Radiation shielding structure design system and method based on MCNP

Abstract

The invention specifically discloses a radiation shielding structure design system and method based on MCNP. The system comprises a model establishing module used for establishing an optimized mathematical model containing a shielding performance target and engineering constraints; the geometric modeling module is used for constructing a parameterized geometric model of the shielding body to be optimized, and the key size of the parameterized geometric model is defined by a group of adjustable design variables; the input file generation module is connected with the geometric modeling module; the analog calculation module is connected with the input file generation module; the optimization control module is connected with the geometric modeling module and the analog calculation module, and the optimization control module is used for adopting an optimization algorithm and circulating the following steps until an optimal shielding scheme meeting the optimization mathematical model is obtained: controlling the input file generation module to generate an analog calculation input file according to a current design variable, and the analog calculation module is controlled to execute analog calculation to obtain a shielding performance result, and the design variable is adjusted according to the shielding performance result.

Description

Technical Field

[0001] This invention belongs to the field of nuclear radiation protection technology, and specifically relates to a radiation shielding structure design system and method based on MCNP. Background Technology

[0002] In fields such as nuclear power plants, nuclear medicine, industrial flaw detection, and nuclear technology research, effective radiation shielding is crucial for ensuring the safety of personnel and the environment. Its fundamental purpose is to attenuate the radiation field intensity to below specified limits through appropriate material and structural design, thereby ensuring that occupational exposure of workers does not exceed dose constraints and meeting the protection requirements of the general public.

[0003] Currently, traditional radiation shielding design mainly relies on empirical formulas (such as semi-empirical formulas and accumulation factor methods) and manual calculations. This method is effective for simple geometries and homogeneous media, but it has significant limitations: 1. Insufficient accuracy: Empirical formulas have large errors when dealing with complex geometries, non-uniform materials, and neutron-photon coupling fields, which may lead to overly conservative designs (increasing cost and weight) or insufficient designs (posing safety hazards).

[0004] 2. Inefficiency: The design process relies heavily on the experience of engineers. Each modification to the scheme requires tedious manual calculations, making it difficult to perform systematic optimization with multiple parameters and objectives.

[0005] 3. Difficulty in handling complex problems: Traditional methods are almost unable to accurately assess the shielding effect of shielding bodies with complex structures such as holes, steps, and curved pipes.

[0006] Therefore, providing an automated, refined, and optimized method for designing radiation shielding structures to improve the design efficiency and shielding performance of radiation shielding structures is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] In view of the above-mentioned shortcomings of the existing technology, the purpose of this invention is to provide a radiation shielding structure design system and method based on MCNP. The radiation shielding design structure of this invention has high design accuracy and high degree of automation, which can effectively improve the design efficiency and shielding performance of radiation shielding structures.

[0008] The technical solution of this invention is implemented as follows: A radiation shielding structure design system based on MCNP includes a model building module, a geometric modeling module, an input file generation module, a simulation calculation module, and an optimization control module.

[0009] The model building module is used to build an optimization mathematical model that includes shielding performance targets and engineering constraints.

[0010] The geometric modeling module is used to construct a parametric geometric model of the shield to be optimized. The key dimensions of the parametric geometric model are defined by a set of adjustable design variables.

[0011] The input file generation module is connected to the geometric modeling module; the simulation calculation module is connected to the input file generation module.

[0012] The optimization control module is connected to the geometric modeling module and the simulation calculation module. The optimization control module is used to employ an optimization algorithm and repeat the following steps until the optimal shielding scheme that satisfies the optimization mathematical model is obtained: controlling the input file generation module to generate the input file for simulation calculation based on the current design variables, controlling the simulation calculation module to perform simulation calculation to obtain the shielding performance result, and adjusting the design variables based on the shielding performance result.

[0013] Furthermore, it also includes a result verification module, which is connected to the simulation calculation module and the optimization control module. The result verification module is used to perform uncertainty analysis on the shielding performance results. When the uncertainty is greater than a preset threshold, the simulation calculation module is controlled to increase the number of simulated particles and re-execute the simulation calculation.

[0014] Furthermore, the optimization control module employs one of the following optimization algorithms: parameter scanning method, genetic algorithm, or particle swarm optimization algorithm.

[0015] Furthermore, the shielding performance target includes a dose equivalent rate limit at a specified point outside the shield; the engineering constraints include one or more of the following: total weight of the shield, maximum package size, and cost.

[0016] Furthermore, the design variables include one or more of the following: the thickness, angle, and radius of curvature of the shielding layer.

[0017] Furthermore, the input file generation module includes an input file template and a script program; the input file template contains placeholders associated with design variables; the script program is used to fill the placeholders with the values ​​of the current design variables to generate the input file for simulation calculation.

[0018] Furthermore, the input file generation module is also used to set an optimized description for repeating structures or gate elements in the input file.

[0019] Furthermore, the simulation calculation module executes the simulation calculation by calling the Monte Carlo particle transport program through a script; and after the simulation calculation is completed, it automatically parses and extracts the shielding performance results from the output file.

[0020] Furthermore, the optimal shielding scheme is a numerical combination of design variables that satisfy the optimization mathematical model.

[0021] This invention also provides a radiation shielding structure design method based on MCNP, specifically including the following steps: S1: Establish an optimization mathematical model that includes shielding performance targets and engineering constraints; S2: Construct a parametric geometric model of the shield to be optimized, wherein the key dimensions of the parametric geometric model are defined by a set of adjustable design variables; S3: Generate the MCNP input file from the current design variables; S4: Perform MCNP calculations to obtain shielding performance results; S5: Perform uncertainty analysis on the shielding performance results. If the uncertainty is less than or equal to the preset threshold, output the shielding performance results; if the uncertainty is greater than the preset threshold, increase the number of simulated particles and proceed to step S4. S6: Use an optimization algorithm to evaluate the shielding performance results. If the convergence condition is met, output the current design variable as the optimal shielding scheme; if the convergence condition is not met, proceed to step S7. S7: Adjust the design variables based on the shielding performance results, and repeat steps S3-S6.

[0022] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention establishes an optimization mathematical model that includes masked performance objectives and engineering constraints through a model building module, providing clear guidance for subsequent optimization processes. The geometric modeling module constructs a parametric geometric model whose key dimensions are defined by adjustable design variables, enabling a systematic exploration of the design space and overcoming the limitations of traditional methods that can only obtain local optima. The connection between the input file generation module and the simulation calculation module automates the simulation calculation, avoiding the tedious process of manually adjusting parameters and repeated calculations.

[0023] 2. The optimization control module of this invention employs an optimization algorithm to iteratively execute steps such as generating input files, performing simulation calculations, obtaining shielding performance results, and adjusting design variables until the optimal shielding scheme that satisfies the optimization mathematical model is obtained. This iterative optimization mechanism enables the system to efficiently find the global optimal solution in a broad design space, thereby minimizing the weight, volume, or cost of the shielding body while meeting safety standards.

[0024] 3. Compared to traditional methods, this invention can improve design efficiency, shorten the design cycle, and reduce human error. Through an automated and refined design process, it can ensure the repeatability and comparability of design solutions, thereby providing a reliable, efficient, and optimized solution for the shielding structure design of critical facilities such as nuclear reactors, radioactive material transport containers, or medical radiation equipment. Attached Figure Description

[0025] Figure 1 This is a flowchart of the design method described in this invention. Detailed Implementation

[0026] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0027] In the fields of nuclear technology applications and radiation protection, the design of radiation shielding structures is a crucial step. Traditional radiation shielding design methods often rely heavily on empirical formulas and manual calculations. This not only leads to long design cycles and low efficiency but also makes it difficult to guarantee calculation accuracy. Especially when dealing with complex structures, the optimization process is time-consuming and labor-intensive, and it is difficult to systematically find the optimal solution. For example, suppose we need to design a container for transporting radioactive materials, whose shielding layer must meet strict dose equivalent rate limits, while also considering engineering constraints such as the container's total weight, maximum external dimensions, and cost. In this situation, if traditional manual design methods are still used, engineers need to repeatedly try different combinations of shielding materials, thicknesses, and geometries, and perform numerous simulations to evaluate the shielding performance of each scheme. This process is not only inefficient but also makes it difficult to guarantee finding an optimal shielding scheme under all constraints. This may result in a container that is too heavy, too large, or too costly, or even fails to meet safety standards. Failure to address these problems will severely restrict the further development and application of nuclear technology in fields such as medicine, energy, and industry, increasing the cost and risk of radiation protection.

[0028] Based on this, this invention proposes a radiation shielding structure design system based on MCNP, aiming to overcome the limitations of traditional design methods and achieve automation, refinement, and optimization in radiation shielding structure design. The system includes a model building module, a geometric modeling module, an input file generation module, a simulation calculation module, and an optimization control module.

[0029] The model building module is used to build an optimization mathematical model that includes shielding performance targets and engineering constraints.

[0030] The geometric modeling module is used to construct a parametric geometric model of the shield to be optimized. The key dimensions of the parametric geometric model are defined by a set of adjustable design variables.

[0031] The input file generation module is connected to the geometric modeling module; the simulation calculation module is connected to the input file generation module.

[0032] The optimization control module is connected to the geometric modeling module and the simulation calculation module. The optimization control module is used to employ an optimization algorithm and repeat the following steps until the optimal shielding scheme that satisfies the optimization mathematical model is obtained: controlling the input file generation module to generate the input file for simulation calculation based on the current design variables, controlling the simulation calculation module to perform simulation calculation to obtain the shielding performance result, and adjusting the design variables based on the shielding performance result.

[0033] This invention integrates a model building module, a geometric modeling module, an input file generation module, a simulation calculation module, and an optimization control module. It can systematically find the lightest, smallest, or lowest-cost shielding solution while meeting safety standards, significantly improving design efficiency and shielding performance.

[0034] To better understand the radiation shielding structure design system described in this invention, the key terms and implementation environment involved will be explained in detail below. This system is mainly applied in the fields of nuclear technology and radiation protection. Its core lies in combining the Monte Carlo particle transport program (MCNP) with optimization algorithms to achieve automated design of radiation shielding structures.

[0035] Among them, shielding performance targets refer to the safety standards that need to be achieved in radiation protection design, such as the dose equivalent rate limit at a specified point outside the shield.

[0036] Engineering constraints refer to the actual limitations that need to be met during the design process, such as the total weight of the shield, the maximum outer size, and the cost.

[0037] Design variables refer to parameters that can be adjusted during the optimization process, such as the thickness, angle, and radius of curvature of the shielding layer.

[0038] Optimization algorithms are mathematical methods used to search for optimal solutions in a design variable space, such as parametric scanning, genetic algorithms, or particle swarm optimization.

[0039] Simulation calculations typically refer to using Monte Carlo particle transport programs (such as MCNP) to simulate the radiation field in order to obtain shielding performance results.

[0040] This system can be deployed in various computing environments, such as personal workstations, server clusters, or cloud computing platforms, to meet the shielding design requirements of different scales and complexities.

[0041] The core of the radiation shielding structure design system proposed in this invention lies in achieving automation, refinement, and optimization of the radiation shielding structure through modular design and iterative optimization processes.

[0042] Specifically, the system includes a model building module primarily used to establish an optimization mathematical model that incorporates shielding performance objectives and engineering constraints. This model forms the basis of the entire optimization process, expressing design objectives (such as minimizing the dose rate) and constraints (such as weight, size, and cost) mathematically. For example, when designing a neutron radiation source transport container, the shielding performance objective can be set as a dose equivalent rate limit at 30 cm from the outer surface of the container, such as less than 25 µSv / h. Simultaneously, engineering constraints may include a total shield weight not exceeding 100 kg and a maximum container shell radius not exceeding 30 cm. These objectives and constraints are integrated into a single mathematical model to guide the subsequent optimization process.

[0043] Furthermore, the system also includes a geometric modeling module for constructing a parametric geometric model of the shield to be optimized. The key dimensions of this parametric geometric model are defined by a set of adjustable design variables. For example, for the aforementioned neutron radiation source transport container, its shielding layer could consist of an inner polyethylene moderator layer and an outer lead shielding layer. In this case, the thickness d of the polyethylene moderator layer... poly and the thickness d of the lead shielding layer pb These can be defined as design variables. Parametric modeling facilitates the adjustment of the values ​​of these design variables, thereby exploring different shielding structure configurations. As a preferred implementation, the geometric modeling module can use 3D modeling software (such as CAD software) or scripting programs (such as Python scripts) to create and modify geometric models to achieve automated modeling.

[0044] The input file generation module here is connected to the geometric modeling module. Its function is to generate the input files required for simulation calculations based on the values ​​of the current design variables. These input files are typically input files for Monte Carlo particle transport programs (such as MCNP), and contain information such as the geometry of the shield, material composition, radiation source terms, and the dose tally that needs to be recorded. For example, in the neutron radiation source transport container example above, the input file generation module will generate the input files based on the currently set d... poly and d pb The value automatically generates an MCNP input file that accurately describes the spherical shell model with a specific polyethylene and lead layer thickness.

[0045] The simulation module connects to the input file generation module and is used to perform simulation calculations to obtain shielding performance results. This module typically executes calculations by calling a Monte Carlo particle transport program (such as MCNP) via a script. The MCNP program can accurately simulate particle transport processes in complex geometries and materials, thus obtaining an accurate radiation dose rate distribution. In the example above, the simulation module runs the MCNP program, calculates the dose equivalent rate at 30 cm from the outer surface of the container based on the input file, and outputs it as the shielding performance result.

[0046] Finally, the optimization control module, connected to the geometric modeling and simulation calculation modules, forms the core of the entire system. This module employs an optimization algorithm, iteratively executing the following steps until the optimal shielding scheme satisfying the optimization mathematical model is obtained: First, the input file generation module generates the input file for simulation calculation based on the current design variables; second, the simulation calculation module executes simulation calculations to obtain shielding performance results; finally, the design variables are adjusted based on these shielding performance results. This iterative process continues until the optimal design scheme satisfying all shielding performance objectives and engineering constraints is found. For example, the optimization control module can use a genetic algorithm. In each iteration, the genetic algorithm evaluates the "fitness" of the current combination of design variables based on the dose rate obtained from the simulation calculation and the current constraints (such as weight and size), and generates a new set of design variable combinations. This process is repeated until a design with the lowest dose rate that satisfies all constraints is found. poly and d pb combination.

[0047] Compared to traditional radiation shielding structure design methods, the radiation shielding structure design system provided by this invention significantly improves design efficiency and optimization levels through automation and intelligence. The model building module transforms complex engineering problems into clear mathematical models, providing a definite direction for optimization. The parameterization capability of the geometric modeling module simplifies and efficiently adjusts design variables, avoiding tedious manual modeling work. The automated integration of the input file generation module and the simulation calculation module enables seamless conversion from design parameters to simulation results, greatly shortening the calculation cycle. Most importantly, the optimization control module employs intelligent optimization algorithms, enabling systematic searching across a broad design space, avoiding getting trapped in local optima, and finding a globally optimal shielding scheme that is lighter, smaller, or performs better. This not only improves design accuracy but also significantly reduces design costs and time, making radiation protection design more efficient and reliable.

[0048] Through the coordinated operation of the above modules, the radiation shielding structure design system of the present invention can achieve automation, refinement, and optimization of radiation shielding structure design. The entire process can be summarized in the following steps: First, in the model building module, the performance objectives and engineering constraints of the shielding design are clearly defined. For example, for a neutron radiation source transport container, the performance objective can be set as a dose equivalent rate limit of no more than 25 µSv / h at 30 cm from the outer surface of the container, while the engineering constraints include a total weight of the shielding body not exceeding 100 kg and a maximum radius of the container shell not exceeding 30 cm. These objectives and constraints are integrated into an optimization mathematical model to guide the subsequent optimization process.

[0049] Next, in the geometry modeling module, a parametric geometric model of the shield to be optimized is constructed. The key dimensions of this parametric geometric model are defined by a set of adjustable design variables. For example, the shielding layer of the container can be modeled as a spherical shell structure with an inner polyethylene moderating layer and an outer lead shielding layer. The thickness d of the polyethylene moderating layer... poly and the thickness d of the lead shielding layer pb These are defined as adjustable design variables. By adjusting the values ​​of these design variables, the geometry of the shield can be changed.

[0050] Subsequently, the input file generation module generates the input file required by the Monte Carlo particle transport program (such as MCNP) based on the values ​​of the current design variables. This input file details the geometry of the shield, material composition, radiation source terms (e.g., a 252Cf spontaneous fission neutron source), and dose tallies that need to be recorded (e.g., setting an F4 counter card 30 cm outside the container to calculate the neutron dose).

[0051] Then, the simulation module performs MCNP simulations and extracts shielding performance results, such as radiation dose rates at key locations, from the output file. For example, the MCNP program will calculate the shielding performance at the current d poly and d pb Under combined conditions, the dose equivalent rate at 30 cm from the outer surface of the container.

[0052] Finally, the optimization control module uses a preset optimization algorithm (such as a genetic algorithm) to evaluate whether the current design scheme meets the objectives and constraints in the optimization mathematical model based on the shielding performance results obtained from simulation calculations. If the current scheme does not meet the requirements or there is still room for optimization, the optimization control module will adjust the values ​​of the design variables and feed them back to the input file generation module, thus starting a new round of "modeling-calculation-evaluation" cycle. This iterative process will continue until the optimal shielding scheme that satisfies all constraints and has the lowest dose rate is found. For example, if the current d poly =10 cm, d pb The combination of 5 cm failed to meet the dose equivalent rate limit or weight constraint. The optimization control module may generate new combinations of design variables, such as d. poly =11 cm, d pb = 4 cm, and restart the simulation calculation and evaluation process. Through this iterative process, the optimal combination of shielding layer thicknesses is automatically found, thereby automating and optimizing the radiation shielding structure design process.

[0053] In some of the above embodiments, the radiation shielding structure design system establishes an optimized mathematical model through a model building module, constructs a parametric geometric model through a geometric modeling module, generates input files for simulation calculations through an input file generation module, performs simulation calculations through a simulation calculation module, and cyclically adjusts design variables through an optimization control module to obtain the optimal shielding scheme. However, in actual simulation calculations, especially Monte Carlo simulations, the results often exhibit statistical uncertainty. If the number of particles in the simulation calculation is insufficient, or the simulation parameters are set improperly, the accuracy of the shielding performance results may be low, thus affecting the accuracy of the optimization control module's adjustment of design variables. This could even result in the final optimal shielding scheme not being the true optimal solution, or failing to meet the expected shielding performance requirements. To address this issue and ensure the reliability of the simulation calculation results, this invention further provides a scheme for uncertainty analysis of the shielding performance results.

[0054] In specific implementation, the system also includes a result verification module, which is connected to the simulation calculation module and the optimization control module. The result verification module is used to perform uncertainty analysis on the shielding performance results. When the uncertainty is greater than a preset threshold, the simulation calculation module is controlled to increase the number of simulated particles and re-execute the simulation calculation.

[0055] Here, the result verification module is a functional unit specifically designed to evaluate the reliability of simulation results. This module receives the shielding performance results output by the simulation module and performs uncertainty analysis on them. Uncertainty analysis refers to evaluating the confidence interval or error range of the simulation results using statistical methods to quantify the reliability of the results. When the uncertainty exceeds a preset threshold, it indicates that the accuracy of the current simulation results is insufficient to support subsequent optimization decisions. In this case, the result verification module will instruct the simulation module to increase the number of simulated particles. The number of simulated particles refers to the number of particles used to simulate particle transport processes in Monte Carlo simulations. Increasing the number of simulated particles can generally improve the statistical accuracy of the simulation results and reduce uncertainty. The preset threshold can be set according to the needs of the actual application and the required accuracy of the results; for example, it can be set as a certain percentage of relative or absolute error.

[0056] Here, a result verification module is introduced to perform uncertainty analysis on the shielding performance results obtained by the simulation calculation module, thereby effectively solving the statistical uncertainty problem that may exist in the simulation calculation results. When the result verification module detects that the uncertainty of the shielding performance result is greater than a preset threshold, it considers the reliability of the current simulation result insufficient. At this time, the result verification module will promptly control the simulation calculation module to increase the number of simulated particles and re-execute the simulation calculation. This process ensures that the shielding performance results on which the optimization control module is based have sufficient accuracy and reliability, avoiding optimization direction deviations or suboptimal solutions caused by low-precision simulation results. Through this iterative verification mechanism, the system can self-correct and improve, thereby ensuring that the final optimal shielding scheme is based on highly reliable data.

[0057] The above technical solutions significantly improve the robustness of the radiation shielding structure design system and the accuracy of the optimization results. Introducing a result verification module allows the system to dynamically evaluate the quality of simulation calculations during optimization iterations and adjust simulation parameters as needed, ensuring that each optimization decision is based on reliable data. This not only avoids resource waste and design risks caused by insufficient simulation accuracy but also enables more efficient convergence to the true optimal solution, resulting in a safer, more economical radiation shielding structure that better meets actual engineering needs.

[0058] In practical applications, the result verification module can employ various uncertainty analysis methods. For example, it can use statistical measures such as standard deviation, relative standard deviation, or confidence intervals to quantify uncertainty. Specifically, after the simulation calculation module completes a simulation calculation and outputs the shielding performance results, the result verification module receives these results. Assuming the shielding performance result is a certain dose equivalent rate value, the result verification module will calculate the standard deviation of this dose equivalent rate value based on multiple sub-simulations or statistical sampling methods. If the ratio of the calculated standard deviation to the dose equivalent rate value (i.e., the relative standard deviation) exceeds a preset 5% threshold, the result verification module will determine that the uncertainty is too high. At this time, the result verification module will send an instruction to the simulation calculation module, requesting it to reduce the number of simulated particles from the current 10. 7 Increase to 10 8 The simulation module, upon receiving the instruction, adjusts its internal particle transport parameters to increase the number of simulated particles and performs the simulation again until the result verification module determines that the uncertainty meets the requirements.

[0059] In some of the above embodiments, in practical applications, the optimization algorithm adopted by the optimization control module has a decisive impact on the optimization efficiency of the system and the quality of the final solution. If the optimization algorithm is not chosen properly, it may lead to slow convergence of the optimization process, or even get stuck in local optima, making it impossible to effectively find the optimal solution that satisfies the shielding performance target and engineering constraints. In order to further improve the optimization efficiency and solution quality of the radiation shielding structure optimization design system, this invention provides an improved solution, namely, that the optimization control module adopts a specific optimization algorithm.

[0060] Specifically, the optimization control module employs one of the following optimization algorithms: parameter scanning method, genetic algorithm, or particle swarm optimization algorithm.

[0061] The parameter scanning method is one approach to find the optimal solution by traversing a preset range of parameters. Specifically, this method systematically changes the values ​​of one or more design variables within a given range, at certain step sizes or intervals, and performs simulation calculations and performance evaluations for each combination of design variables. By comparing the masking performance results under different combinations of design variables, the optimal parameter combination can be identified. This method is suitable for situations with a small number of design variables and a limited parameter range, ensuring the finding of the globally optimal solution. However, the computational cost increases significantly with the number of design variables and the scanning precision.

[0062] A Genetic Algorithm (GA) is a global optimization algorithm that simulates natural selection and genetic mechanisms. This algorithm searches for the optimal solution in the solution space of a design variable by mimicking selection, crossover, and mutation operations in biological evolution. Initially, a population of design variables is randomly generated, with each individual representing a potential masking scheme. Then, selection is made based on the masking performance (fitness) of each scheme; schemes with higher fitness have a greater probability of being retained. Through crossover, the characteristics of two parent schemes are combined to generate new offspring schemes; through mutation, some characteristics of the offspring schemes are randomly changed to increase population diversity and avoid getting trapped in local optima. These operations are repeated until a termination condition is met (e.g., reaching the maximum number of iterations or finding a sufficiently good solution). Genetic algorithms have strong global search capabilities, are suitable for complex, nonlinear optimization problems, and do not have strict requirements on the form of the objective function.

[0063] Particle Swarm Optimization (PSO) is a swarm intelligence-based optimization algorithm inspired by the foraging behavior of birds. In this algorithm, each "particle" represents a potential shielding solution and moves within the solution space of the design variables. Each particle adjusts its velocity and position based on its own optimal position (individual optimum) and the optimal position found by the entire swarm (global optimum). Specifically, a particle updates its velocity vector based on its current position, current velocity, individual optimum, and global optimum, and thus updates its position vector. Through continuous iteration, the particle swarm gradually converges towards the global optimum. PSO has advantages such as simple implementation, fast convergence, and few parameters, and performs exceptionally well in handling continuous optimization problems.

[0064] Here, by introducing one of the following algorithms—parameter scanning, genetic algorithm, or particle swarm optimization—into the optimization control module, the problems of slow convergence speed and susceptibility to local optima inherent in traditional radiation shielding methods can be effectively solved. When using the parameter scanning method, the system can systematically traverse the design space, ensuring that a globally optimal solution is found within a limited range of design variables and parameters. When using the genetic algorithm, its mechanism of simulating biological evolution enables the system to effectively perform a global search in a complex, nonlinear design space, continuously optimizing the shielding scheme through selection, crossover, and mutation operations, avoiding the trap of local optima. When using the particle swarm optimization algorithm, its swarm intelligence characteristics allow particles to quickly gather towards the optimal solution region, significantly improving the convergence speed and efficiency of the optimization process through guidance from individual and global optima. The introduction of these optimization algorithms allows the optimization control module to find the optimal shielding scheme that satisfies the shielding performance target and engineering constraints more efficiently and accurately. Compared to systems without specified optimization algorithms, this effectively avoids local optima problems during the optimization process, ensuring that a globally or near-globally optimal solution is found. For example, the global search capabilities of genetic algorithms and particle swarm optimization enable the system to explore potential solutions more comprehensively when facing complex and varied design spaces, thereby achieving superior shielding performance and more economical engineering costs. At the same time, the introduction of these algorithms makes the optimization process more intelligent and automated, reducing the need for manual intervention and improving design efficiency.

[0065] In some preferred embodiments, the optimization control module can employ a genetic algorithm to find the optimal shielding scheme. For example, when designing a multi-layered composite shielding structure, design variables may include the thickness of each shielding layer, material type, and interlayer spacing. Since these variables may have complex nonlinear relationships, traditional gradient descent methods may struggle to handle them effectively. In this case, the genetic algorithm can first randomly generate an initial population containing multiple shielding schemes, each scheme representing a combination of design variables. Then, a simulation calculation module evaluates the shielding performance of each scheme and calculates its fitness based on the evaluation results. Next, the optimization control module performs selection, crossover, and mutation operations based on the fitness to generate a new population. For example, schemes with high fitness (i.e., schemes with good shielding performance and that meet engineering constraints) have a higher probability of being selected for crossover, thus passing on superior genes (combinations of design variables) to the next generation. Through multiple iterations, the schemes in the population gradually converge to the optimal or near-optimal shielding structure. This approach effectively explores a broad design space, ultimately finding a solution that achieves the best balance between shielding performance and engineering constraints.

[0066] In practical engineering applications, the selection of shielding performance targets and engineering constraints has a significant impact on the final optimization results. Therefore, this invention defines the specific content of shielding performance targets and engineering constraints to improve the practicality and reliability of the optimization results.

[0067] Specifically, the shielding performance target refers to the dose equivalent rate limit at a designated point outside the shield. The dose equivalent rate is a crucial indicator for evaluating radiation shielding effectiveness. By setting an upper limit for the dose equivalent rate at a specific location outside the shield, it can be ensured that the shielding design meets safety standards and regulatory requirements. In practice, based on the actual application scenario, areas or equipment requiring key protection can be selected as designated points, and dose equivalent rate limits can be set according to the relevant radiation safety standards.

[0068] Engineering constraints include at least one of the following: total weight of the shielding structure, maximum external dimensions, and cost. These constraints reflect the limitations imposed on shielding design in actual engineering projects. For example, in large facilities such as nuclear power plants, the size of the shielding structure is limited by the spatial layout; while cost is an important factor that needs to be considered in all engineering designs. By incorporating these engineering constraints into the optimization mathematical model, it can be ensured that the optimized shielding scheme meets both the shielding performance requirements and the actual engineering constraints.

[0069] Clearly defining the shielding performance targets and engineering constraints allows for a more targeted optimization process, enabling the rapid identification of shielding solutions that meet practical needs. The dose equivalent rate limit directly reflects the shielding effectiveness, while engineering constraints such as total weight, maximum package size, and cost ensure the feasibility and economy of the optimized solution.

[0070] As a preferred implementation, dose equivalent rate limits at multiple designated points outside the shield can be considered simultaneously, while constraints on total weight and cost can be imposed. For example, dose equivalent rate limits can be set for three key locations outside the shield, while limiting the total weight of the shield to no more than 1 ton and the cost to no more than 100,000 yuan. In this way, a shielding solution with superior overall performance can be obtained, better meeting the needs of actual engineering projects.

[0071] The selection of design variables is crucial to the optimization results. Considering the complexity of shielding structures in practical applications, the selection of design variables needs to be flexible and adaptable. The design variables of this invention include at least one of the following: the thickness, angle, and radius of curvature of the shielding layer.

[0072] The thickness of the shielding layer directly affects the penetration depth of rays within the shield, making it a key factor influencing shielding effectiveness. Adjusting the thickness of the shielding layer effectively controls the attenuation of rays, achieving the desired shielding effect. The angle of the shielding layer determines the direction of ray incidence on the shield's surface; different angles result in different scattering and absorption paths within the shield. Optimizing the angle of the shielding layer alters the ray's transmission trajectory within the shield, thereby improving shielding efficiency. The radius of curvature of the shielding layer affects the geometry of the shield; different radii of curvature lead to focusing or diverging effects on rays. Adjusting the radius of curvature of the shielding layer changes the ray distribution within the shield, thus optimizing the shielding effect.

[0073] Here, by using the thickness, angle, and radius of curvature of the shielding layer as design variables, the structure of the shield can be adjusted more precisely, resulting in a superior shielding solution. Specifically, the optimization control module can adjust these design variables to change the geometry and material distribution of the shield, thereby affecting the transmission and attenuation of rays within the shield. Through continuous iterative optimization, an optimal combination of design variables can be found, enabling the shield to achieve the best shielding performance while meeting engineering constraints. Compared to traditional fixed-structure shields, this allows for customized design based on actual needs, better meeting diverse application scenarios. Furthermore, by optimizing the angle and radius of curvature of the shielding layer, the shielding material can be utilized effectively, thereby reducing the weight and cost of the shield.

[0074] As a specific implementation method, an optimized design of a radiation shielding structure for medical applications can be considered. In this application, the shielding structure needs to effectively shield X-rays to protect the safety of medical personnel and patients. In this case, the thickness, angle, and radius of curvature of the shielding layer can be used as design variables, and an optimization algorithm can be used to find the optimal shielding scheme. For example, the thickness of the shielding layer can be adjusted to achieve the expected dose equivalent rate limit; the angle of the shielding layer can be adjusted to change the direction of the rays incident on the surface of the shielding structure; and the radius of curvature of the shielding layer can be adjusted to optimize the distribution of rays within the shielding structure. Through continuous iterative optimization, an optimal combination of design variables can be found, enabling the shielding structure to achieve the best shielding performance while meeting the constraints of total weight and maximum external dimensions.

[0075] In some embodiments of this application, the input file generation module needs to generate an input file for simulation calculations based on the current design variables. However, if the input file template lacks placeholders associated with the design variables, or if the script program cannot accurately fill the placeholders with the values ​​of the design variables, the generated input file may be incorrect, thereby affecting the accuracy of the simulation calculations. To address this, the present invention provides an improved input file generation module, which includes an input file template and a script program. The input file template contains placeholders associated with the design variables, and the script program is used to fill the placeholders with the values ​​of the current design variables to generate the input file for simulation calculations.

[0076] The input file template is a predefined text file with a specific format and structure. This template contains placeholders related to design variables; for example, specific tags (such as "$thickness$" and "$angle$") can be used to represent the variables that need to be filled. The script is an executable piece of code that reads the values ​​of the current design variables and fills these values ​​into the corresponding placeholders in the input file template according to predetermined rules, thereby generating a complete input file that can be used by the simulation calculation module.

[0077] Specifically, the input file template can use various common text formats, such as XML, JSON, or plain text. Placeholder design should be flexible to accommodate different types of design variables. For example, numerical placeholders can be used directly for numerical variables such as thickness and angle; string placeholders can be used for character variables such as material type. The script can be implemented using various programming languages, such as Python, MATLAB, or Fortran. The script needs to have functions such as reading design variable values, parsing the input file template, and replacing placeholders. As a preferred implementation, the script can employ a parametric design method, that is, defining the mapping relationship between the input file template and design variables as parameters, thus allowing users to easily adjust them according to actual needs.

[0078] This invention achieves automatic generation of input files by setting placeholders associated with design variables in the input file template and using a script to fill these placeholders with the values ​​of the current design variables. This method avoids the tedious process of manually editing input files, reduces human error, and improves the efficiency and accuracy of optimization design. Furthermore, since the input file generation process is automated, it can significantly shorten the optimization design cycle and improve design efficiency.

[0079] In some implementations, when processing shields with repeating structures or cells, performing detailed modeling and simulation calculations for each repeating cell leads to a significant increase in computational load and a decrease in optimization efficiency. To address this, the present invention reduces computational resource consumption and improves optimization efficiency by including optimization descriptions for repeating structures or cells in the input file generation module.

[0080] The optimized description of repeating structures or cells refers to the approach of no longer describing each repeating structural unit individually when describing the geometry of a shield. Instead, the repeating pattern and quantity are defined using specific instructions or parameters. Specifically, keywords such as "repetition factor" and "array" can be used in the input file to define repeating structures, or techniques like "cell filling" can be used to describe the repeating arrangement of cells. For example, for a shielding wall composed of multiple identical modules, the geometry of only one module can be described, and then the repeating factor can be used to specify the number of times that module repeats in the horizontal and vertical directions, thus quickly constructing the geometric model of the entire shielding wall.

[0081] This effectively reduces the size of the input file and the complexity of the simulation calculation, thereby improving optimization efficiency. Specifically, by setting optimization descriptions for repeating structures or cells in the input file, detailed modeling and simulation calculations for each repeating element can be avoided, thus reducing computational resource consumption. Furthermore, optimization descriptions can simplify the geometric model construction process, improve modeling efficiency, and shorten the overall optimization design cycle.

[0082] For example, consider a shield composed of multiple identical hexagonal honeycomb structures. Without optimization description, each hexagonal honeycomb structure would need to be modeled individually, and its geometric parameters would need to be described in detail in the input file. However, by adopting the technical solution of this application, the geometric parameters of only one hexagonal honeycomb structure can be described, and then the repeating arrangement and quantity of this structure on the plane can be specified through the "array" command, thereby quickly constructing the geometric model of the entire shield. This method not only reduces the size of the input file but also reduces the complexity of the simulation calculation and improves optimization efficiency.

[0083] In practical applications, the way the simulation module performs simulation calculations has a significant impact on computational efficiency and the accuracy of the results. Therefore, this paper proposes that the simulation module use scripts to call the Monte Carlo particle transport program to perform simulation calculations, thereby improving computational efficiency and the accuracy of the results.

[0084] Monte Carlo particle transport programs are used to simulate particle transport processes in a medium through random sampling. Specifically, these programs can simulate the transport of particles such as photons, neutrons, and electrons in shielding materials, thereby evaluating the shielding performance of the shield. As a preferred implementation, Monte Carlo particle transport programs such as MCNP and Geant4 can be used. Calling the Monte Carlo particle transport program via script enables automated simulation calculations, reducing manual intervention and improving computational efficiency. Furthermore, the Monte Carlo method can handle shields with complex geometries and material compositions well, resulting in more accurate shielding performance evaluation results.

[0085] For example, a Python script can be used to call the MCNP program to perform simulation calculations. First, a template for the MCNP input file is defined in the Python script, and placeholders associated with the design variables are set. Then, based on the current values ​​of the design variables, the Python script fills the placeholders with the values, generating the MCNP input file. Finally, the Python script calls the MCNP program to perform the simulation calculations and extracts the masking performance results from the MCNP output file. After completing the simulation calculations, the simulation module automatically parses and extracts the masking performance results from the output file, thereby reducing manual intervention and improving efficiency and accuracy.

[0086] The simulation module generates output files containing a large amount of simulation data, such as particle transport trajectories, energy deposition distribution, and dose distribution. Shielding performance results refer to specific parameters extracted from this simulation data to evaluate the shielding effect, such as dose equivalent rate at a specified location and average flux rate in a specific area. Automatic parsing and extraction of shielding performance results means that the system can automatically identify relevant data in the output files and convert them into values ​​usable by the optimization and control module.

[0087] Specifically, the simulation calculation module can automatically parse and extract data using a script. This script predefines the location and format of the masking performance results in the output file, and then reads the output file to extract the relevant data according to predetermined rules. For example, the script can search for lines containing specific keywords (such as "dose rate") and then extract the dose equivalent rate value from those lines.

[0088] Furthermore, to ensure the accuracy of the extraction, the script can also perform data validation. For example, it can check whether the extracted values ​​are within a reasonable range or compare them with other known parameters. If anomalies are found, a warning is issued or the extraction is repeated. This avoids manually reviewing and recording data in the output file, reducing human error and improving the efficiency and reliability of the optimization design. In addition, the automatic parsing and extraction functions can easily integrate simulation results with other analysis tools or databases, thereby enabling more comprehensive masking design and evaluation.

[0089] Defining the optimal shielding scheme is a crucial issue. To address this, this invention proposes that the optimal shielding scheme is a numerical combination of the design variables that satisfies an optimization mathematical model.

[0090] Specifically, the numerical combination of design variables refers to a specific set of values, each corresponding to an adjustable design variable in the parametric geometric model. These design variables can be the thickness, angle, and radius of curvature of the shielding layer, among others. By adjusting the values ​​of these design variables, the geometry and material distribution of the shielding structure can be altered, thereby affecting its shielding performance.

[0091] In this context, satisfying the optimal mathematical model means that the numerical combination of design variables can achieve the optimal shielding performance target while simultaneously meeting engineering constraints. The shielding performance target can be the dose equivalent rate limit at a specified point outside the shielding body, while engineering constraints can be the total weight of the shielding body, the maximum outer dimensions, and the cost, etc. The optimal mathematical model expresses these targets and constraints in the form of mathematical formulas, so that the optimal combination of design variables that meets the conditions can be solved through optimization algorithms.

[0092] Therefore, by defining the optimal shielding scheme as a numerical combination of design variables that satisfy the optimization mathematical model, the optimization objective can be described more accurately, and a clear optimization direction can be provided for the optimization algorithm. This definition not only considers shielding performance but also comprehensively takes into account engineering constraints, thus yielding more practical and feasible shielding schemes.

[0093] A radiation shielding structure design method based on MCNP is presented, employing the aforementioned MCNP-based radiation shielding structure design system. The flowchart of this method is shown below. Figure 1 As shown, the specific steps include: S1: Establish an optimization mathematical model that includes shielding performance targets and engineering constraints; S2: Construct a parametric geometric model of the shield to be optimized, wherein the key dimensions of the parametric geometric model are defined by a set of adjustable design variables; S3: Generate the MCNP input file from the current design variables; S4: Perform MCNP calculations to obtain shielding performance results; S5: Perform uncertainty analysis on the shielding performance results. If the uncertainty is less than or equal to the preset threshold, output the shielding performance results; if the uncertainty is greater than the preset threshold, increase the number of simulated particles and proceed to step S4. S6: Use an optimization algorithm to evaluate the shielding performance results. If the convergence condition is met, output the current design variable as the optimal shielding scheme; if the convergence condition is not met, proceed to step S7. S7: Adjust the design variables based on the shielding performance results, and repeat steps S3-S6.

[0094] Finally, it should be noted that the above embodiments of the present invention are merely illustrative examples and not intended to limit the implementation of the invention. Those skilled in the art can make other variations and modifications based on the above description. It is impossible to exhaustively list all possible implementations here. All obvious variations or modifications derived from the technical solutions of this invention are still within the scope of protection of this invention.

Claims

1. A MCNP-based radiation shielding structure design system, characterized by, The system comprises a model establishing module, a geometric modeling module, an input file generating module, a simulation calculation module and an optimization control module. The model establishing module is configured to establish an optimization mathematical model comprising shielding performance targets and engineering constraints. The geometric modeling module is configured to construct a parameterized geometric model of the shielding body to be optimized, wherein key dimensions of the parameterized geometric model are defined by a set of adjustable design variables. The input file generating module is connected to the geometric modeling module, and the simulation calculation module is connected to the input file generating module. The optimization control module is connected to the geometric modeling module and the simulation calculation module, and is configured to adopt an optimization algorithm to control the input file generating module to generate an input file for simulation calculation according to current design variables, and control the simulation calculation module to perform simulation calculation to obtain shielding performance results, and adjust the design variables according to the shielding performance results, until an optimal shielding scheme satisfying the optimization mathematical model is obtained.

2. The MCNP-based radiation shielding structure design system of claim 1, wherein, The system further comprises a result checking module connected to the simulation calculation module and the optimization control module, and configured to perform uncertainty analysis on the shielding performance results, and control the simulation calculation module to increase the number of simulated particles and perform simulation calculation again when the uncertainty is greater than a preset threshold.

3. The MCNP-based radiation shielding structure design system of claim 1, wherein, The optimization algorithm adopted by the optimization control module is one of a parameter scanning method, a genetic algorithm or a particle swarm algorithm.

4. The MCNP-based radiation shielding structure design system of claim 1, wherein, The shielding performance targets comprise dose equivalent rate limits of specified points outside the shielding body, and the engineering constraints comprise one or more of total weight, maximum outer package size and cost of the shielding body.

5. The MCNP-based radiation shielding structure design system of claim 1, wherein, The design variables comprise one or more of thickness, angle and radius of curvature of the shielding layer.

6. The MCNP-based radiation shielding structure design system of claim 1, wherein, The input file generating module comprises an input file template and a script program, the input file template comprises placeholders associated with the design variables, and the script program is configured to fill numerical values of the current design variables into the placeholders to generate the input file for simulation calculation.

7. The MCNP-based radiation shielding structure design system of claim 1, wherein, The input file generating module is further configured to set optimization descriptions for repeating structures or cells in the input file.

8. The MCNP-based radiation shielding structure design system of claim 1, wherein, The simulation calculation module performs simulation calculation by calling a Monte Carlo particle transport program through a script, and automatically parses and extracts the shielding performance results from an output file after the simulation calculation is completed.

9. The MCNP-based radiation shielding structure design system of claim 1, wherein, The optimal shielding scheme is a numerical combination of the design variables satisfying the optimization mathematical model.

10. A method for designing a radiation shielding structure based on MCNP, characterized by, The system comprises the following steps: S1: establishing an optimization mathematical model comprising shielding performance targets and engineering constraints; S2: constructing a parameterized geometric model of the shielding body to be optimized, wherein key dimensions of the parameterized geometric model are defined by a set of adjustable design variables; S3: generating an MCNP input file from the current design variables; S4: performing MCNP calculation to obtain shielding performance results; S5: performing uncertainty analysis on the shielding performance results, and outputting the shielding performance results if the uncertainty is less than or equal to a preset threshold, or increasing the number of simulated particles and entering step S4 if the uncertainty is greater than the preset threshold. S6: The optimization algorithm is used to evaluate the shielding performance result, if the convergence condition is met, the current design variable is output as the optimal shielding scheme; if the convergence condition is not met, step S7 is entered; S7: The design variable is adjusted according to the shielding performance result, and steps S3-S6 are repeated.

Images