switching a function corresponding to the input module based on the result of the comparison. A program stored on a non-transitory computer-readable recording medium, comprising instructions for performing a step of controlling the decision support system in a safe mode based on user input for the input module switched above, wherein the step of switching the function corresponding to the input module includes, when a logic check result or performance test result for the decision support system fails based on the comparison result, outputting a warning notification through the operator indicator and switching the function corresponding to the input module from a test function to a warning recognition function, and based on user input for the input module switched to the warning recognition function, switching the function corresponding to the input module from the warning recognition function to a screen control function, and wherein the step of controlling in a safe mode includes, when the last information screen among a plurality of information screens output based on user input for the input module switched to the screen control function is output, switching the function corresponding to the input module from the screen control function to a safe mode function.

Citation Information

Patent Citations

  • Apparatus and method for verification human factors of nuclear energy instrument

    KR101591300B1

  • Intelligent condition monitoring method and system for nuclear power plants

    KR1020210109206A

  • The system which synthetic diagnoses the control boardwhich is belong to a control element drive mechanismcontrol system

    KR100759212B1

  • Reinforcement learning model learning method and rule-based nuclear power plant operation automation method using the same

    KR1020230111059A

  • Process management system and method of nuclear power plant

    KR1020240147261A