A method, system and storage medium for intelligent resiliency control of a nuclear facility sampling loop

Abstract

The present application relates to the technical field of nuclear facility sampling detection, and proposes a nuclear facility sampling loop intelligent resilience control method, system and storage medium, which synchronously acquires medium state and radiation space characteristics, inputs a digital twin integrated with a nuclide migration model, simulates nuclide diffusion paths under abnormal conditions and quantitatively evaluates the risks to equipment and personnel, generates alternative reconstruction strategies with the goal of maintaining sampling functions and the constraint of minimizing impact, and decomposes into automatic control and manual operation instructions after critical safety and radiation exposure simulation verification. The system accordingly comprises a three-layer architecture of perception execution, intelligent control and monitoring interaction, and adopts distributed edge computing to realize fast and safe interlocking. The present application realizes intelligent resilience control from passive alarm isolation to active prediction-optimization-verification, and can quickly restore core sampling functions under the premise of ensuring radiation safety when the system is abnormal.

Description

Technical Field

[0001] This invention relates to the field of nuclear facility sampling and testing technology, and in particular to a method, system and storage medium for intelligent resilience control of nuclear facility sampling loops. Background Technology

[0002] The sampling loops for radioactive liquids and gases in nuclear facilities (such as nuclear power plants and reprocessing plants) are critical links in monitoring system status and ensuring operational safety. These loops are constantly exposed to complex radiation fields, and the media may be radioactive, corrosive, or carry entrainment risks. Traditional sampling control relies heavily on preset logic and manual intervention. When faced with anomalies such as pipe scaling / corrosion, micro-leakage, or filter clogging, it is difficult to quickly and quantitatively assess the comprehensive impact of these anomalies on downstream equipment contamination, personnel accessibility, and sample representativeness. Furthermore, it lacks the ability to autonomously generate and verify optimal recovery strategies that maintain core sampling functions while ensuring absolute personnel safety. With the development of digital twin and intelligent decision-making technologies, a new technological path has been provided for building an intelligent resilient control system capable of sensing risks, predicting impacts, and dynamically reconfiguring.

[0003] An existing invention patent with publication number CN119064075A discloses an online alarm-linked rapid sampling device for nuclear power plants. Manual shut-off valve A is connected to an impurity filter; manual shut-off valve A is connected to a sampling pump; the sampling pump is connected to an electronic flow meter and an electric valve B; the sampling pump is connected to a control unit; the control unit is connected to the electronic flow meter; the control unit is connected to electric valve A; the control unit is connected to electric valve B; the electronic flow meter is connected to electric valve A; electric valve A is connected to a sampling bottle; and manual shut-off valve B is connected to a sampling bottle support device and electric valve B.

[0004] As described above, the control logic of such existing technical solutions is static and passive. They cannot predict the migration and diffusion paths of radionuclides and their quantitative impact on downstream equipment, personnel radiation safety, and sampling representativeness when an anomaly occurs. Furthermore, they cannot dynamically generate and verify optimal equipment reconfiguration strategies based on this prediction, prioritizing personnel radiation safety and sampling functionality. This results in the system often only being able to adopt conservative isolation or shutdown measures when facing anomalies, making it difficult to quickly and intelligently restore or maintain core sampling functions while ensuring safety. Summary of the Invention

[0005] In view of this, the present invention proposes a method, system and storage medium for intelligent resilient control of nuclear facility sampling loops based on digital twin prediction of nuclide migration and dynamic generation and verification of a safe reconfiguration strategy, which can restore or maintain the core sampling function under the premise of ensuring safety. This solves the problem that the static passive control of existing equipment cannot predict the impact of nuclide migration and dynamically generate and verify the optimal reconfiguration strategy, thus making it impossible to restore or maintain the core sampling function.

[0006] The technical solution of this invention is implemented as follows: On the one hand, the present invention provides a smart resilience control method for sampling loops in nuclear facilities, comprising the following steps: Simultaneously acquire and fuse medium state characteristics and radiation spatial characteristics; medium state characteristics include quantitative parameters of radioactive scaling / corrosion visual images, activity concentrations of key nuclides in the effluent, and determination results of whether gaseous samples carry droplet risks; radiation spatial characteristics include the local three-dimensional dose rate field distribution of the sampling valve group and the radiation zoning level sequence of the current flow path. The medium state characteristics and radiation space characteristics are input into a digital twin that integrates a radionuclide pipeline deposition model to simulate the migration and diffusion paths of key nuclides under abnormal development trends. Based on the simulation results, the impact of the migration and diffusion of key nuclides on downstream equipment activity accumulation, personnel accessibility, and sampling representativeness is quantitatively evaluated. With the goal of maintaining the highest priority sampling function and minimizing the impact of the aforementioned assessment as the optimization guide, the expected collective dose of the reconstruction path is used as the optimization variable, and the integrity of the radiation zone barrier, the remaining available dose of the equipment, and the taboo of cross-contamination between media are used as constraints. One or more equipment action sequences are generated in the digital twin as alternative reconstruction strategies. In the digital twin, critical safety verification and dynamic radiation irradiation simulation are performed on each alternative reconstruction strategy. After the results of critical safety verification and dynamic radiation irradiation simulation both meet the preset safety threshold, the verified optimal strategy is decomposed into automatic control commands and operation tasks requiring manual intervention. For operations involving high radiation areas, the master-slave teleoperated robotic arm is guided to complete the operation outside the shielded window through an augmented reality interface.

[0007] Based on the above technical solutions, preferably, the medium state characteristics are obtained through a radiation-resistant miniature endoscope, an online gamma spectrometer, and a humidity and particle sensor, specifically including the following steps: The sequential images of radioactive scaling / corrosion inside pipelines acquired by a radiation-resistant miniature endoscope are processed to obtain quantitative parameters of radioactive scaling / corrosion based on the sequential images. The quantitative parameters of radioactive scaling / corrosion include the coverage and average thickness of radioactive scaling on the inner wall of the pipeline, as well as the depth, area and distribution density of corrosion pits. The energy spectrum measured by the online gamma spectrometer is analyzed to identify and calculate the net count rate of the characteristic peaks of predefined key nuclides. After detection efficiency and self-absorption correction, it is converted into the activity concentration of key nuclides in the effluent. The synchronous readings of humidity and particulate sensors are fused to determine whether there is a risk of entrained droplets in the gaseous sample. When the relative humidity is continuously higher than the saturation threshold and the aerosol particle size distribution shows droplet characteristics, it is determined that there is a risk of entrained droplets in the gaseous sample.

[0008] Based on the above technical solutions, a preferred approach utilizes digital twins to simulate the migration and diffusion pathways of key nuclides under abnormal development trends, specifically including: When a micro-leak in the pipeline or an abnormal blockage in the filter is detected, the system takes the location of the leak / pressure difference at the blockage point, the current medium flow rate, temperature, nuclide activity concentration, and pipeline inner wall roughness as inputs, and calls the radionuclide airborne diffusion model or liquid-borne migration model in the digital twin to calculate the spatiotemporal distribution of key nuclides in the complex pipeline network and surrounding environment, predicts the deposition activity of key nuclides on the inner wall of specific downstream equipment, and the level of air or surface contamination near the predetermined maintenance path.

[0009] Based on the above technical solutions, the preferred approach, minimizing the impact on assessment, is achieved through the following optimization method: The impact on downstream equipment activity accumulation, personnel accessibility, and sampling representativeness is quantified into three synergistically optimizable sub-objectives, specifically including: Corresponding to the downstream equipment activity accumulation index, the sub-objective is to minimize the new surface contamination activity of downstream key equipment caused by nuclide migration after the strategy is implemented; Corresponding to the personnel accessibility index, the sub-objective is to minimize the product of the dose rate and the expected dwell time in the area that personnel must approach for the implementation strategy or subsequent maintenance. Corresponding to the sampling representativeness index, the sub-objective is to minimize the distortion of the concentration of key nuclides in the sample caused by flow path switching or medium mixing; The optimization process searches for the Pareto solution set that optimizes the combination of the above sub-objectives within the feasible solution space that satisfies all rigid constraints using a multi-objective optimization algorithm.

[0010] Based on the above technical solutions, the preferred critical security check is as follows: In the digital twin, the effective neutron multiplication factor is calculated for the final configuration of the system after implementing the alternative reconfiguration strategy. This ensures that the effective neutron multiplication factor is lower than the preset critical safety limit and leaves sufficient conservative margin. The final configuration includes the geometric arrangement of all devices, the dielectric filling state, and the reflector conditions. Dynamic radiation exposure simulation, for tasks requiring human intervention, uses a real-time updated three-dimensional dose rate field as a background to dynamically calculate the dose rate changes of the virtual personnel model along the operation trajectory, affecting its whole body and major organs, and performs time integration to obtain the expected effective dose for the entire operation process. For tasks involving master-slave teleoperated robotic arms, the simulation uses a real-time updated three-dimensional dose rate field as a background to dynamically calculate the dose rate changes of the virtual personnel model along the operation trajectory, affecting its whole body and major organs, and performs time integration to obtain the expected absorbed dose for the entire operation process.

[0011] On the other hand, the present invention provides a system applying the above-mentioned intelligent resilience control method for nuclear facility sampling loops, comprising: The perception and execution layer includes a multimodal sensor array for acquiring medium state characteristics and radiation space characteristics, an execution valve pump, and at least one master-slave teleoperated robotic arm; The intelligent control layer includes an intelligent master control unit for running the digital twin, performing impact assessments, generating and verifying refactoring strategies, an anomaly prediction and hierarchical protection unit, and a functional dynamic refactoring unit. The monitoring and interaction layer includes a display terminal for providing an augmented reality interface and a control panel with integrated haptic feedback.

[0012] Based on the above technical solutions, preferably, the multimodal sensor array includes a radiation-resistant miniature endoscope, an online gamma spectrometer, and a radiation field imaging unit; The system adopts a distributed edge computing architecture, deploying regional control nodes with local decision-making capabilities in physical proximity to high-radiation sampling areas; The area control node is configured to quickly execute preset radiation safety interlocking logic based on local sensor data, and when communication with the intelligent control layer is interrupted, it autonomously coordinates the devices in the area to maintain at least one minimum safe sampling or isolation flow path according to the last received valid strategy or local fixed logic.

[0013] Based on the above technical solutions, the preferred local solidification logic includes: When the gamma dose rate in this area is detected to exceed the threshold A, the inlet valves of all sampling branches in this area are automatically closed. When a sudden drop in pressure in a branch is detected, accompanied by an increase in radiation, it is determined to be a leak, and the isolation valves at both ends of the branch are automatically closed and emergency drainage is initiated.

[0014] The augmented reality interface is configured to overlay the real-time three-dimensional dose rate field as a heatmap with the device's three-dimensional model, real-time status, and reconstruction strategy flow verified by the digital twin. It can also calculate and prompt the expected cumulative dose for performing the task in real time based on the task list and virtual personnel trajectory.

[0015] Based on the above technical solutions, the preferred anomaly prediction and hierarchical protection unit is configured to execute a three-level response mechanism: The first level of response generates an alert log and marks the device health as declining when a slow degradation trend is detected in the device's performance parameters. The second-level response, when it is confirmed that the equipment malfunction has affected the function but the system can be maintained through reconstruction, triggers the dynamic functional reconstruction unit and generates a pre-maintenance work order simultaneously; The third-level response, when a fault that could lead to the immediate release of radioactive materials is detected, triggers a hardwired safety logic unit that is completely independent of the software system, which directly takes over the safety-level equipment and executes a safety shutdown sequence.

[0016] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described intelligent resilience control method for nuclear facility sampling loops.

[0017] The intelligent resilience control method, system, and storage medium for nuclear facility sampling loops of the present invention have the following advantages over the prior art: (1) By addressing the critical risk of radionuclides migrating in the sampling loop of nuclear facilities due to anomalies such as microleakage or blockage in the pipeline, the diffusion path is simulated using a digital twin, and the impact on downstream equipment contamination, personnel accessibility, and sample representativeness is quantitatively assessed. Based on this, the system can proactively generate a reconfiguration strategy aimed at maintaining sampling function and minimizing radiation consequences, and pre-complete critical safety verification and personnel radiation dose simulation in virtual space. This effectively solves the problem of existing technologies being passively responsive and unable to quickly restore sampling function while ensuring radiation safety, and realizes proactive resilience control from alarm-isolation to prediction-optimization-verification.

[0018] (2) By setting up a radiation-resistant endoscope to directly observe radioactive scaling and corrosion inside the pipeline, and combining this with an online gamma spectrometer to accurately measure the activity concentration of key nuclides, a realistic source term and boundary condition are provided for nuclide migration prediction. The gaseous / liquid-borne radionuclide migration model adopted can accurately predict the deposition and contamination distribution of nuclides in complex pipelines and surrounding environments during leaks or blockages, thereby transforming the originally difficult-to-quantify radiation consequences into specific equipment contamination activity and spatial dose field data, providing a direct basis for safety decisions.

[0019] (3) By quantifying the core constraints of safe operation of nuclear facilities—namely, equipment contamination accumulation, personnel radiation exposure, and sample representativeness distortion—into co-optimizable sub-objectives, a multi-objective optimization algorithm is used to find the Pareto solution set that achieves the optimal balance between the above objectives within the feasible solution space that satisfies rigid safety constraints such as radiation zone barrier integrity and remaining available dose of equipment. Before the strategy is implemented, the system performs neutron effective multiplication factor calculation and dynamic radiation exposure simulation on the alternative reconstruction strategies in a digital twin to ensure that no scheme will trigger critical risks and to accurately predict and control the occupational radiation dose of personnel, thereby meeting the stringent requirements of nuclear safety regulations for the optimization principle of radiation protection from a design perspective.

[0020] (4) The system architecture of this invention separates intelligent optimization decision-making from local hard safety logic. The built-in solidified logic of the regional control node ensures that even in the most extreme case of communication interruption, the most basic radiation safety isolation can still be performed. The three-level response mechanism realizes hierarchical control from equipment degradation early warning, dynamic functional reconstruction to emergency safety shutdown, ensuring that anomalies can be controlled at an appropriate level. The augmented reality interface visualizes the invisible radiation field, directly guiding the operation of the remote-controlled robotic arm, minimizing the need for personnel to enter high-radiation areas, and fundamentally improving the level of operational safety. Attached Figure Description

[0021] 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 these drawings without creative effort.

[0022] Figure 1 This is a flowchart of the intelligent resilience control method for facility sampling loops according to the present invention; Figure 2 This is a framework diagram of the system of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0024] In the description of the embodiments of the present invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of the present invention based on the specific circumstances.

[0025] In the description of the embodiments of the present invention, it should be noted that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of the present invention.

[0026] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0027] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0028] The following disclosure provides numerous different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. Additionally, examples of various specific processes and materials are provided in this invention; however, those skilled in the art will recognize the applicability of other processes and / or the use of other materials.

[0029] like Figure 1 As shown, the intelligent resilience control method for nuclear facility sampling loops of the present invention includes the following steps: Step S1: Simultaneously acquire and fuse medium state characteristics and radiation space characteristics.

[0030] This step simultaneously acquires and integrates medium state characteristics and radiation spatial characteristics. Medium state characteristics include quantitative parameters of radioactive scaling / corrosion, activity concentrations of key nuclides in the effluent, and the determination results of whether gaseous samples carry the risk of droplets. Radiation spatial characteristics include the local three-dimensional dose rate field distribution of the sampling valve assembly and the radiation zoning sequence traversed by the current flow path.

[0031] The medium state characteristics were obtained using a radiation-resistant miniature endoscope, an online gamma spectrometer, and humidity and particle sensors, specifically including the following steps: The sequential images of radioactive scaling / corrosion inside pipelines acquired by a radiation-resistant miniature endoscope are processed to obtain and quantify the coverage and average thickness of radioactive scaling on the inner wall of the pipeline, as well as the depth, area, and distribution density of corrosion pits.

[0032] Specifically, scale coverage It can be defined as the pixel area of ​​the scaled region. Total visible pixel area of ​​the inner wall The ratio: ; average thickness Calculations can be made using 3D reconstruction or laser ranging data. ; in, For the first The thickness of each sampling point This represents the total number of sampling points.

[0033] Depth of corrosion pits ,area and distribution density It is obtained through the following image segmentation and morphological analysis steps, specifically including: Image and Processing: The acquired sequence of images of the inner wall of the pipeline is converted to grayscale, contrast is enhanced, and noise is filtered, for example, by using median filtering or Gaussian filtering, to improve image quality.

[0034] Corrosion pit region segmentation: Thresholding methods, such as the Otsu method, or edge detection algorithms, such as the Canny operator, are used to initially separate potential corrosion pit regions from the preprocessed image, generating a binary image. In this context, the target area (corrosion pit) has a pixel value of 1, while the background has a pixel value of 0.

[0035] Morphological post-processing, binary images Morphological opening operations are performed, followed by erosion and then dilation, to remove small noise points. Then, morphological closing operations are performed to fill small holes inside the target area, resulting in a more complete image of the eroded pit region. .

[0036] Connectivity analysis and parameter extraction: Connectivity markers were used to identify each individual corrosion pit region. For the first... One area of ​​corrosion pits: area By counting the total number of pixels within the connected region. And combined with camera calibration parameters (the actual area corresponding to each pixel) The calculation yielded the following: .

[0037] depth Utilizing the principle of stereo vision, depth information is calculated based on parallax by matching corresponding points of the same corrosion pit in adjacent frames or binocular endoscopic images; or by estimating depth information using preset scale information and image shadow and texture features.

[0038] Distribution density : Statistical unit analysis area Number of corrosion pits inside ,Right now .in, This is the actual area calculated based on the image field of view and calibration parameters.

[0039] The energy spectrum measured by the online gamma spectrometer was analyzed to identify and calculate the net count rate of characteristic peaks of predefined key nuclides. After detection efficiency and self-absorption correction, this was converted into the activity concentration of key nuclides in the effluent. The calculation formula can be expressed as: ; in, The net count rate of the characteristic peak. For detection efficiency, It is a self-absorption correction factor. For measuring time.

[0040] The readings from the humidity and particulate sensors are fused together to determine whether a gaseous sample carries a risk of droplet entrainment. If the relative humidity remains above the saturation threshold and the aerosol particle size distribution exhibits droplet characteristics, the gaseous sample is deemed to carry a risk of droplet entrainment. This judgment logic can be formalized as follows: like > and Then it is determined that there is a risk; among which Relative humidity, The saturation threshold This refers to the aerosol particle size distribution. This is a set of particle size distributions that characterize droplet features.

[0041] Step S2 simulates the nuclide migration pathway and quantitatively assesses the impact.

[0042] The characteristics of the medium state and the radiation space are input into a digital twin that integrates a radionuclide pipeline deposition model to simulate the migration and diffusion paths of key nuclides under abnormal development trends. Based on the simulation results, the impact of the migration and diffusion of key nuclides on the activity accumulation of downstream equipment, personnel accessibility, and sampling representativeness is quantitatively evaluated.

[0043] Specifically, digital twins are used to simulate the migration and diffusion pathways of key nuclides under abnormal development trends, including: When a micro-leak in the pipeline or an abnormal blockage in the filter is detected, the system takes the location of the leak / pressure difference at the blockage point, the current medium flow rate, temperature, nuclide activity concentration, and pipeline inner wall roughness as inputs, and calls the radionuclide gaseous diffusion model or liquid migration model in the digital twin to calculate the spatiotemporal distribution of key nuclides in the complex pipeline network and surrounding environment, predict the deposition activity of key nuclides on the inner wall of specific downstream equipment, and the level of air or surface contamination near the predetermined maintenance path.

[0044] Specifically, the digital twin selects and calls the appropriate migration model based on the characteristics of the medium state, especially the determination of whether there is a risk of entrained droplets in the gaseous sample: when it is determined that there is a risk of entrained droplets, the liquid migration model is called first or combined with the model for simulation; when it is determined that there is no risk of entrained droplets, the gaseous diffusion model is called for simulation. The migration and diffusion process can be described by the following general transport equation: ; in, Nuclide concentration; For time; For the velocity field; The diffusion coefficient is denoted as . For example, a source term, such as a leak point; Nuclide decay constant. The spatiotemporal distribution of nuclides is obtained by solving this equation. .

[0045] The digital twin uses numerical methods to solve the above partial differential equations. The specific solution process includes: Based on the geometric model of the pipeline network and its surrounding environment, a three-dimensional computational mesh is generated, and the computational domain is spatially discretized. Initial time Concentration field Determined by the background concentration under normal operating conditions. Boundary conditions include: setting source term boundaries (given flux or concentration) at leak or blockage points; setting no-slip or deposition boundaries at solid walls; and setting free outflow or given ambient concentration at open boundaries.

[0046] The transport equations are spatially discretized using the finite volume method for any control volume. Its discretization equation can be expressed as: ; in, For control body At time step The average concentration of nuclides; For control body Volume; For time steps; Representative control body Each facet; For noodles The diffusion coefficient on; For noodles Concentration gradient on; For control body Source term strength within; is the nuclide decay constant.

[0047] Time discretization is performed using a first-order implicit Euler scheme, i.e., the time derivative term... In the control body Discretized as ,in, and These represent the current and next time steps, respectively. Combining the spatial discretization equations with the temporal discretization scheme yields information about the unknowns. Large sparse linear equation systems: ; in, It is a coefficient matrix; Let be the concentration vector on all grid cells in the next time step; To include current time step information The source term and the right-hand side vectors of the boundary conditions are also included. The linear equations can be solved using a computational fluid dynamics solver or a custom program employing iterative algorithms such as the conjugate gradient method or GMRES to obtain the values ​​for each time step. Nuclide concentration on each grid cell ,Right now .

[0048] The obtained spatiotemporal concentration distribution Stored and used for subsequent deposition calculations and dose assessments.

[0049] To quantitatively assess the impact on activity accumulation in downstream equipment, the digital twin further employed a deposition model to calculate the deposition activity of nuclides on the inner wall of the equipment. This deposition process can be modeled as follows: ; in, The deposition activity per unit area of ​​the inner wall of the equipment is expressed in Bq / m². The deposition rate is expressed in m / s and its value is related to the flow regime, surface properties, and nuclide properties. The nuclide concentration in the fluid near the equipment surface is expressed in Bq / m³ and is obtained from the aforementioned transport equation. The decay constant of a nuclide is expressed in s. -1 ; By applying this differential equation within the policy execution time window By performing integration, the increased surface contamination activity of downstream key equipment that may be caused by nuclide migration can be quantitatively calculated. This provides direct input for subsequent optimization objectives.

[0050] Step S3: Generate alternative refactoring strategies.

[0051] With the goal of maintaining the highest priority sampling function and minimizing the impact of the aforementioned assessment as the optimization guideline, the expected collective dose of the reconstruction path is used as the optimization variable. The integrity of the radiation zone barrier, the remaining available dose of the equipment, and the taboo of cross-contamination between media are used as constraints. One or more equipment action sequences are generated in the digital twin as alternative reconstruction strategies.

[0052] Minimizing the impact of the assessment is achieved through the following optimization approach: the impact on downstream equipment activity accumulation, personnel accessibility, and sampling representativeness is quantified into three synergistically optimizable sub-objectives, specifically including: Corresponding to the downstream equipment activity accumulation index, the sub-objective is to minimize the additional surface contamination activity of downstream key equipment that may result from nuclide migration after the strategy is implemented. .

[0053] Corresponding to the personnel accessibility index, the sub-objective is to minimize the product of the dose rate and the expected dwell time in the area that personnel must approach for implementation strategies or subsequent maintenance. This expected radiation dose... It can be quantified as: ; in, For position Dose rate at the site For the estimated length of stay, This refers to the operational trajectory of a person or robotic arm.

[0054] Corresponding to the sampling representativeness index, the sub-objective is to minimize the distortion of the concentration of key nuclides in the sample caused by flow path switching or medium mixing. This distortion... It can be defined as: ; in, The concentration of key nuclides under standard operating conditions. The concentration after flow path switching or mixing.

[0055] The rigid safety constraints that the optimization process must satisfy are specifically defined as follows: Any reconfiguration strategy must not disrupt or bypass established physical barriers to radiation zones (such as shielding walls or sealed doors). This constraint is defined by a constraint function. To formalize it. Among them, decision variables... This represents a specific alternative reconfiguration strategy, namely a sequence of ordered equipment operation instructions (such as opening valve V001 and closing pump P002). Function Its function is to examine sequences Does the sequence contain any explicitly prohibited operations (e.g., opening isolation valve ISO-101 or instructing a robotic arm to pass through shielded door H-203)? If any such violation is included, then 0; conversely, if the sequence Fully compliant, then 0. Therefore, this constraint condition The zero-guarantee optimization algorithm will only generate policies that do not contain specific violations.

[0056] For equipment with cumulative dose limits (such as certain radiation-sensitive instruments), the additional dose expected to be received after the strategy is implemented must not exceed its remaining available dose limit. This constraint can be expressed as: ,in To execute the strategy The expected additional dose for the device.

[0057] To prevent accidental mixing of media with different radioactivity levels, the strategy must not lead to connectivity between specific flow paths, which is prohibited in the process design. This constraint is defined by a constraint function. To formalize it. Function Based on the pipeline network model in the digital twin, the execution strategy is simulated. The final system state is then checked, and it is determined whether any prohibited flow paths are connected. If so, then... ;otherwise, .

[0058] The optimization process searches for the Pareto solution set that optimizes the combination of the above sub-objectives within the feasible solution space that satisfies all the above rigid constraints using a multi-objective optimization algorithm (such as the constrained NSGA-II algorithm). This constrained multi-objective optimization problem can be fully formalized as follows: ; ; in, For decision variables, it represents an alternative reconfiguration strategy, i.e., a specific sequence of equipment actions; These correspond to the three sub-objective functions mentioned above. For the first A constraint function, The total number of constraints.

[0059] Step S4: Security verification, simulation verification, and instruction decomposition.

[0060] In the digital twin, critical safety verification and dynamic radiation irradiation simulation are performed on each alternative reconstruction strategy. After the results of critical safety verification and dynamic radiation irradiation simulation both meet the preset safety threshold, the verified optimal strategy is decomposed into automatic control commands and operation tasks requiring manual intervention. For operations involving high radiation areas, the master-slave teleoperated robotic arm is guided to complete the operation outside the shielded window through an augmented reality interface.

[0061] The critical safety check specifically involves: In the digital twin, the effective neutron multiplication factor is calculated for the final system configuration after implementing alternative reconfiguration strategies. This ensures that the effective neutron multiplication factor is below a preset critical safety limit and that sufficient conservative margin is maintained. The final configuration includes the geometric arrangement of all devices, the dielectric filling state, and the reflector conditions.

[0062] Among these factors, the geometric arrangement of the equipment determines the probability and spatial distribution of neutron leakage; the state of the medium filling (such as the type, concentration, and density of the solution) directly affects the neutron moderation capability and absorption cross-section; and the reflector conditions affect the proportion of neutrons reflected back to the active region of the system. These parameters together constitute the physical model inputs necessary for critical safety analysis.

[0063] Specifically, it is necessary to ensure that the calculated effective neutron multiplication factor is accurate. satisfy: ,in To maintain a conservative margin, This is a preset critical safety limit. The neutron transport equations are usually obtained by solving the Monte Carlo method (such as MCNP).

[0064] Dynamic radiation exposure simulation, for tasks requiring human intervention, uses a real-time updated three-dimensional dose rate field as a background to dynamically calculate the dose rate changes of a virtual personnel model as it moves along the operation trajectory, affecting its whole body and major organs. Time integration is then performed to obtain the predicted effective dose for the entire operation. The effective dose for the whole body can be calculated using the following formula: ; in, Representing different tissues or organs; For the organization equivalent dose rate, The tissue weighting factor is a dimensionless coefficient that reflects the relative sensitivity of different tissues to the stochastic effects of radiation (such as cancer). Its specific value adopts the recommended values ​​in International Commission on Radiation Protection (ICRP) Publication No. 103. Indicates time The score is calculated, and the score range is the total time taken by the personnel to perform the operation. It is a position vector, and it is time. The function describes the operational trajectory of a person (or the end effector of a robotic arm) moving in a three-dimensional radiation field.

[0065] Organizational weighting factor The specific values ​​are shown in the table below:

[0066] For the operation tasks of master-slave teleoperated robotic arms, the dose rate changes of the virtual personnel model and its main organs are dynamically calculated as the virtual personnel model moves along the operation trajectory against the background of a real-time updated three-dimensional dose rate field, and time integration is performed to obtain the expected absorbed dose for the entire operation process. Specifically, its dynamic radiation exposure simulation aims to assess the radiation exposure experienced by the robotic arm during task execution, in order to predict radiation damage to its critical components and maintenance needs. The simulation calculates the absorbed dose rate at the locations traversed by key components such as the joint drive units, control units, and end effectors of the robotic arm as they move along the operating trajectory verified by the digital twin. Integrating over time yields the expected absorbed dose of the component. : ; in, Indicates time The integral is the total time taken for the robotic arm to perform the operation.

[0067] like Figure 2 As shown, the intelligent resilience control system for nuclear facility sampling loops using the above method of the present invention includes: The system adopts a layered distributed architecture, comprising a monitoring and interaction layer, an intelligent control layer, and a sensing and execution layer from top to bottom. Data and command exchange between these layers occurs via industrial Ethernet, fiber optics, or an industrial wireless private network with redundant ring network capabilities, ensuring real-time and reliable communication. The system acquires multimodal data from the physical world through the sensing and execution layer, performs intelligent analysis and decision-making driven by a digital twin in the intelligent control layer, and finally completes closed-loop control of the physical loop and personnel interaction through the actuators in the monitoring and interaction layer and the sensing and execution layer.

[0068] The sensing and execution layer includes a multimodal sensor array for acquiring medium state characteristics and radiation space characteristics, an execution valve pump, and at least one master-slave teleoperated robotic arm; the sensing and execution layer serves as the interface for interaction between the system and the physical sampling loop, specifically including a multimodal sensor array, an execution valve pump, and at least one master-slave teleoperated robotic arm.

[0069] The intelligent control layer includes an intelligent master control unit for running the digital twin, performing impact assessments, generating and verifying refactoring strategies, an anomaly prediction and hierarchical protection unit, and a functional dynamic refactoring unit.

[0070] The intelligent control unit runs a digital twin. This digital twin is built on Unity3D, Unreal Engine, or a dedicated simulation platform, integrating geometric models, including precise 3D models of all equipment, pipes, valves, and supporting structures in the sampling loop; physical and process models, integrating a radionuclide pipeline deposition model, a computational fluid dynamics solver, a radiation dose calculation engine, and a multi-objective optimization algorithm library. The intelligent control unit receives real-time data from the perception and execution layer, driving the digital twin to run synchronously, performing a series of computational tasks such as nuclide migration simulation, impact assessment, strategy optimization, and safety verification as described in the method embodiment.

[0071] The monitoring and interaction layer includes a display terminal providing an augmented reality interface and a control panel with integrated haptic feedback. Specifically, the augmented reality interface is presented through a large stereoscopic display screen, and the control panel with integrated haptic feedback is used to operate the master-slave telescopic robotic arm. The force feedback handle on the control panel can realistically reproduce the force / torque when the slave arm interacts with the environment, providing the operator with a sense of presence and ensuring the accuracy and safety of precise operations.

[0072] Furthermore, the multimodal sensor array includes a radiation-resistant miniature endoscope, an online gamma spectrometer, and a radiation field imaging unit; The radiation-resistant miniature endoscope employs a tungsten alloy or lead composite material shielding shell and is equipped with a radiation-hardened optical lens and image transmission bundle. It is fixedly mounted via a flange onto a dedicated observation window at sampling branch bends, before or after filters, or near sampling valves. The endoscope's built-in miniature CMOS image sensor acquires sequential images of the pipeline's inner wall, transmitting them via the fiber optic image transmission bundle or radiation-hardened cable to an image processing unit located in a low-radiation zone. The image processing unit uses machine learning algorithms (such as the U-Net network) to segment and extract features from the images, quantifying scale coverage, thickness, and corrosion pit parameters.

[0073] The online gamma-ray spectrometer preferably employs a high-purity germanium detector and is housed in a lead-shielded sampling chamber. The detector is aligned with the flowing medium to acquire gamma-ray spectra in real time. The spectrometer's built-in multichannel analyzer and embedded processor perform spectrum stabilization, background subtraction, and characteristic peak identification, and upload the nuclide identification results and net count rate of characteristic peaks via network. Activity concentration is calculated using a formula, where the detection efficiency ε and the self-absorption correction factor f are pre-calibrated experimentally.

[0074] Radiation field imaging unit: This unit consists of multiple radiation detector arrays arranged around the sampling valve group area. Using the spatial response data of the detector arrays, combined with deconvolution algorithms or Kriging interpolation, the two-dimensional or three-dimensional dose rate field distribution of the area is reconstructed in real time and transmitted as data.

[0075] It also includes process sensors for monitoring pressure, temperature, flow rate, and pH, as well as microenvironment sensors for detecting humidity and aerosol particle size distribution in gaseous samples, which together form a comprehensive sensing network.

[0076] Actuators include pneumatic / electric isolation valves, regulating valves, sampling pumps, etc. These devices are all equipped with intelligent positioners or frequency converters to receive standard industrial control signals from the control layer, enabling precise opening control, start-stop, and speed regulation.

[0077] The actuator includes a master-slave teleoperated robotic arm, with at least one set deployed at the shielded window between the high-radiation sampling room and the operating corridor. The master arm (operating end) is located at the control panel in the low-radiation zone, while the slave arm (actuating end) is located in the high-radiation zone. High-precision synchronous movement and force feedback are achieved between the master and slave arms via a real-time motion controller and force feedback system, enabling the operator to perform precise disassembly, assembly, and operation of valve handwheels, connectors, and other equipment within the high-radiation zone from a safe area.

[0078] The system adopts a distributed edge computing architecture, deploying regional control nodes with local decision-making capabilities in physical proximity to high-radiation sampling areas.

[0079] The area control node is configured to quickly execute preset radiation safety interlocking logic based on local sensor data, and when communication with the intelligent control layer is interrupted, it autonomously coordinates the devices in the area to maintain at least one minimum safe sampling or isolation flow path according to the last received valid strategy or local fixed logic.

[0080] The local solidification logic includes: detecting that the gamma dose rate in the local area exceeds a threshold. When this occurs, the inlet valves of all sampling branches in the area will be automatically shut off via an interlocking mechanism.

[0081] When a sudden drop in pressure in a branch is detected, accompanied by an increase in radiation, it is determined to be a leak, and the isolation valves at both ends of the branch are automatically closed and emergency drainage is initiated.

[0082] The augmented reality interface is configured to overlay the real-time three-dimensional dose rate field as a heatmap with the device's three-dimensional model, real-time status, and reconstruction strategy flow verified by the digital twin. It can also calculate and prompt the expected cumulative dose for performing the task in real time based on the task list and virtual personnel trajectory.

[0083] Furthermore, the anomaly prediction and hierarchical protection unit is configured to execute a three-level response mechanism: The first level of response generates an alert log and marks the device health as declining when a slow degradation trend is detected in the device's performance parameters. Specifically, the system continuously monitors the time series of equipment performance parameters. When trend analysis or machine learning models identify a parameter exhibiting a slow degradation trend (e.g., the closing time of a pneumatic valve increases linearly), but the function has not yet been lost, an early warning log is generated. In the digital twin interface, the corresponding equipment is marked as having declining health, such as by turning the icon yellow, and a pre-maintenance message is displayed.

[0084] The second-level response, when it is confirmed that the equipment malfunction has affected the function but the system can be maintained through reconstruction, triggers the dynamic functional reconstruction unit and generates a pre-maintenance work order simultaneously; Specifically, when it is confirmed that a device malfunction has affected its core sampling function, but the system as a whole still has redundant paths, the dynamic functional reconfiguration unit is immediately triggered to initiate a dynamic reconfiguration process aimed at maintaining the highest priority sampling function. At the same time, a pre-maintenance work order is automatically generated and pushed to the maintenance management system.

[0085] The functional dynamic reconfiguration unit receives alternative reconfiguration strategies (Pareto solutions) from the intelligent master control unit. It first performs critical safety verification and dynamic radiation exposure simulation (calculating the effective dose H) for each strategy in the digital twin. After successful verification, the optimal strategy is decomposed into two types of executable instructions: one is a sequence of device control instructions that can be automatically executed by the distributed control system; the other is a list of structured operational tasks that require manual or remotely operated robotic arms, containing steps, tools, expected dose, and augmented reality guidance information.

[0086] The third-level response, when a fault that could lead to the immediate release of radioactive materials is detected, triggers a hard-wired safety logic unit that is completely independent of the software system, which directly takes over the safety-level equipment and executes a safety shutdown sequence. Specifically, when sensor data fusion identifies a fault that could immediately lead to the release of radioactive materials (such as a pipeline rupture or compromised containment integrity), a hard-wired safety logic unit, completely independent of the software system, is triggered. This unit, implemented by a safety-grade programmable logic controller, is directly connected via hard-wiring to safety-grade equipment such as emergency isolation valves and emergency pumps. Once triggered, this unit will ignore any upper-level software instructions and directly execute a pre-programmed safety shutdown sequence, achieving the highest level of safety protection.

[0087] The present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described intelligent resilience control method for nuclear facility sampling loops.

[0088] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent resilience control of a sampling loop in a nuclear facility, characterized in that, Includes the following steps: Simultaneously acquire and fuse medium state characteristics and radiation space characteristics; The medium state characteristics include quantitative parameters of radioactive scaling / corrosion, activity concentration of key nuclides in the effluent, and the determination results of whether the gaseous sample carries the risk of droplets; the radiation spatial characteristics include the local three-dimensional dose rate field distribution of the sampling valve group and the radiation zoning level sequence of the current flow path. The medium state characteristics and radiation space characteristics are input into a digital twin that integrates a radionuclide pipeline deposition model to simulate the migration and diffusion paths of key nuclides under abnormal development trends. Based on the simulation results, the impact of the migration and diffusion of key nuclides on downstream equipment activity accumulation, personnel accessibility, and sampling representativeness is quantitatively evaluated. With the goal of maintaining the highest priority sampling function and minimizing the impact of the aforementioned assessment as the optimization guide, the expected collective dose of the reconstruction path is used as the optimization variable, and the integrity of the radiation zone barrier, the remaining available dose of the equipment, and the taboo of cross-contamination between media are used as constraints. One or more equipment action sequences are generated in the digital twin as alternative reconstruction strategies. In the digital twin, critical safety verification and dynamic radiation irradiation simulation are performed on each alternative reconstruction strategy. After the results of critical safety verification and dynamic radiation irradiation simulation both meet the preset safety threshold, the verified optimal strategy is decomposed into automatic control commands and operation tasks requiring manual intervention. For operations involving high radiation areas, the master-slave teleoperated robotic arm is guided to complete the operation outside the shielded window through an augmented reality interface.

2. The intelligent resilience control method for nuclear facility sampling loops as described in claim 1, characterized in that, The medium state characteristics are obtained using a radiation-resistant miniature endoscope, an online gamma spectrometer, and humidity and particle sensors, specifically including the following steps: The sequential images of radioactive scaling / corrosion inside pipelines acquired by a radiation-resistant miniature endoscope are processed to obtain quantitative parameters of radioactive scaling / corrosion based on the sequential images. The quantitative parameters of radioactive scaling / corrosion include the coverage and average thickness of radioactive scaling on the inner wall of the pipeline, as well as the depth, area and distribution density of corrosion pits. The energy spectrum measured by the online gamma spectrometer is analyzed to identify and calculate the net count rate of the characteristic peaks of predefined key nuclides. After detection efficiency and self-absorption correction, it is converted into the activity concentration of key nuclides in the effluent. The synchronous readings of humidity and particulate sensors are fused to determine whether there is a risk of entrained droplets in the gaseous sample. When the relative humidity is continuously higher than the saturation threshold and the aerosol particle size distribution shows droplet characteristics, it is determined that there is a risk of entrained droplets in the gaseous sample.

3. The intelligent resilience control method for nuclear facility sampling loops as described in claim 1, characterized in that, The migration and diffusion pathways of key nuclides under abnormal development trends are simulated using digital twins, specifically including: When a micro-leak in the pipeline or an abnormal blockage in the filter is detected, the system takes the location of the leak / pressure difference at the blockage point, the current medium flow rate, temperature, nuclide activity concentration, and pipeline inner wall roughness as inputs, and calls the radionuclide airborne diffusion model or liquid-borne migration model in the digital twin to calculate the spatiotemporal distribution of key nuclides in the complex pipeline network and surrounding environment, predicts the deposition activity of key nuclides on the inner wall of specific downstream equipment, and the level of air or surface contamination near the predetermined maintenance path.

4. The intelligent resilience control method for nuclear facility sampling loops as described in claim 1, characterized in that, Minimizing the impact of the assessment is achieved through optimization by quantifying the impact on downstream equipment activity accumulation, personnel accessibility, and sampling representativeness into three synergistically optimizable sub-objectives, specifically including: Corresponding to the downstream equipment activity accumulation index, the sub-objective is to minimize the new surface contamination activity of downstream key equipment caused by nuclide migration after the strategy is implemented; Corresponding to the personnel accessibility index, the sub-objective is to minimize the product of the dose rate and the expected dwell time in the area that personnel must approach for the implementation strategy or subsequent maintenance. Corresponding to the sampling representativeness index, the sub-objective is to minimize the distortion of the concentration of key nuclides in the sample caused by flow path switching or medium mixing; The optimization process searches for the Pareto solution set that optimizes the combination of the above sub-objectives within the feasible solution space that satisfies all rigid constraints using a multi-objective optimization algorithm.

5. The intelligent resilience control method for nuclear facility sampling loops as described in claim 1, characterized in that, The critical security check specifically refers to: In the digital twin, the effective neutron multiplication factor is calculated for the final configuration of the system after implementing the alternative reconfiguration strategy to ensure that the effective neutron multiplication factor is lower than the preset critical safety limit and has sufficient conservative margin. The final configuration includes the geometric arrangement of all devices, the dielectric filling state and the reflector conditions. The dynamic radiation simulation, for tasks requiring human intervention, uses a real-time updated three-dimensional dose rate field as a background to dynamically calculate the dose rate changes of the virtual personnel model along the operation trajectory, affecting its whole body and major organs, and performs time integration to obtain the expected effective dose for the entire operation process. For tasks involving master-slave teleoperated robotic arms, the simulation uses a real-time updated three-dimensional dose rate field as a background to dynamically calculate the dose rate changes of the virtual personnel model along the operation trajectory, affecting its whole body and major organs, and performs time integration to obtain the expected absorbed dose for the entire operation process.

6. A system applying the intelligent resilience control method for nuclear facility sampling loops as described in any one of claims 1 to 5, characterized in that, include: The perception and execution layer includes a multimodal sensor array for acquiring the state characteristics and radiation space characteristics of the medium, an execution valve pump, and at least one master-slave teleoperated robotic arm; The intelligent control layer includes an intelligent main control unit for running the digital twin, performing impact assessments, generating and verifying the reconstruction strategy, an anomaly prediction and hierarchical protection unit, and a functional dynamic reconstruction unit. The monitoring and interaction layer includes a display terminal for providing the augmented reality interface and a control panel with integrated haptic feedback.

7. The system as described in claim 6, characterized in that: The multimodal sensor array includes a radiation-resistant miniature endoscope, an online gamma spectrometer, and a radiation field imaging unit; The system adopts a distributed edge computing architecture, deploying regional control nodes with local decision-making capabilities in physical proximity to the high-radiation sampling area; The area control node is configured to quickly execute preset radiation safety interlocking logic based on local sensor data, and when communication with the intelligent control layer is interrupted, it autonomously coordinates the devices in the area to maintain at least one minimum safe sampling or isolation flow path according to the last received valid strategy or local fixed logic.

8. The system as described in claim 7, characterized in that, The local firmware logic includes: When the gamma dose rate in this area is detected to exceed the threshold A, the inlet valves of all sampling branches in this area are automatically closed. When a sudden drop in pressure in a branch is detected, accompanied by an increase in radiation, it is determined to be a leak. The isolation valves at both ends of the branch are automatically closed and the emergency drainage is activated. The augmented reality interface is configured to overlay the real-time three-dimensional dose rate field as a heatmap with the device's three-dimensional model, real-time status, and reconstruction strategy flow verified by the digital twin. It can also calculate and prompt the expected cumulative dose for performing the task in real time based on the task list and virtual personnel trajectory.

9. The system as described in claim 8, characterized in that, The anomaly prediction and hierarchical protection unit is configured to execute a three-level response mechanism: The first level of response generates an alert log and marks the device health as declining when a slow degradation trend is detected in the device's performance parameters. The second-level response, when it is confirmed that the equipment malfunction has affected the function but the system can be maintained through reconstruction, triggers the dynamic functional reconstruction unit and generates a pre-maintenance work order simultaneously; The third-level response, when a fault that could lead to the immediate release of radioactive materials is detected, triggers a hardwired safety logic unit that is completely independent of the software system, which directly takes over the safety-level equipment and executes a safety shutdown sequence.

10. A computer-readable storage medium, characterized in that: The system contains a computer program that, when executed by a processor, implements the intelligent resilience control method for the nuclear facility sampling loop as described in any one of claims 1 to 5.

Images