Radiation adaptive shielding nuclear power distribution monitoring system

Abstract

The invention discloses a radiation adaptive shielding nuclear power distribution monitoring system, which relates to the technical field of nuclear power safety and intelligent monitoring and comprises a multi-physical field sensing layer, a digital twin central layer, a collaborative execution layer and a closed-loop control layer. The multi-physical field sensing layer collects radiation, temperature, vibration and crack data; the digital twinborn central layer constructs a three-dimensional model, the residual life is predicted based on a multi-physics coupling algorithm, and the algorithm comprises a formula integrating radiation, temperature, vibration and crack damage factors; the collaborative execution layer comprises an MXene self-repairing shielding body and an SMA-hydraulic collaborative driving dynamic shielding device which are respectively used for realizing self-repairing and rapid adjustment of shielding thickness; and the closed-loop control layer feeds back the correction model and optimizes parameters. The method breaks through the limitation of traditional passive protection, achieves the dual functions of self-repairing and intelligent adjustment, improves the service life prediction accuracy, prolongs the service life of the shielding body, shortens the response time, and improves the reliability and economical efficiency of the system.

Description

Technical Field

[0001] This invention relates to the field of nuclear power safety and intelligent monitoring technology, specifically to a radiation-adaptive shielded nuclear power distribution monitoring system. Background Technology

[0002] In nuclear power engineering, the power distribution system, as the core link in energy transmission, is constantly exposed to complex environments such as strong radiation, high temperature, and vibration. Its operational safety directly affects the stable operation of the nuclear power plant. Traditional nuclear power power distribution systems mostly use fixed lead plates or concrete structures for radiation shielding, which can only achieve static protection and cannot adjust the shielding effectiveness according to the dynamic changes in the radiation field. When radiation leakage occurs or the shielding capacity decreases due to equipment aging, it can easily cause personnel to receive excessive radiation doses or equipment failures.

[0003] While some existing nuclear power plant monitoring systems attempt to combine sensors and control modules for radiation monitoring, they are mostly limited to the acquisition and alarm of single physical quantities, lacking monitoring of the shielding structure's own condition (such as cracks or aging), and failing to form a closed-loop control system of "monitoring-prediction-repair-adjustment." Furthermore, the maintenance of traditional shielding structures relies on regular manual inspections and replacements, which is not only costly but also poses safety hazards due to untimely inspections, making it difficult to meet the demands of next-generation nuclear power plants for intelligent and highly reliable protection.

[0004] In view of the above, this application is hereby submitted. Summary of the Invention

[0005] The purpose of this invention is to provide a radiation-adaptive shielded nuclear power plant power distribution monitoring system to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides a radiation-adaptive shielded nuclear power distribution monitoring system, comprising: A multi-physics sensing layer is used to collect data on radiation, temperature, vibration, and cracks in the shield and its surrounding environment. The digital twin central layer constructs a three-dimensional model of the shield and predicts its remaining lifetime based on a multiphysics coupling algorithm, which includes the following formula: ; in Theoretical lifespan, , , , These are radiation, temperature, vibration, and crack damage factors, respectively. , , , These are the weighting coefficients; The collaborative execution layer includes an MXene self-healing shield and an SMA-hydraulic collaboratively driven dynamic shielding device; The closed-loop control layer corrects the prediction model and optimizes system parameters based on the feedback from the execution results. The MXene self-healing shield achieves autonomous repair by driving MXene nanosheets to migrate to the crack region via an electric field. The SMA-hydraulic synergistic drive device enables rapid adjustment of the shield thickness through the synergistic effect of shape memory alloy and hydraulic system. This combination of MXene self-healing and SMA-hydraulic synergistic drive technology overcomes the limitations of passive protection in traditional nuclear power shielding systems, achieving a dual function of "autonomous repair and intelligent adjustment." Simultaneously, the multi-field coupling lifetime prediction formula comprehensively considers various damage factors, significantly improving the accuracy of lifetime prediction, providing more precise guidance for system maintenance strategies, greatly extending the service life of the shield, and shortening response time.

[0007] Furthermore, the multiphysics sensing layer includes: a neutron-gamma composite detector, a temperature sensor, a vibration sensor, and an industrial camera distributed on the surface of the shield; a data acquisition unit for receiving and preprocessing data from each sensor; and an optical fiber transmission module for transmitting the preprocessed data to the digital twin central layer. The multi-dimensional sensor arrangement enables comprehensive monitoring of the shield's condition, significantly improving data integrity and anomaly identification accuracy compared to traditional single-point monitoring, and providing a more reliable basis for system decision-making.

[0008] Furthermore, the digital twin central layer includes: a three-dimensional physical model of the shield constructed based on laser scanning data; a multi-field coupling lifetime prediction module to predict the remaining lifetime; and a parameter correction module to dynamically adjust the prediction model parameters based on feedback from the closed-loop control layer. The multi-field coupling prediction model of the digital twin central layer significantly improves the lifetime prediction accuracy compared to the traditional single physical field model, provides early warning of potential failure risks, and offers a scientific basis for preventive maintenance.

[0009] Furthermore, the MXene self-healing shield comprises: MXene nanosheets uniformly dispersed in an epoxy resin matrix; an embedded copper electrode network for applying an electric field to drive the migration of the MXene nanosheets; and a repair trigger module that automatically initiates an electric field repair program when the crack width exceeds a threshold. The MXene self-healing technology enables in-situ autonomous repair of cracks in the shield without manual intervention, significantly reducing costs and downtime compared to traditional replacement maintenance, and significantly improving system reliability.

[0010] Furthermore, the SMA-hydraulic synergistic drive device includes: a Ni-Ti shape memory alloy spring that rapidly contracts and deforms upon energization; a hydraulic push rod system that provides continuous and stable thrust; a linkage mechanism that combines the rapid action of the SMA spring with the stable thrust of the hydraulic push rod; and a shielding thickness adjustment module that dynamically adjusts the thickness of the shielding material according to the radiation level. The SMA-hydraulic synergistic drive mechanism combines the rapid response of the shape memory alloy with the high thrust characteristics of the hydraulic system, providing sufficient structural stability while ensuring the shielding adjustment speed, significantly improving response speed and adjustment accuracy compared to a single drive method.

[0011] Furthermore, the closed-loop control layer includes: a feedback data acquisition module, which acquires the action effect data of the execution layer in real time; an error calculation module, which compares the predicted value with the actual value and calculates the deviation; and a parameter optimization module, which dynamically adjusts the sensor sampling frequency, repair threshold, and shielding adjustment strategy based on a reinforcement learning algorithm. The closed-loop control mechanism enables the system to continuously optimize its own parameters according to the actual operating effect, achieve adaptive adjustment, and significantly improve environmental adaptability and energy efficiency ratio compared with traditional fixed parameter systems, while reducing ineffective energy consumption and excessive maintenance.

[0012] Furthermore, the radiation-adaptive shielding nuclear power distribution monitoring system is characterized in that the MXene nanosheets are prepared by the following process: obtaining Ti3AlC2Tx nanosheets by etching Ti3AlC2 powder with HF; dispersing the nanosheets in an epoxy resin matrix and adding a silane coupling agent to improve the dispersion uniformity; and molding and curing by vacuum casting process. This preparation process ensures the uniform distribution of MXene nanosheets in the matrix, improves shielding effectiveness and self-healing ability, while enhancing the mechanical properties and radiation aging resistance of the material, and extending the service life of the shield.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. By combining MXene self-healing shielding with a dynamic shielding device driven by SMA-hydraulic synergy, the limitations of traditional shielding systems in terms of "passive protection" have been overcome. The MXene self-healing shielding uses an electric field to drive nanosheets to autonomously migrate and repair cracks, restoring shielding effectiveness without manual intervention and solving the problem of traditional shielding requiring downtime for replacement when cracks appear. The SMA-hydraulic synergy device combines the rapid response of shape memory alloys with the stable thrust of a hydraulic system to achieve millisecond-level adjustment of shielding thickness, significantly improving the response speed to sudden radiation compared to a single drive method. The synergistic effect of these two technologies achieves, for the first time, a dual function of "autonomous repair and intelligent adjustment," significantly extending the service life of the shielding and shortening the response time.

[0014] 2. By constructing a digital twin central layer based on a multi-physics coupling algorithm, and integrating multi-dimensional data such as radiation, temperature, vibration, and cracks, a custom-defined lifetime prediction formula is used to accurately predict the remaining lifetime of the shielding body. This formula comprehensively considers the coupled effects of multiple damage factors, significantly improving prediction accuracy compared to traditional single-physics models. It can provide early warning of potential failure risks, offering a scientific basis for preventative maintenance and avoiding over-maintenance or insufficient protection due to inaccurate lifetime assessments.

[0015] 3. Through the collaboration of the multi-physics sensing layer and the closed-loop control layer, a complete closed loop of "perception-decision-execution-feedback" is formed. The multi-dimensional sensor deployment significantly improves data integrity and anomaly identification accuracy compared to traditional single-point monitoring; the closed-loop control mechanism, based on reinforcement learning, dynamically optimizes system parameters to achieve adaptive adjustment, significantly improving environmental adaptability and energy efficiency compared to traditional fixed-parameter systems, reducing ineffective energy consumption and manual intervention costs, and greatly improving economy and reliability while ensuring the safe operation of the nuclear power distribution system. Attached Figure Description

[0016] Figure 1 A schematic diagram of a nuclear power plant power distribution monitoring system with radiation adaptive shielding. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 This invention provides a technical solution: a nuclear power plant power distribution monitoring system with radiation adaptive shielding. We know that a complete adaptive shielding system needs to achieve closed-loop control of "perception-decision-execution-feedback". Therefore, the system is divided into four core layers: a multi-physics sensing layer, a digital twin central layer, a collaborative execution layer, and a closed-loop control layer. The multi-physics sensing layer is responsible for collecting radiation, temperature, vibration, and crack data of the shielding body and its surrounding environment; the digital twin central layer analyzes the data, constructs a virtual model, and predicts the state; the collaborative execution layer completes self-repair and shielding adjustment according to the central instructions; and the closed-loop control layer feeds back the execution results to the sensing layer to achieve dynamic optimization.

[0019] Each layer is connected via industrial Ethernet, and the data transmission protocol uses Modbus TCP / IP. After 72 hours of continuous testing, the total number of data packets transmitted via fiber optic cable was 3.24 × 10⁻⁶. 6 There were 28 lost data packets, resulting in a bit error rate of 8.6 × 10⁻⁶.-6 Far below the industrial-grade requirement of 10 -4 This ensures communication stability in environments with strong electromagnetic interference.

[0020] This embodiment uses a 10kV high-voltage switchgear in the No. 3 reactor building of a nuclear power plant as an example. During normal operation, the measured neutron flux in this area is 1.2 × 10⁻⁶. 8 -8.5×10 9 n / cm 2 The gamma-ray dose rate fluctuates between 52 and 198 μSv / h, with an average daytime temperature of 38°C and an average nighttime temperature of 35°C. The vibration acceleration caused by the proximity of the main pump unit remains stable at 0.12-0.48g. Based on the operation and maintenance records of the past 5 years, the neutron shielding efficiency of the shielding in this area decreases by about 2% annually due to radiation aging, and the crack width caused by vibration increases by an average of 0.03 mm annually. Under the traditional maintenance model, an average of 12 manual interventions are required annually, with a single intervention costing approximately 80,000 yuan.

[0021] I. Construction of the Multiphysics Sensing Layer Step 2: Sensor Selection and Layout Plan 1. Radiation sensor: Neutron detector: Selected The glass scintillation detector has a thermal neutron detection efficiency of ≥85%, an energy response range of 0.025 eV-1 MeV, and a 316L stainless steel casing that can withstand radiation doses up to 10. 5 Gy. Gamma-ray detector: Employs a CdZnTe semiconductor detector capable of distinguishing gamma rays of different energies. Arrangement: One set of neutron-gamma-ray composite detectors is installed on each of the four sides (front, rear, left, and right) of the switchgear shield, at a height of 1.5m above the ground. The installed detectors underwent continuous testing for one month, collecting a total of 8760 data sets, and the average value was calculated. Standard deviation Theoretical data shows that its measurement error is ≤±3%, which meets the system accuracy requirements.

[0022] 2. Temperature Field Monitoring Data Processing: Armored thermocouples (Type K) were selected, with a measurement range of -20-120℃ and an accuracy of ±0.5℃. The armor layer was made of Inconel 600, which is corrosion-resistant and radiation-resistant. Twenty thermocouples were arranged at 30cm intervals on the shielding surface, forming a temperature field monitoring grid. The temperature data collected by the 20 thermocouples ranged from 35.2-64.8℃, and the minimum value was calculated. maximum value The temperature value of 45℃ at a certain moment is normalized. The normalized result was calculated to be 0.33, which is consistent with the theoretical data.

[0023] 3. Vibration sensor: Select a piezoelectric MEMS sensor with a range of 0-10g and an accuracy of ±0.01g, and attach it to the center of the top of the shield. After Fourier transform, the vibration data will be analyzed to determine the dominant frequency. The vibration frequency was consistent with that of the main pump unit, and it was determined to be the main source of vibration.

[0024] 4. Crack Image Feature Extraction: A 2-megapixel industrial camera, equipped with an LED ring light source and an 8mm lens focal length, was installed to capture images of the shielding surface every 10 minutes. The weld seam images captured by the industrial camera were processed using the Canny operator to identify a crack of length... ,width The crack area was calculated. , accounting for the surface area of ​​the shielding body ( The proportion is The repair threshold has not yet been reached.

[0025] Data transmission reliability verification: After 72 hours of continuous testing, the total number of data packets transmitted via fiber optic cable was 3.24 × 10⁻⁶. 6 There were 28 lost data packets, resulting in a bit error rate of 8.6 × 10⁻⁶. -6 Far below the industrial-grade requirement of 10 -4 It meets the communication needs in environments with strong electromagnetic interference.

[0026] Traditional techniques rely on single-point monitoring, while the 4D data acquisition in this embodiment improves the accuracy of anomaly identification. According to theoretical data, the coordinated deployment of multiple sensors can reduce radiation field reconstruction errors.

[0027] II. Model Construction of the Central Layer of Digital Twin Step 3: 3D Model and Physical Parameter Modeling Point cloud data of the shielding structure was acquired through laser scanning. After modeling in SolidWorks, the key dimensions were determined: length 1200mm, width 800mm, height 1500mm, lead plate thickness 8.2mm (design value 8mm, error due to manufacturing process), and polyethylene layer thickness 20.1mm. Specifically, the lead plate density is 11.34g / cm³. 3 Thermal conductivity 35 W / (m·K), neutron macroscopic cross section 0.15 cm -1 Polyethylene: density 0.95 g / cm³ 3 Thermal conductivity 0.4 W / (m·K), hydrogen atom number density 5 × 10⁻⁶ 22 atoms / cm 3 MXene composite material: density 1.2 g / cm³ 3 Thermal conductivity 0.8 W / (m·K), shielding effectiveness ≥40 dB (100 kHz-1 GHz).

[0028] After inputting the data into COMSOL Multiphysics, the linear attenuation coefficient of the lead plate for 662 keV gamma rays was calculated to be 0.78 cm⁻¹. -1 The moderating length of polyethylene for thermal neutrons is 9.2 cm, which is consistent with the theoretical data (lead: 0.77 cm). -1 The deviation of ≤2% for polyethylene (9.0cm) verifies the accuracy of the model.

[0029] Step 4: Data Preprocessing and Feature Extraction Analysis of radiation data for a specific hour: mean dose rate peak The variance of the fluctuation was calculated. ; Temperature field gradient The maximum gradient was calculated to be 0.5℃ / cm (the temperature difference between the top and bottom of the switch cabinet) based on the temperature difference between adjacent sensors. After Fourier transform, the dominant frequency of the vibration data The frequency of the vibration was consistent with that of the main pump unit, and it was determined to be the main source of vibration. Feature extraction results of the crack image show that the crack widths measured in three consecutive measurements were 0.08 mm, 0.09 mm, and 0.09 mm, respectively, with an average width of 0.087 mm calculated. Repair has not yet been triggered.

[0030] Step 5: Construction and Training of the Life Prediction Model The model training used 8,000 sets of normal data and 2,000 sets of fault data, including 400 sets of radiation exceeding the standard, 600 sets of thermal aging, 500 sets of vibration fatigue, and 500 sets of crack failure cases.

[0031] We use LSTM (Long Short-Term Memory) network to process time-series data and combine it with CNN (Convolutional Neural Network) to extract image features, thus constructing an LSTM-CNN hybrid model.

[0032] Input layer: LSTM receives temporal features of radiation, temperature, and vibration (9 features in total). , , , , , , , ); The CNN receives crack images (256×256 pixels); Output layer: Remaining lifetime (Unit: year)

[0033] Based on Miner's fatigue accumulation theory and considering the effects of multi-field coupling, the life prediction formula is derived as follows: ; in: The theoretical lifespan under no-damage conditions is 15 years. (Cumulative radiation damage) ); (Cumulative temperature damage) ); (Vibration cumulative damage,) ); (Crack damage factor, (This refers to the surface area of ​​the shielding body).

[0034] Weighting coefficient calculation: The weighting coefficients in the lifespan formula were calculated using multiple linear regression analysis. (Radiation damage weight), calculated based on 1000 radiation-dominated failure cases; (Temperature damage weight), obtained by fitting 800 sets of thermal aging data; (Vibration damage weight), verified through 600 sets of vibration fatigue cases; (Crack damage weight) is determined based on 500 sets of crack failure data.

[0035] Theoretical data shows that the goodness of fit of this weight combination is... It is superior to the single-factor model. ).

[0036] Model training results: The model was trained using 8,000 sets of normal data and 2,000 sets of fault data (including 400 cases of radiation exceeding limits, 600 cases of thermal aging, 500 cases of vibration fatigue, and 500 cases of crack failure). After 500 epochs of training, the average prediction error on the validation set was 7.8%. The predicted lifespan of the new shield was 14.8 years, with a deviation of 1.3% from the theoretical data (15 years). For the shield that had been in operation for 5 years, the predicted remaining lifespan was 8.2 years, and the actual remaining lifespan, as determined by follow-up testing, was 8.0 years, with a calculated error of 2.5%, both of which meet the design requirement of ≤8%.

[0037] In existing technologies, life prediction of nuclear power equipment often uses a single physical model without considering the effects of multi-field coupling. However, the hybrid model in this embodiment combines multi-source data with a custom life formula, achieving for the first time a comprehensive life assessment of "radiation-heat-vibration-crack," improving prediction accuracy by more than 2 times. Theoretical data shows that the accuracy of early warning 90 days in advance can reach 90%.

[0038] III. Action Implementation in the Collaborative Execution Layer Step 6: Preparation and Repair Mechanism of MXene Self-Healing Shield Material preparation: MXene nanosheets: Ti3AlC2TxMXene was obtained by etching Ti3AlC2 powder (99% purity) with HF, with a sheet diameter of 500-800nm ​​and a thickness of 1-2nm; Composite material formulation: MXene (5wt%) + epoxy resin (80wt%) + polyamide curing agent (15wt%), with 0.5wt% silane coupling agent (KH550) added to improve dispersibility; Molding process: vacuum casting (vacuum degree -0.09MPa), curing at 60℃ for 4 hours to form a 10mm thick shielding layer, with built-in copper electrodes (5cm spacing).

[0039] When the digital twin predicts crack width At that time, a repair command is triggered.

[0040] Crack repair test: An artificially created 0.15mm wide crack was used to apply a 50V voltage, and the migration speed of MXene was observed. ; The calculation showed that the required migration distance to repair a 0.15mm crack was 0.3cm (simultaneous migration from both sides), the theoretical repair time was 2.5 minutes, and the actual test time was 3 minutes. The deviation was caused by the inhomogeneity of the material. After the repair is completed, the filling area is calculated using ultrasonic testing. Total crack area Repair efficiency It is 90% higher than the design requirements.

[0041] Step 7: Adjustment of the dynamic shielding drive device SMA spring: Material: Ni-Ti alloy (Ni content 55at%), phase transition temperature 60℃, free length 10cm, contraction rate 20%; Hydraulic push rod: stroke 0-10cm, thrust 500N, response time 0.5s; Shielding module: composed of lead-polyethylene composite plate (thickness 5-15cm adjustable), connected to the drive device via guide rail.

[0042] When the neutron flux increases to 6×10 10n / cm 2 / s (exceeding the threshold of 5×10) 10 n / cm 2 / s): When a 3A current is applied to the SMA spring, it contracts to 8cm within 2 seconds, causing the shielding module to move 2cm. The hydraulic push rods work synchronously, increasing the shielding thickness from 10cm to 15cm within 3 seconds. The neutron shielding efficiency at this point is calculated as follows: At a thickness of 10cm, the shielding efficiency is ; At a thickness of 15cm, ; The actual measured shielding rate was 98.8%, with a deviation of 0.18% from the theoretical data. The total response time of the adjustment process is 0.8 seconds, which meets the design requirements (≤1 second).

[0043] In the prior art, the repair of the shielding body relies on manual replacement, and dynamic adjustment mostly adopts hydraulic drive (response time > 1 second). However, the MXene self-healing technology in this embodiment realizes the autonomous repair of cracks, and the repair efficiency is higher than that of manual repair. The SMA-hydraulic collaborative drive shortens the response time to 0.1 seconds and reduces energy consumption by 40% compared with hydraulic drive.

[0044] IV. Feedback Optimization of Closed-Loop Control Layer Step 8: Feedback Data Acquisition and Error Correction After a certain shielding adjustment: The predicted neutron shielding rate was 98.98%, the actual measurement was 98.2%, and the calculation yielded... No corrections are needed; After another repair, the predicted repair efficiency was 92%, but the actual efficiency was 88%, resulting in a margin of error. It is still within the allowable range (≤10%). When the prediction error of a certain vibration reaches 12%, the LSTM weights are corrected: After correction, the error was reduced to 7%.

[0045] Step 9: Dynamic Optimization Strategy Based on one year of operational data: with a sampling interval of 10 seconds under normal operating conditions, the annual data volume is 3.15 × 10⁻⁶. 7 After the abnormal operating conditions (accounting for approximately 5%) were shortened to 1 second, the total data volume increased to 4.73 × 10⁻⁶. 7 The calculations show that energy consumption is reduced by 45% compared to 1-second sampling throughout the entire time period; The repair threshold of the new shield is 0.1 mm. After 3 years of operation, the threshold should be adjusted to 0.08 mm based on the crack propagation rate. In practical applications, this reduces the occurrence rate of severe cracks from 2 times / year to 0 times. The shielding adjustment strategy, enhanced by learning and training, improved the matching degree between thickness and radiation level from 60% to 91%, and theoretical data shows that it can reduce ineffective adjustment energy consumption by 30%.

[0046] This implementation method demonstrates that the lifespan prediction formula, derived from 100,000 sets of historical data, reduces the error from 25% in existing technologies to 7.8%, achieving a 92% accuracy rate for 90-day advance warning, a 53% improvement over traditional methods (60%), thus enhancing the precision of data-driven approaches. Through the synergistic effect of self-repair and dynamic adjustment, the theoretical repair efficiency of MXene reaches 90%, while the actual efficiency reaches 93.75%. Combined with the 0.8-second response speed of SMA-hydraulic drive, the radiation exceedance time is reduced from an average of 15 hours per year to 3 hours, and the annual radiation dose to personnel is reduced from 2.1 mSv to 0.5 mSv, exceeding the IAEA limit. Furthermore, the calculated annual maintenance cost is reduced from 800,000 yuan (traditional) to 320,000 yuan, with a reduction of 8 manual interventions per year (saving 640,000 yuan), a 45% reduction in energy consumption (saving 40,000 yuan), and an investment payback period of approximately 3 years, demonstrating significant engineering application value.

[0047] As we know from existing technologies, balancing the "safety-economy" of nuclear power shielding systems has always been a challenge for the industry. This embodiment verifies the feasibility of the technical solution through quantitative theoretical data. For the first time, it deeply integrates MXene self-healing, digital twin prediction, and dynamic shielding drive to form a complete adaptive closed loop, providing a scalable paradigm for the intelligent protection of nuclear power distribution systems. It has significant industrial value and provides a scalable paradigm for the intelligent protection of nuclear power distribution systems.

Claims

1. A radiation-adaptive shielded nuclear power plant power distribution monitoring system, characterized in that: include: A multi-physics sensing layer is used to collect data on radiation, temperature, vibration, and cracks in the shield and its surrounding environment. The digital twin central layer constructs a three-dimensional model of the shield and predicts its remaining lifetime based on a multiphysics coupling algorithm, which includes the following formula: ； in Theoretical lifespan, , , , These are radiation, temperature, vibration, and crack damage factors, respectively. , , , These are the weighting coefficients; The collaborative execution layer includes an MXene self-healing shield and an SMA-hydraulic collaboratively driven dynamic shielding device; The closed-loop control layer corrects the prediction model and optimizes system parameters based on the feedback from the execution results. The MXene self-healing shield achieves autonomous repair by driving MXene nanosheets to migrate to the crack region through an electric field, and the SMA-hydraulic synergistic drive device achieves rapid adjustment of the shield thickness through the synergistic effect of shape memory alloy and hydraulic system.

2. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 1, characterized in that: The multiphysics sensing layer includes: a neutron-gamma composite detector, a temperature sensor, a vibration sensor, and an industrial camera distributed on the surface of the shield; a data acquisition unit for receiving and preprocessing data from each sensor; and an optical fiber transmission module for transmitting the preprocessed data to the digital twin central layer.

3. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 1, characterized in that: The digital twin central layer includes: a three-dimensional physical model of the shield constructed based on laser scanning data; a multi-field coupling lifetime prediction module to predict the remaining lifetime; and a parameter correction module to dynamically adjust the prediction model parameters based on feedback from the closed-loop control layer.

4. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 1, characterized in that: The MXene self-healing shield comprises: MXene nanosheets uniformly dispersed in an epoxy resin matrix; an embedded copper electrode network for applying an electric field to drive the migration of the MXene nanosheets; and a repair trigger module that automatically initiates an electric field repair procedure when the crack width exceeds a threshold.

5. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 1, characterized in that: The SMA-hydraulic co-drive device includes: a Ni-Ti shape memory alloy spring that rapidly contracts and deforms when energized; a hydraulic push rod system that provides continuous and stable thrust; and a linkage mechanism that combines the rapid movement of the SMA spring with the stable thrust of the hydraulic push rod.

6. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 1, characterized in that: The closed-loop control layer includes: a feedback data acquisition module, which acquires the action effect data of the execution layer in real time; an error calculation module, which compares the predicted value with the actual value and calculates the deviation; and a parameter optimization module, which dynamically adjusts the sensor sampling frequency, repair threshold, and shielding adjustment strategy based on a reinforcement learning algorithm.

7. The nuclear power plant power distribution monitoring system with radiation adaptive shielding as described in claim 4, characterized in that: The radiation-adaptive shielding nuclear power distribution monitoring system is characterized in that the MXene nanosheets are prepared by the following process: Ti3AlC2Tx nanosheets are obtained by etching Ti3AlC2 powder with HF; the nanosheets are dispersed in an epoxy resin matrix, and a silane coupling agent is added to improve the dispersion uniformity; and the nanosheets are formed and cured by a vacuum casting process.

Images