RIA prediction model and construction method of optical fiber after radiation in high-energy ct environment

Abstract

The application belongs to the technical field of optical fiber communication and radiation effect cross, and particularly relates to a high-energy CT environment optical fiber radiation RIA prediction model and a construction method, comprising: obtaining radiation experiment data of a multi-layer composite structure optical fiber; determining the absorption dose rate distribution of each layer of the optical fiber; constructing a multi-defect group superimposed RIA dynamics prediction primary model according to the radiation experiment data and the absorption dose rate distribution; adopting a staged fitting method to optimize the RIA dynamics prediction primary model, and obtaining the RIA prediction model. A closed loop is formed from experiment acquisition, dose simulation, model construction to parameter optimization, the RIA prediction model can quantitatively estimate the optical fiber transmission performance degradation based on measurable process parameters and post-irradiation time. The RIA prediction model can be directly embedded into the process design flow of industrial CT detection, and provides a quantitative basis for detection parameter optimization, irradiation risk assessment and post-detection processing strategy making, and realizes effective transformation from mechanism research to engineering application.

Description

Technical Field

[0001] This invention belongs to the field of interdisciplinary technology of optical fiber communication and radiation effects, specifically relating to a prediction model and construction method for RIA after optical fiber radiation under high-energy CT environment. Background Technology

[0002] In fiber optic guidance systems, the guiding fiber serves as the crucial carrier for data transmission, and its structural integrity and transmission stability directly impact guidance accuracy and reliability. Industrial CT systems utilize high-energy X-ray beams for multi-angle projection scanning and reconstruction of the target's internal structure, offering irreplaceable advantages in the geometric reconstruction, defect identification, and quality assessment of the guiding fiber coil. However, high-energy X-rays themselves interact with the multilayer material system of the guiding fiber through photoelectric effects, Compton scattering, and ionization, generating various point defects such as electron-hole pairs, non-bridging oxygen pore centers (NBOHC), and self-trapped hole centers (STH). This increases the optical absorption cross-section, leading to the accumulation of radiation-induced attenuation (RIA). Excessively high RIA increases fiber transmission loss, severely affecting the data link stability of the guidance or fiber optic sensing system.

[0003] Current research on fiber optic irradiation damage mainly focuses on the damage response of communication fibers, erbium-doped fibers, and detection fibers under steady-state or low-dose-rate irradiation conditions. For single-layer fibers, several models based on color center generation and recombination dynamics exist, including second-order dynamics models, power-law defect generation models, and tensile exponential annealing models. However, these models are typically only applicable to the fiber core or homogeneous material systems and are not suitable for the damage behavior of multilayer composite fibers in high-dose-rate industrial CT environments.

[0004] On the one hand, the guiding fiber has a multilayered combination of functional materials, with significant differences in atomic composition, electron trapping cross section, and density between the layers. This results in strong inhomogeneity in X-ray energy deposition between the layers, with dose rates differing by several orders of magnitude. Traditional models cannot simultaneously handle the differences in the proportion of color centers generated due to this spatial inhomogeneity.

[0005] On the other hand, industrial CT scenarios typically employ high-energy radiation sources, with dose rates often exceeding 5 Gy / s, far higher than the typical dose rates in fiber optic communication irradiation experiments (usually <1 Gy / s). At such high dose rates, the defect generation rate exhibits nonlinear enhancement, and the composite competition between defects is intense. Traditional power-law models are insufficiently accurate in describing the multidimensional nonlinear evolution of RIA over time, dose rate, and cumulative dose, lacking integrated modeling capabilities for transient relaxation, saturation effects, and damage recovery dynamics, making it difficult to directly guide the optimization of industrial CT detection parameters and radiation safety design.

[0006] Therefore, it is urgent to establish a predictive model that can accurately describe the multidimensional evolution behavior of multilayer composite fiber RIA under high-energy CT detection environment, so as to realize reliable prediction of fiber performance degradation under different dose rates, cumulative doses and post-irradiation time conditions, and provide a theoretical basis for developing safe and efficient CT detection processes. Summary of the Invention

[0007] The purpose of this invention is to provide a predictive model and construction method for RIA after fiber optic radiation in high-energy CT environment, so as to solve the technical problem that existing technologies cannot directly guide the optimization of industrial CT detection parameters and radiation safety design.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for constructing a prediction model for post-fiber optic radiation-induced irradiation (RIA) under high-energy CT conditions, comprising: Obtain radiation experimental data of multilayer composite optical fibers; determine the absorbed dose rate distribution of each layer of the optical fiber; Based on radiation experimental data and absorbed dose rate distribution, a primary model for predicting the dynamics of RIAs with multiple defect families superimposed is constructed. A staged fitting method was used to optimize the primary model for RIA dynamics prediction, resulting in the RIA prediction model.

[0009] Preferably, obtaining radiation experimental data of multilayer composite optical fibers includes: A multi-layered composite structure guiding fiber was obtained as a sample fiber, and the length and initial optical power of each sample fiber were measured and recorded. The sample optical fibers were grouped according to a preset experimental matrix and placed under an industrial CT system for X-ray irradiation at a specified tube voltage and duration. After irradiation, the real-time output optical power of each sample optical fiber was rapidly measured and recorded at multiple preset discrete time points. Based on the initial optical power, real-time output optical power and sample fiber length, the RIA value of each sample fiber at each discrete time point is calculated to form a multidimensional experimental dataset containing dose rate, cumulative dose, time and RIA value. Outlier detection and processing were performed on the multidimensional experimental dataset. Outliers were identified using intragroup consistency test and time series sliding window analysis, and data correction was performed using a hierarchical imputation method to obtain radiation experimental data of multilayer composite optical fiber used for model construction.

[0010] Preferably, the RIA value of each sample fiber at each discrete time point is calculated using the following formula:

[0011] In the formula, Before irradiation at wavelength Initial optical power at the location, Time after irradiation at the same wavelength t Measured optical power, L The length of the optical fiber sample.

[0012] Preferably, the step of constructing a primary model for predicting the dynamics of RIAs with multiple defect families superimposed based on radiation experimental data and absorbed dose rate distribution includes: Based on radiation experimental data, the input variables are determined to be post-irradiation time, radiation dose rate, and cumulative radiation dose. Based on the input variables and RIA dynamics, a primary RIA dynamics prediction model is constructed based on three defect families: short lifetime, medium lifetime, and stable lifetime, as shown in the following equation:

[0013]

[0014]

[0015] In the formula, i = 1, 2, 3, representing the components of short-lifetime / medium-lifetime / stable defects, respectively, and t represents the time after irradiation. Indicates cumulative dose. Indicates dose rate, This represents the magnitude of the generation of the i-th type of defect under this condition. This indicates possible baseline drift or long-term irreversible damage. It is the time stretching index of various defects; This represents the dose saturation value for the color center defect family. For the first dose rate enhancement term, This is the second dose rate enhancement term; The stretch index is the first time scale. The stretching index is the second time scale. Indicates the intensity generated after radiation; This indicates the spontaneous annealing capability of the same defect family.

[0016] Preferably, the step of using a staged fitting method to optimize the primary model for RIA dynamics prediction to obtain the RIA prediction model includes: A subset of data that has reached a preset stability threshold after irradiation is selected from the experimental dataset. The primary model for RIA dynamic prediction is simplified to include only the contribution of the stable defect family. Using the data subset, the amplitude function parameters and constant terms of the stable defect family are obtained by fitting using the nonlinear least squares method. The amplitude function parameters and constant terms of the fitted stable defect family are fixed. The complete experimental dataset containing all time points is used as input to fit the amplitude function parameters of all remaining defect families and the time evolution function parameters of all defect families in the model to obtain a preliminary set of full parameter solutions. Centered on the initial full-parameter solution, multiple sets of different initial parameter value vectors are randomly generated within the preset range of each parameter value; for each set of initial parameter value vectors, a nonlinear least squares fitting algorithm is run independently to re-optimize all adjustable parameters of the model, and the results are recorded, as well as the sum of squared residuals after each fitting is completed. Compare the sum of squared residuals after each fitting, and select the set of parameters with the smallest fitting error as the optimal parameter set of the model; determine the RIA prediction model based on the optimal parameter set.

[0017] Preferably, determining the absorbed dose rate distribution of each layer of the optical fiber includes: Based on the actual material composition, density, and geometric parameters of each functional layer of the guiding fiber, an equivalent multi-layered cylindrical geometric model was established in Monte Carlo simulation software and placed in an air medium to simulate the actual irradiation environment. Using specialized energy spectrum calculation software, the X-ray emission energy spectrum of an industrial CT system after passing through a specific filter is simulated under several preset tube voltage conditions. The generated energy spectrum data is then normalized and used as the incident photon source input for subsequent Monte Carlo simulations. For each tube voltage, a complete physical model of photon-matter interaction is used to track a large number of photon histories in Monte Carlo simulation software, and the deposition energy data of X-rays in each functional layer of the optical fiber are recorded and statistically analyzed. Based on the definition of absorbed dose rate, the deposition energy data per unit mass of each layer obtained from Monte Carlo simulation are converted to calculate the average absorbed dose rate of the fiber core, cladding, coating and reinforcing layer under different tube voltage conditions, and the absorbed dose rate distribution of each layer of the fiber is obtained.

[0018] In a second aspect, the present invention provides a system for constructing a prediction model for post-fiber optic radiation-induced irradiation (RIA) under high-energy CT conditions, comprising: The data acquisition unit is used to acquire radiation experimental data of multilayer composite optical fibers and determine the absorbed dose rate distribution of each layer of the optical fiber. The model building unit is used to construct a primary model for predicting the dynamics of RIAs with multiple defect families based on radiation experimental data and absorbed dose rate distribution. The model optimization unit is used to optimize the primary model for RIA dynamics prediction using a staged fitting method, thereby obtaining the RIA prediction model.

[0019] In a third aspect, the present invention provides a prediction model for RIA after fiber optic radiation in a high-energy CT environment, which is constructed using the method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment described in any one of the above claims.

[0020] In a fourth aspect, the present invention provides an electronic device, including a processor and a memory, wherein the processor is configured to execute a computer program stored in the memory to implement a method for constructing a post-fiber optic radiation RIA prediction model under high-energy CT environment as described in any one of the preceding claims.

[0021] In a fifth aspect, the present invention provides a computer-readable storage medium storing at least one instruction that, when executed by a processor, implements the method for constructing a post-fiber optic radiation RIA prediction model under high-energy CT conditions as described in any one of the preceding claims.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: 1) This invention is the first to construct a high-precision prediction model for the complex irradiation environment of multilayer composite optical fibers, overcoming the limitations of traditional homogeneous models. Existing studies are mostly based on a single material or assume a homogeneous medium, which cannot accurately describe the physical reality of highly non-uniform radiation energy deposition in guided optical fibers during industrial CT inspection due to the different materials of the core, cladding, coating, and reinforcing layers. This invention is the first to accurately simulate the absorbed dose rate distribution of each functional layer under different tube voltages using the Monte Carlo method, quantifying this non-uniformity and introducing it as a key input into the model, thereby achieving a true physical characterization of the actual complex structure. 2) A dynamic model of "multi-defect family superposition" deeply coupled with physical mechanisms and experimental data was established, achieving a precise description of the multidimensional and nonlinear evolutionary behavior of RIA. Traditional models often struggle to simultaneously characterize the complex coupling relationships and nonlinear saturation characteristics of radiation-induced decay (RIA) with dose rate, cumulative dose, and post-irradiation time. This invention innovatively proposes a phenomenological framework based on the competitive evolution of short, medium, and long-lifetime defect families. This model not only couples the microscopic physical processes of defect generation, transformation, and annealing, but also stably solves numerous highly coupled parameters through a staged fitting and multi-starting-point global optimization strategy, thus successfully reproducing and predicting the macroscopic dynamic behaviors of RIA, such as relaxation rise, graded annealing, and dose rate saturation effects. This invention provides a complete solution from basic research to engineering applications, offering direct engineering guidance value. Existing research largely focuses on mechanistic exploration or performance testing under specific conditions, lacking systematic prediction and optimization tools for actual testing scenarios. This invention not only establishes a high-precision prediction model (with a prediction error of ≤0.02 dB / km for stable RIAs in independent validation), but also, based on the model and experimental data, explicitly proposes specific and quantifiable engineering guidelines, such as a safe operating window (e.g., dose rate ≤5.4 Gy / s, cumulative dose ≤40 kGy) and a post-detection resting recovery period (≥2 hours). This provides direct and reliable theoretical tools and practical evidence for optimizing process parameters, radiation safety design, and performance assurance in guided fiber optic cable CT inspection, achieving a closed loop from theoretical model to engineering implementation. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the multilayer composite structure of the guidance fiber according to an embodiment of the present invention; Figure 3 This is a schematic diagram of Fluka multilayer fiber modeling according to an embodiment of the present invention; Figure 4 The incident energy spectrum calculated by SpecCale in this embodiment of the invention is shown; where (a) is 160 kV; (b) is 190 kV; (c) is 220 kV; and (d) is 250 kV. Figure 5 The X-ray flux distribution diagrams under different X-ray source voltages in this embodiment of the invention are shown; where (a) is 160kV; (b) is 190kV; (c) is 220kV; and (d) is 250kV. Figure 6 This is a data processing flowchart of an embodiment of the present invention; Figure 7 This is a graph showing the average RIA over time at different dose rates according to an embodiment of the present invention. Figure 8 The graphs show the changes in stable RIA with irradiation time at different dose rates according to embodiments of the present invention; where (a) is 5.2 Gy / s; (b) is 5.4 Gy / s; (c) is 5.7 Gy / s; and (d) is 6.1 Gy / s. Figure 9 This is a system structure block diagram according to an embodiment of the present invention; Figure 10This is a structural block diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0025] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0026] See Figure 1 This application discloses a method for constructing a prediction model of RIA after fiber optic radiation in a high-energy CT environment, including: S1: Obtain radiation experimental data of multilayer composite optical fiber; determine the absorbed dose rate distribution of each layer of the optical fiber; S2: Based on radiation experimental data and absorbed dose rate distribution, a primary model for predicting the dynamics of RIAs with multiple defect families superimposed is constructed. S3: The primary model for RIA dynamic prediction is optimized by using a staged fitting method to obtain the RIA prediction model.

[0027] This invention, while acquiring radiation experimental data, independently determines the absorbed dose rate distribution of each functional layer through Monte Carlo simulation. For the first time, it quantitatively introduces the non-uniformity of energy deposition in multilayer structures into the modeling process, enabling subsequent model construction to be based on multi-dimensional input data that more closely reflects physical reality. Addressing the characteristics of high dose rates and complex defect evolution in high-energy CT environments, this invention employs a multi-defect family superposition approach to construct a primary model, treating the macroscopic RIA as the sum of contributions from defect families with different dynamic behaviors. This framework can adapt to the actual physical processes of nonlinearly enhanced defect generation rates and the coexistence and competition of multiple defects under high dose rates, solving the problem of traditional models' single structure and poor adaptability when describing the multi-factor coupled evolution of RIAs from a mechanistic perspective. This invention employs a phased fitting method, rationally dividing the parameter solution sequence by first determining some parameters and then gradually optimizing the overall model. This effectively reduces the complexity of parameter inversion, improves the stability and operability of the model solution, and makes the complex dynamic model feasible for engineering applications. The invention forms a closed loop from experimental acquisition, dose simulation, model construction to parameter optimization. The resulting RIA prediction model can quantitatively predict the degradation of fiber optic transmission performance based on measurable process parameters (tube voltage, irradiation duration) and post-irradiation time. This method can be directly embedded into the process design flow of industrial CT inspection, providing quantitative basis for inspection parameter optimization, irradiation risk assessment, and post-inspection processing strategy formulation, achieving an effective transformation from mechanistic research to engineering applications.

[0028] In some embodiments, obtaining radiation experimental data of multilayer composite optical fibers includes: A multi-layered composite structure guiding fiber was obtained as a sample fiber, and the length and initial optical power of each sample fiber were measured and recorded. The sample optical fibers were grouped according to a preset experimental matrix and placed under an industrial CT system for X-ray irradiation at a specified tube voltage and duration. After irradiation, the real-time output optical power of each sample optical fiber was rapidly measured and recorded at multiple preset discrete time points. Based on the initial optical power, real-time output optical power and sample fiber length, the RIA value of each sample fiber at each discrete time point is calculated to form a multidimensional experimental dataset containing dose rate, cumulative dose, time and RIA value. Outlier detection and processing were performed on the multidimensional experimental dataset. Outliers were identified using intragroup consistency test and time series sliding window analysis, and data correction was performed using a hierarchical imputation method to obtain radiation experimental data of multilayer composite optical fiber used for model construction.

[0029] In some embodiments, the RIA value of each sample fiber at each discrete time point is calculated using the following formula:

[0030] In the formula, Before irradiation at wavelength Initial optical power at the location, Time after irradiation at the same wavelength t Measured optical power, L The length of the optical fiber sample.

[0031] In some embodiments, constructing a primary model for predicting the dynamics of RIAs with multiple defect families superimposed based on radiation experimental data and absorbed dose rate distribution includes: Based on radiation experimental data, the input variables are determined to be post-irradiation time, radiation dose rate, and cumulative radiation dose. Based on the input variables and RIA dynamics, a primary RIA dynamics prediction model is constructed based on three defect families: short lifetime, medium lifetime, and stable lifetime, as shown in the following equation:

[0032]

[0033]

[0034] In the formula, i = 1, 2, 3, representing the components of short-lifetime / medium-lifetime / stable defects, respectively, and t represents the time after irradiation. Indicates cumulative dose. Indicates dose rate, This represents the magnitude of the generation of the i-th type of defect under this condition. This indicates possible baseline drift or long-term irreversible damage. It is the time stretching index of various defects; This represents the dose saturation value for the color center defect family. For the first dose rate enhancement term, This is the second dose rate enhancement term; The stretch index is the first time scale. The stretching index is the second time scale. Indicates the intensity generated after radiation; This represents the spontaneous annealing capability of the same defect family. Addressing the complex kinetic characteristics of nonlinearly enhanced defect generation rates under high dose rate irradiation, intense inter-defect composite competition, and post-irradiation relaxation rise and graded annealing, this invention innovatively decomposes the total RIA into the superimposed contributions of three defect families: short-lived, medium-lived, and stable-lived. This framework not only couples the microscopic mechanisms of defect generation, transformation, and annealing but also simultaneously characterizes the nonlinear saturation and evolution of RIA with respect to three core variables: dose rate, cumulative dose, and post-irradiation time. This solves the problem of insufficient accuracy in multidimensional prediction by traditional power-law models or single-exponential models.

[0035] In some embodiments, the step of optimizing the primary model for RIA dynamics prediction using a staged fitting method to obtain the RIA prediction model includes: A subset of data that has reached a preset stability threshold after irradiation is selected from the experimental dataset. The primary model for RIA dynamic prediction is simplified to include only the contribution of the stable defect family. Using the data subset, the amplitude function parameters and constant terms of the stable defect family are obtained by fitting using the nonlinear least squares method. The amplitude function parameters and constant terms of the fitted stable defect family are fixed. The complete experimental dataset containing all time points is used as input to fit the amplitude function parameters of all remaining defect families and the time evolution function parameters of all defect families in the model to obtain a preliminary set of full parameter solutions. Centered on the initial full-parameter solution, multiple sets of different initial parameter value vectors are randomly generated within the preset range of each parameter value; for each set of initial parameter value vectors, a nonlinear least squares fitting algorithm is run independently to re-optimize all adjustable parameters of the model, and the results are recorded, as well as the sum of squared residuals after each fitting is completed. By comparing the sum of squared residuals after each fitting, the set of parameters with the smallest fitting error is selected as the optimal parameter set for the model; the RIA prediction model is then determined based on the optimal parameter set. Addressing the technical challenges of numerous model parameters, strong coupling, and susceptibility to local optima, this invention introduces a staged fitting and multi-starting-point optimization strategy. This method first locks down stable defect family parameters based on stable stage data, and then determines all dynamic parameters through global optimization using multiple sets of random initial values. This significantly improves the robustness and accuracy of parameter solving, ensuring the model possesses good convergence and repeatability.

[0036] In some embodiments, determining the absorbed dose rate distribution of each layer of the optical fiber includes: Based on the actual material composition, density, and geometric parameters of each functional layer of the guiding fiber, an equivalent multi-layered cylindrical geometric model was established in Monte Carlo simulation software and placed in an air medium to simulate the actual irradiation environment. Using specialized energy spectrum calculation software, the X-ray emission energy spectrum of an industrial CT system after passing through a specific filter is simulated under several preset tube voltage conditions. The generated energy spectrum data is then normalized and used as the incident photon source input for subsequent Monte Carlo simulations. For each tube voltage, a complete physical model of photon-matter interaction is used to track a large number of photon histories in Monte Carlo simulation software, and the deposition energy data of X-rays in each functional layer of the optical fiber are recorded and statistically analyzed. Based on the definition of absorbed dose rate, the deposition energy data per unit mass of each layer obtained from Monte Carlo simulation are converted to calculate the average absorbed dose rate of the fiber core, cladding, coating and reinforcing layer under different tube voltage conditions, and the absorbed dose rate distribution of each layer of the fiber is obtained.

[0037] Existing technologies are mostly based on the assumption of single-layer or homogeneous materials, which cannot handle the highly non-uniform radiation energy deposition problem caused by differences in the materials of the core, cladding, coating, and reinforcing layers of guiding optical fibers. This invention accurately obtains the absorbed dose rate distribution of each functional layer through Monte Carlo simulation, quantifies this interlayer non-uniformity and introduces it into the model, fundamentally breaking through the limitations of traditional homogeneous models, and making the model input closer to the actual physical scenario of industrial CT inspection.

[0038] This application also discloses a prediction model for RIA after fiber optic radiation in a high-energy CT environment, characterized in that it is constructed using the method described in any one of the above-mentioned methods for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment.

[0039] Example 1 This invention addresses the shortcomings of existing technologies by providing a prediction model and construction method for radiation-induced attenuation (RIA) after fiber optic radiation in a high-energy CT environment. This model, driven by a fusion of physical mechanisms and experimental data, can accurately predict the multidimensional nonlinear evolution of radiation-induced attenuation of multilayer composite guided fibers after high-energy X-ray irradiation, considering dose rate, cumulative dose, and time.

[0040] The technical solution of the present invention includes the following steps: Step 1: Prepare experimental equipment and the optical fiber guiding the sample required for the experiment; Step 2: Perform pre-radiation tests on all sample optical fibers and record their precise length and initial transmission performance; Step 3: Design and complete the full-factor multilayer composite fiber irradiation experiment; Step 4: Use the Monte Carlo method to simulate and calculate the absorbed dose rate under different X-ray tube voltage conditions in this experimental scenario; Step 5: Experimental data processing and damage mechanism analysis; Step 6: Physical framework construction and variable selection for RIA prediction model after multilayer composite fiber optic radiation under high-energy CT detection environment. Step 7: Solving and optimizing the parameters of the RIA prediction model; Step 8: Verify model performance.

[0041] As a preferred option, the specific method for step 1 is as follows: Step 1.1: Select a guiding optical fiber of the specified model and specifications, composed of a multi-layered composite structure (typically including a germanium-doped fiber core, a silica cladding, an acrylate coating, and a polytetrafluoroethylene reinforcing layer), as the experimental sample, and prepare a sufficient quantity of samples with a length of 1 km. Its multi-layered composite structure is as follows: Figure 2 As shown; Step 1.2: Select a high-precision industrial CT system as the irradiation source. Confirm that its key parameters, such as the X-ray source tube voltage range, maximum tube power, and detector size, meet the experimental requirements. Step 1.3: Prepare a high-stability light source, a high-precision optical power meter, and standard jumpers for calibration. Simultaneously, prepare a high-precision optical time-domain reflectometer for accurate fiber length measurement. Step 1.4: Prepare temperature and humidity control equipment to ensure that the ambient temperature in the experimental area is stable at 20±2℃ and the relative humidity is controlled at 35%-40%. Prepare auxiliary tools such as non-metallic sample holders, lint-free cloths, alcohol, fiber optic cleavers, and fusion splicers. Step 1.5: Before the experiment begins, calibrate all measuring instruments (optical power meter, OTDR). Physically connect and functionally debug the industrial CT system, optical testing system, and sample fixing device to ensure the synchronization and stability of data acquisition; Step 1.6: Compile standard operating procedures that include equipment operation steps, sample processing procedures, data recording formats, and radiation safety protection measures, and train the laboratory personnel.

[0042] As a preferred option, the specific method for step 2 is as follows: Step 2.1: Use a lint-free cloth dampened with alcohol to carefully clean the end faces of both ends of each fiber sample to be tested, ensuring that there are no dust, oil or other contaminants, in order to reduce connection loss. Step 2.2: Measure the length of each fiber sample using a high-precision OTDR device. Analyze the OTDR curves to determine and record the precise length of each fiber with an accuracy better than ±0.06%. Step 2.3: Connect the two ends of the standard jumper to the stable light source and the high-precision optical power meter, respectively. Turn on the light source and wait for the optical power meter reading to stabilize sufficiently. Then, set this value as the reference value for the measurement system to complete the system calibration. Step 2.4: Measure and record the initial optical power: Replace the standard jumper cable on the fiber sample to be tested and connect it between the light source and the optical power meter. After the reading stabilizes, record the optical power value P1. Then, reverse the connections at both ends of the fiber and measure and record the optical power value P2 again. Step 2.5: Calculate and record initial transmission performance: Take the average of the two measurement results as the initial optical power P0 of the fiber, and record P0 as the reference value P0(λ) for subsequent calculation of radiation-induced attenuation. Repeat steps 2.1 to 2.4 for all sample fibers; Step 2.6: Organize and archive the data, including the fiber number, precise length, and initial optical power P0(λ) of all sample fibers. Based on the subsequent experimental design (different combinations of dose rates and irradiation durations), assign a unique experimental group number to each fiber.

[0043] As a preferred option, the specific method for step 3 is as follows: Step 3.1: Design a full factorial experimental matrix and determine the core variables and their levels. Using X-ray tube voltage and irradiation duration as two core factors, and controlling the radiation dose rate and cumulative dose respectively, design a 4x4 full factorial experiment. Step 3.2: Randomly assign the fiber optic samples tested in Step 2 to the 16 experimental condition combinations mentioned above, with 4 fibers assigned to each condition for repeated experiments. Regularly wind or lay each fiber on a non-metallic support (such as a cylinder) and place it at the center of the CT system stage. Step 3.3: For each experimental group, set the corresponding tube voltage, tube current, and preset irradiation duration in the industrial CT control software. Start the CT system and irradiate the sample group with X-rays. Ensure that geometric parameters such as the distance from the X-ray source to the sample (SOD) remain constant; Step 3.4: Immediately after each irradiation session, start timing. At the predetermined precise time points (5, 15, 30, 60, 120, 240, 480 minutes after irradiation), quickly connect the fiber under test to the optical power testing system; Step 3.5: At each monitoring time point, following the method in Step 2.4, measure and record the optical power P(λ,t) of the optical fiber at the current time t. After the measurement is completed, return the sample to its fixed position as soon as possible until the next monitoring time point. Step 3.6: For each sample at each time point, calculate the radiation-induced attenuation value at that moment using formula (1). Finally, obtain a dataset containing sample ID, tube voltage, irradiation duration, time point t, and RIA value.

[0044] (1) in Before irradiation at wavelength The optical power at that location, Time after irradiation at the same wavelength t Measured optical power, L The length of the optical fiber sample.

[0045] As a preferred option, the specific method for step 4 is as follows: Step 4.1: Based on the actual material composition, density, and geometric dimensions (diameter, thickness) of each layer (core, cladding, coating, and reinforcing layer) of the guiding fiber, establish an equivalent multilayer cylindrical model in the Monte Carlo simulation software (FLUKA), and set it to be placed in an air medium, such as... Figure 3 As shown; Step 4.2: Using the professional software SpekCalc, simulate the X-ray energy spectrum of an industrial CT system after passing through a specific filter (e.g., 1mm Al) at target tube voltages (160, 190, 220, 250 kV). Figure 4 As shown. The generated energy spectrum data, after being normalized, is used as the simulated incident photon source; Step 4.3: Select an electromagnetic physics model in the simulation software that includes the entire process of photoelectric effect, Compton scattering, and electron pair generation. Set a sufficiently large particle history number (e.g., 10^8) to ensure statistical accuracy, and define the energy deposition recording region; Step 4.4: Run the simulation program for the four energy spectral sources corresponding to different source voltages, track the photon transport process in the multilayer fiber structure, and record the energy deposition in each layer of material. The simulated X-ray flux distribution under different source voltages is as follows: Figure 5 As shown; Step 4.5: After the simulation is completed, extract the energy data per unit mass of material deposited in each layer output by the software. ; Step 4.6: According to the definition of absorbed dose rate = (dE dep The simulated unit mass deposition energy ( / dt·m) is converted into absorbed dose rate in Gy / s. The average absorbed dose rate of each functional layer of the optical fiber under different tube voltages is compiled and output to form a dose rate distribution table.

[0046] As a preferred option, the specific method for step 5 is as follows: Step 5.1: Process the raw RIA data using a multidimensional outlier detection method. The specific process is as follows: Figure 6 As shown. First, intragroup consistency analysis was performed (using an improved Grubbs test based on median and MAD). Then, a time series sliding window MAD detection was used to identify and label outlier data points. Step 5.2: For the marked outliers, a tiered imputation strategy is adopted. Normal values ​​from adjacent time points of the same sample are used for interpolation replacement first; if no adjacent normal values ​​are available, the average measurement value of other samples within the same experimental group at the same time point is used for replacement. Step 5.3: Plot the curves of mean RIA versus time at different dose rates (tube voltages), as follows: Figure 7 As shown. Analyze the common characteristics of the curves; Step 5.4: Take the average RIA over a sufficiently long period after irradiation (e.g., 2, 4, or 8 hours) as the "stable RIA". Plot the curves of stable RIA versus irradiation duration (i.e., cumulative dose) at different dose rates, as shown below. Figure 8 As shown, observe its monotonic, linear, or nonlinear saturation growth trend; Step 5.5: Plot the curve of stable RIA as a function of dose rate (tube voltage) under the same irradiation duration, and observe whether it shows monotonically increasing and nonlinear saturation characteristics. Step 5.6: Based on existing literature, link the above macroscopic laws with microscopic damage mechanisms.

[0047] As a preferred option, the specific method for step 6 is as follows: Step 6.1: Based on experimental analysis, the three core macroscopically measurable input variables of the model are determined to be: post-irradiation time t, radiation dose rate, etc. Cumulative radiation dose D; Step 6.2: To describe the complex RIA dynamics, a phenomenological model framework is proposed: the total RIA is considered as the superposition result of several defect families with different annealing dynamic characteristics. In this invention, three defect families are considered: short lifetime, medium lifetime, and stable lifetime. Step 6.3: Establish the overall equation of the model: (2) In the formula, i = 1, 2, 3, representing the components of short-lifetime / medium-lifetime / stable defects, respectively, and t represents the time after irradiation. Indicates cumulative dose. Indicates dose rate, This represents the magnitude of the generation of the i-th type of defect under this condition. This indicates possible baseline drift or long-term irreversible damage. It is the time stretching index of various defects; Step 6.4: Design the amplitude function The specific form of which simultaneously reflects the saturation effect of D and Enhancement effect: (3) in, This represents the dose saturation value for the color center defect family. , This is a dose rate enhancement term used to capture the nonlinear enhancement effect of color center defect formation efficiency at high dose rates. The formation and disappearance rates of color centers together determine the characteristics of RIA changes with time and dose; Step 6.5: Describe the relaxation of defect concentration using the generalized stretching index form: (4) In the formula, , The stretching exponent on a time scale is used to control defect dynamics. It represents the post-irradiation generation intensity, which indicates the dynamic intensity of a certain defect family continuing to grow and transform into a higher absorption cross section after irradiation stops, reflecting the transformation behavior of metastable defects into stable color centers. This indicates the spontaneous annealing capability of the same defect family, describing the thermal relaxation and recombination process of defects on the post-radiation timescale.

[0048] As a preferred option, the specific method for step 7 is as follows: Step 7.1, Phased Fitting Strategy – Phase 1: Fitting Stable Defect Parameters. Select all RIA data with t ≥ 120 min from the dataset (considered as the stable phase). Assuming that short-life defects have annealed at this point, the model simplifies to: (5) Using this subset of data, the magnitude function parameters α3,D of the stable defect family (i=3) are fitted using the nonlinear least squares method. s,3 ,k dr,3 ,k s,3 And the constant C; Step 7.2, Staged Fitting Strategy – Second Stage: Initial Fitting of All Dynamic Parameters. Fix the parameters obtained in the first stage. Using the complete experimental dataset (all time points), initialize and fit all remaining parameters in the model (including α values ​​from other defect families). i D s,i k dr,i k s,i , and τ of all races i β i A gen,i B heal,i Perform the first global fitting to obtain an initial set of solutions; Step 7.3: To avoid local optima, take the initial solution of the second stage mentioned above as the center and randomly generate multiple sets of 20 different initial parameter value vectors in the vicinity of its parameter values ​​(within the range of 0.4 to 1.6 times); Step 7.4: Using each set of random initial values, run the nonlinear least squares fitting algorithm independently to optimize all adjustable parameters of the model (except for the parameters fixed in the first stage), and record the sum of squared residuals after each set of fitting is completed. Step 7.5: Compare the sum of squared residuals of all multi-start point fitting results, and select the set of parameter solutions with the smallest residuals as the final optimal parameter set of the RIA prediction model; Step 7.6: Use the optimal parameter set to calculate the model's predicted values ​​on the training data and compare them with the experimental values. Calculate the overall coefficient of determination R² and other indicators to confirm the model's ability to describe the training data.

[0049] As a preferred option, the specific method for step 8 is as follows: Step 8.1: Using a guided fiber sample of the same specifications that was not involved in the above modeling process, conduct a new set of irradiation experiments covering part of the modeling interval and including extrapolation conditions. Strictly follow the methods in Steps 2 and 3 to obtain its independent RIA time series data; Step 8.2: Extract stable RIA data for t ≥ 120 min from the validation data. Compare the model-predicted "stable RIA vs. dose rate" curves (for different cumulative doses) with the validation data points to assess trend consistency; Step 8.3: For all stable RIA validation data points, calculate the root mean square error, mean absolute error, and coefficient of determination R² between the model predictions and the measured values ​​according to equations (6), (7), and (8) to conduct a quantitative evaluation: (6) (7) (8) In the formula, and These represent the experimental observations and model predictions, respectively. This is the average value of the experiment. Represents the total number of data points; RMSE and MAE measure the magnitude of absolute error; R² measures the relative explanatory power. Step 8.4: Evaluate the model's extrapolation robustness: Focus on experimental points in the validation data where the dose rate or cumulative dose exceeds the modeling training range. Analyze whether the prediction error at these extrapolation points increases significantly to determine the model's generalization ability; Step 8.5: Summarize all verification results and generate a model performance evaluation report.

[0050] Example 2 Step 1: First, a specific type of multilayer composite guided optical fiber was selected as the experimental object. This fiber consists of a germanium-doped silica core, a pure silica cladding, an acrylic coating, and a polytetrafluoroethylene (PTFE) outer reinforcing layer. A sufficient quantity of 1-kilometer-long samples was prepared for repeated experiments. The core irradiation equipment adopted the high-precision industrial CT system MetroVoxel-4000, whose X-ray source tube voltage can be continuously adjusted within the range of 20-250kV, with a maximum power of 350W, which can meet the requirements of high dose rate irradiation. The measurement system includes a high-stability laser source operating in the 1310nm communication window, a high-precision optical power meter, standard jumpers for system calibration, and an optical time-domain reflectometer with a measurement length accuracy better than ±0.06%. The entire experiment was conducted in a controlled laboratory environment. The temperature was stabilized at 20±2℃ and the humidity was controlled at 35%-40% using a temperature and humidity control system. Non-metallic supports were used to fix the fiber samples to avoid secondary scattering interference. All equipment must undergo rigorous calibration and system integration before the experiment, and standard safety procedures covering equipment operation, radiation protection, and data recording must be established. Step 2: First, carefully clean both ends of the optical fiber with a lint-free cloth to eliminate connection loss. Then, use a high-precision OTDR device to measure the fiber. By analyzing the reflection curve, determine and record the precise length L of each fiber with an accuracy better than ±0.6 meters (0.06%). Next, perform initial optical power measurement: connect the light source and the optical power meter using a standard jumper, and calibrate the system reference after the reading stabilizes; then connect the fiber to the system, record the reading after the power meter stabilizes, and repeat the measurement after swapping the two ends of the fiber. Take the average of the bidirectional measurements as the initial optical power value P0(λ) of the fiber at a specific wavelength (e.g., 1310 nm). This process ensures that random errors introduced by light source fluctuations or connection differences are minimized. Finally, archive all sample data, including number, precise length, and initial optical power, and assign a unique experimental group identifier to each fiber according to the subsequent full-factor experimental design matrix. Step 3: The experiment employed a full factorial design, with the core variables being the X-ray tube voltage (four levels: 160kV, 190kV, 220kV, and 250kV, corresponding to different initial dose rates) and irradiation duration (four levels: 1, 2, 3, and 4 hours, corresponding to different cumulative doses), constituting 16 experimental conditions in a 4×4 configuration. Pre-tested fiber samples were randomly assigned to each condition group, with each group containing multiple fibers to assess repeatability. The samples were regularly wound onto a 25cm diameter cylinder and placed at the center of the CT system turntable, ensuring its axis was perpendicular to the X-ray beam, and the source-sample distance was kept constant at 600mm. For each experimental group, the corresponding tube voltage, tube current (fixed at 0.35mA), and preset duration were set in the control software, and high-energy X-ray irradiation was initiated. Immediately after irradiation, timing was started, and at seven precise time points—5, 15, 30, 60, 120, 240, and 480 minutes—the sample was rapidly moved to the optical testing stage, and the optical power measurement procedure in step 2 was repeated to obtain the post-irradiation optical power P(λ,t). Finally, the radiation-induced attenuation value at that moment was calculated using formula (1). A dataset containing sample ID, tube voltage, irradiation duration, time point t, and RIA value was ultimately obtained.

[0051] (1) in Before irradiation at wavelength The optical power at that location, Time after irradiation at the same wavelength t Measured optical power, L The length of the fiber optic sample; Step 4: To accurately quantify the radiation energy absorbed by each layer of the optical fiber under actual irradiation, a Monte Carlo method was used for dose simulation. Based on the actual material composition, density, and geometry of each layer of the guiding fiber, an equivalent multilayer cylindrical geometric model was established in FLUKA 2021.2 software and placed in air to simulate the actual scattering environment. SpekCalc software was used to simulate the X-ray energy spectrum of an industrial CT system under four target tube voltages, after passing through a tungsten target and a 1mm aluminum filter. The normalized energy spectrum was used as the input photon source for the simulation. A complete electromagnetic transport physics model was enabled in FLUKA, and 10 [units of measurement missing] were tracked for each energy spectrum. 8 The photon history is analyzed to ensure statistical accuracy. After the simulation is completed, the deposition energy per unit mass of each material layer recorded by the software is extracted and calculated according to the absorbed dose rate formula. = (dE dep The average absorbed dose rate (dt·m) was calculated and converted. Finally, the average absorbed dose rate of the core, cladding, coating, and reinforcing layer under four different tube voltages was obtained, revealing the significant non-uniformity of energy deposition in the multilayer structure.

[0052] Step 5: After obtaining the raw experimental data, rigorous data cleaning is performed to improve quality. A multi-dimensional outlier detection strategy is adopted: on the one hand, within the same experimental group, a modified Grubbs criterion based on the median and median absolute deviation is used for intragroup consistency testing; on the other hand, the sliding window MAD method is used for detection of RIA time series data of a single optical fiber. For the identified outliers, a hierarchical imputation method is used, prioritizing the use of normal values ​​from adjacent time points of the same sample for imputation; if no normal values ​​are found, the average value of other samples in the same group at that time point is used as a substitute.

[0053] The processed data were used to analyze macroscopic patterns: the evolution of RIA over time showed a relaxation peak around 30 minutes after irradiation at all dose rates, followed by a gradual stabilization after about 2 hours. This is attributed to the competition between the continuous generation / conversion of defects after irradiation cessation and the thermal annealing of short-lived defects. Analysis of the stable RIA (average of data with t≥120min) revealed that it exhibited monotonically nonlinear growth with increasing cumulative dose and gradually saturated. Simultaneously, at the same irradiation duration, the stable RIA also showed a nonlinear growth and saturation trend with increasing dose rate. The microscopic physical mechanism of these macroscopic nonlinear behaviors can be attributed to the following: high-energy radiation induces point defects such as self-trapped holes and non-bridging oxygen vacancy centers in the silica network. The generation rate of these defects is limited by available lattice sites and forms a dynamic competition with the recombination process between defects. Step 6: Based on experimental phenomena and damage mechanisms, construct the physical framework of the prediction model. The model selects three directly measurable or calculable macroscopic physical quantities as core input variables: post-irradiation time t, radiation dose rate, etc. (Obtained from Monte Carlo simulation and tube voltage mapping), cumulative radiation dose D (calculated from dose rate and irradiation duration). The physical framework adopts a phenomenological model of "multiple defect families superposition," treating the macroscopic RIA as a superposition result contributed by several defect families with different generation and annealing kinetics. Specifically, the overall model expression is defined as: (2) In the formula, i = 1, 2, 3, representing the components of short-lifetime / medium-lifetime / stable defects, respectively, and t represents the time after irradiation. Indicates cumulative dose. Indicates dose rate, This represents the magnitude of the generation of the i-th type of defect under this condition. This indicates possible baseline drift or long-term irreversible damage. It is the time stretching index of various defects.

[0054] Amplitude function The method used to describe the generation amount of type i defects is designed to simultaneously reflect the saturation effect of the cumulative dose and the nonlinear enhancement effect of the dose rate: (3) in, This represents the dose saturation value for the color center defect family. , This is a dose rate enhancement term used to capture the nonlinear enhancement effect of color center defect generation efficiency at high dose rates. The generation and disappearance rates of color centers together determine the characteristics of RIA changes with time and dose.

[0055] Time evolution function This is used to describe the concentration relaxation of this type of defect after irradiation, employing a generalized stretching index to characterize the possible post-irradiation transformation and thermal annealing processes of the defect: (4) In the formula, , The stretching exponent on a time scale is used to control defect dynamics. It represents the post-irradiation generation intensity, which indicates the dynamic intensity of a certain defect family continuing to grow and transform into a higher absorption cross section after irradiation stops, reflecting the transformation behavior of metastable defects into stable color centers. This indicates the spontaneous annealing capability of the same defect family, describing the thermal relaxation and recombination process of defects on the post-radiation timescale; Step 7: To address the issues of numerous model parameters and strong coupling, a phased fitting and multi-starting-point global optimization strategy is adopted. In the first phase, fitting is performed on stable defect parameters. All RIA data with t ≥ 120 min are selected from the dataset. At this point, it can be assumed that short-life defects have been largely annealed, and the model simplifies to: (5) Using this subset of data, the magnitude function parameters α3,D of the stable defect family (i=3) are fitted using the nonlinear least squares method. s,3 ,k dr,3 ,k s,3 And the constant C.

[0056] The second stage involves optimizing all dynamic parameters. The parameters obtained in the first stage are fixed, and a complete spatiotemporal four-dimensional dataset (containing all time points) is used as input to optimize all remaining parameters in the model (including the magnitude parameters α of other defect families). i D s,i k dr,i k s,i The time evolution parameter τ of all defect families i β i A gen,i B heal,iA global fit is performed to obtain an initial solution. To avoid local optima, multiple different initial parameter vectors are randomly generated within a range of 0.4 to 1.6 times each parameter value, centered on this initial solution. For each initial value, a nonlinear least squares fitting algorithm is run independently. Finally, the sum of squared residuals of all fitting results is compared, and the set of parameters with the smallest residuals is selected as the final optimal parameter set of the model, thereby ensuring that the model reaches a global or near-global optimum. Step 8: To objectively evaluate the predictive power and generalization ability of the established model, a completely independent validation set was used for testing. The validation set consisted of guided optical fibers of the same specification that had not participated in the aforementioned modeling process. A series of new irradiation experiments were conducted on it, covering part of the modeling interval and including extrapolation conditions (such as higher voltage, shorter or longer time). First, the model's predictive ability for the "relationship between stable RIA and dose rate" was verified: stable RIA values ​​for t ≥ 120 min were extracted from the validation data, and these were graphically compared with the "RIA-dose rate" curves predicted by the model at different cumulative dose levels. The coefficient of determination R², root mean square error, and mean absolute error between the predicted and measured values ​​were calculated. Second, the model's predictive ability for the "relationship between stable RIA and cumulative dose" was verified: the validation data were grouped by dose rate, and the model's predicted curves of stable RIA changing with cumulative dose within each group were compared with the experimental data points. The corresponding error indices were calculated. In addition, the model's predictive performance at the extrapolation points of the validation set was analyzed in detail to evaluate its robustness. The comprehensive validation results show that the model's prediction accuracy (RMSE) for stable RIAs can reach within 0.02 dB / km, and no systematic bias is observed under extrapolation conditions, proving that the model has good predictive ability and certain engineering practical value.

[0057] Example 3 like Figure 9 As shown, based on the same inventive concept as the above embodiments, the present invention also provides a system for constructing a prediction model for post-fiber optic radiation RIA under high-energy CT conditions, characterized in that it includes: The data acquisition unit is used to acquire radiation experimental data of multilayer composite optical fibers and determine the absorbed dose rate distribution of each layer of the optical fiber. The model building unit is used to construct a primary model for predicting the dynamics of RIAs with multiple defect families based on radiation experimental data and absorbed dose rate distribution. The model optimization unit is used to optimize the primary model for RIA dynamics prediction using a staged fitting method, thereby obtaining the RIA prediction model.

[0058] Example 4 like Figure 10 As shown, the present invention also provides an electronic device 100 for implementing a method for constructing a prediction model of RIA after fiber optic radiation in a high-energy CT environment; The electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on at least one processor 102, and at least one communication bus 104.

[0059] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the method for constructing a prediction model of RIA after fiber optic radiation in high-energy CT environment by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0060] The memory 101 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created based on the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0061] At least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 102 may be a microprocessor or any conventional processor. Processor 102 is the control center of electronic device 100, connecting various parts of electronic device 100 via various interfaces and lines.

[0062] The memory 101 in the electronic device 100 stores multiple instructions to implement a method for constructing a post-fiber radiation RIA prediction model under high-energy CT conditions, and the processor 102 can execute multiple instructions to achieve the following: S1: Obtain radiation experimental data of multilayer composite optical fiber; determine the absorbed dose rate distribution of each layer of the optical fiber; S2: Based on radiation experimental data and absorbed dose rate distribution, a primary model for predicting the dynamics of RIAs with multiple defect families superimposed is constructed. S3: The primary model for RIA dynamic prediction is optimized by using a staged fitting method to obtain the RIA prediction model.

[0063] Example 5 If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, and read-only memory (ROM).

[0064] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0066] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0067] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0068] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment, characterized in that, include: To obtain radiation experimental data of multilayer composite optical fibers; Determine the absorbed dose rate distribution of each layer of the optical fiber; Based on radiation experimental data and absorbed dose rate distribution, a primary model for predicting the dynamics of RIAs with multiple defect families superimposed is constructed. A staged fitting method was used to optimize the primary model for RIA dynamics prediction, resulting in the RIA prediction model.

2. The method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment according to claim 1, characterized in that, Obtain radiation experimental data for multilayer composite optical fibers, including: A multi-layered composite structure guiding fiber was obtained as a sample fiber, and the length and initial optical power of each sample fiber were measured and recorded. The sample optical fibers were grouped according to a preset experimental matrix and placed under an industrial CT system for X-ray irradiation at a specified tube voltage and duration. After irradiation, the real-time output optical power of each sample optical fiber was rapidly measured and recorded at multiple preset discrete time points. Based on the initial optical power, real-time output optical power and sample fiber length, the RIA value of each sample fiber at each discrete time point is calculated to form a multidimensional experimental dataset containing dose rate, cumulative dose, time and RIA value. Outlier detection and processing were performed on the multidimensional experimental dataset. Outliers were identified using intragroup consistency test and time series sliding window analysis, and data correction was performed using a hierarchical imputation method to obtain radiation experimental data of multilayer composite optical fiber used for model construction.

3. The method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment according to claim 2, characterized in that, The RIA value of each sample fiber at each discrete time point is calculated using the following formula: In the formula, Before irradiation at wavelength Initial optical power at the location, Time after irradiation at the same wavelength t Measured optical power, L The length of the optical fiber sample.

4. The method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment according to claim 1, characterized in that, The preliminary model for predicting the dynamics of RIA (Radiation Invasive Affected Areas) based on radiation experimental data and absorbed dose rate distribution is constructed, including: Based on radiation experimental data, the input variables are determined to be post-irradiation time, radiation dose rate, and cumulative radiation dose. Based on the input variables and RIA dynamics, a primary RIA dynamics prediction model is constructed based on three defect families: short lifetime, medium lifetime, and stable lifetime, as shown in the following equation: In the formula, i = 1, 2, 3, representing the components of short-lifetime / medium-lifetime / stable defects, respectively, and t represents the time after irradiation. Indicates cumulative dose. Indicates dose rate, This represents the magnitude of the generation of the i-th type of defect under this condition. This indicates possible baseline drift or long-term irreversible damage. It is the time stretching index of various defects; This represents the dose saturation value for the color center defect family. For the first dose rate enhancement term, This is the second dose rate enhancement term; The stretch index is the first time scale. The stretching index is the second time scale. Indicates the intensity generated after radiation; This indicates the spontaneous annealing capability of the same defect family.

5. The method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment according to claim 1, characterized in that, The step-by-step fitting method is used to optimize the primary model for RIA dynamics prediction, resulting in the RIA prediction model, which includes: A subset of data that reaches a preset stability threshold after irradiation is selected from the experimental dataset. The primary model for RIA dynamic prediction is simplified to include only the contribution of the stable defect family. Using the data subset, the amplitude function parameters and constant terms of the stable defect family are obtained by fitting using the nonlinear least squares method. The amplitude function parameters and constant terms of the fitted stable defect family are fixed. The complete experimental dataset containing all time points is used as input to fit the amplitude function parameters of all remaining defect families and the time evolution function parameters of all defect families in the model to obtain a preliminary set of full parameter solutions. Centered on the initial full-parameter solution, multiple sets of different initial parameter value vectors are randomly generated within the preset range of each parameter value; for each set of initial parameter value vectors, a nonlinear least squares fitting algorithm is run independently to re-optimize all adjustable parameters of the model, and the results are recorded, as well as the sum of squared residuals after each fitting is completed. Compare the sum of squared residuals after each fitting, and select the set of parameters with the smallest fitting error as the optimal parameter set of the model; determine the RIA prediction model based on the optimal parameter set.

6. The method for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment according to claim 1, characterized in that, Determining the absorbed dose rate distribution of each layer of the optical fiber includes: Based on the actual material composition, density, and geometric parameters of each functional layer of the guiding fiber, an equivalent multi-layered cylindrical geometric model was established in Monte Carlo simulation software and placed in air to simulate the actual irradiation environment. Using specialized energy spectrum calculation software, the X-ray emission energy spectrum of an industrial CT system after passing through a specific filter is simulated under several preset tube voltage conditions. The generated energy spectrum data is then normalized and used as the incident photon source input for subsequent Monte Carlo simulations. For each tube voltage, a complete physical model of photon-matter interaction is used to track a large number of photon histories in Monte Carlo simulation software, and the deposition energy data of X-rays in each functional layer of the optical fiber are recorded and statistically analyzed. Based on the definition of absorbed dose rate, the deposition energy data per unit mass of each layer obtained from Monte Carlo simulation are converted to calculate the average absorbed dose rate of the fiber core, cladding, coating and reinforcing layer under different tube voltage conditions, and the absorbed dose rate distribution of each layer of the fiber is obtained.

7. A system for constructing a prediction model for RIA after fiber optic radiation in a high-energy CT environment, characterized in that, include: The data acquisition unit is used to acquire radiation experimental data of multilayer composite optical fibers; Determine the absorbed dose rate distribution of each layer of the optical fiber; The model building unit is used to construct a primary model for predicting the dynamics of RIAs with multiple defect families based on radiation experimental data and absorbed dose rate distribution. The model optimization unit is used to optimize the primary model for RIA dynamics prediction using a staged fitting method, thereby obtaining the RIA prediction model.

8. A prediction model for RIA after fiber optic radiation under high-energy CT conditions, characterized in that, It was constructed using the method for constructing a post-fiber radiation RIA prediction model under high-energy CT environment as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, It includes a processor and a memory, the processor being used to execute a computer program stored in the memory to implement a method for constructing a post-fiber optic radiation RIA prediction model under any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which, when executed by a processor, implements a method for constructing a post-fiber optic radiation RIA prediction model under high-energy CT conditions as described in any one of claims 1 to 6.

Images