放疗逆向计划中的混合多目标优化算法研究

放射治疗中的逆向计划为多目标优化问题.将模拟退火算法和遗传算法相结合,进行多目标优化算法的研究,利用测试函数分析算法的优劣.将基于交叉变异扰动的多目标SA算法应用于放疗逆向计划过程,用基于平均剂量分布的目标函数和基于混合剂量-体积约束的目标函数进行逆向优化计算,说明算法的有效性.     

反应堆辐射屏蔽多目标优化方法研究

为解决基于蒙特卡罗方法的传统屏蔽优化方法效率低、可应用性差的缺点,本文基于精英策略的非支配排序遗传算法（NSGA-Ⅱ）和小批量随机梯度下降法（MBGD）对反应堆屏蔽优化方法进行了研究,同时改进了遗传算法自适应变异率算子以增强遗传算法的全局寻优能力,提出了反应堆屏蔽多目标优化方法。构建反应堆二次屏蔽多目标优化模型,将蒙特卡罗方法与神经网络预测方法输出的屏蔽后归一化中子透射率进行对比,验证了MBGD的准确性。通过神经网络与NSGA-Ⅱ的耦合对屏蔽参数进行约束寻优,能够快速找到屏蔽设计模型的Pareto前沿,可实际应用于反应堆辐射屏蔽多目标优化工程设计。      

基于时间序列分析方法对安全壳泄漏率测量阶段的气体弛豫过程研究

安全壳试验期间安全壳内气体在泄漏率测量平台经历了温度、蒸汽分压等参数波动并再次进入平稳的弛豫过程,本文针对判断新稳态建立的标准、气体弛豫时间、影响弛豫过程的因素等内容进行分析,以期通过计算检验统计量进行平稳性检验的方法给出判断平稳的标准,并对弛豫过程中各项参数的时间序列进行序列分解,对各参数的弛豫过程分别进行分析。研究结果表明,安全壳内气体因安全壳加压造成的蒸汽分压不均匀是影响弛豫时间的主要因素。因此,进一步提出控制蒸汽分压不平衡势,以缩短泄漏率结果的弛豫时间。      

Multi-objective optimal design of vertical natural circulation steam generator

In this work, a hybrid non-dominated sorting genetic algorithm was proposed and utilized to perform the multi-objective optimization design of a natural circulation steam generator, which included minimizing of the weight, the volume and the reactor coolant flow-rate. Sensitivity analysis of the design variables was carried out to study the relationships between the optimization variables and the objective functions, which was also helpful for the explanation of the optimization results. The mathematical model of the steam generator was validated by the RELAP5 code. The results show that the mathematical model has a good agreement with the RELAP5 model after modifying the boiling correlation in the secondary side; the proposed hybrid non-dominated sorting genetic algorithm is able to find much better spread of solutions and better convergence near the true Pareto optimal front compared to the non-dominated sorting genetic algorithm; reactor inlet temperature is the most important variable which influences the distribution of Pareto optimal solutions.     

基于遗传算法的中子屏蔽材料组分优化研究

基于快速非支配排序遗传算法NSGA-II,开展了多目标屏蔽优化设计研究,建立了中子复合屏蔽材料组分的自动优化设计程序。以屏蔽剂量和材料密度最小化为目标,以聚乙烯、铅、硼、锂、铁、铝等材料均匀混合组成30 cm厚平板屏蔽结构为例,验证了优化算法程序的有效性。将基于遗传算法的屏蔽优化方法与设计人员的经验相结合,可更高效地实现多目标屏蔽优化设计。     

基于非支配排序遗传算法的辐射屏蔽多目标优化方法研究

为提高反应堆辐射屏蔽结构设计效率与设计性能,减少传统辐射屏蔽设计方法的主观经验影响。本文基于非支配排序遗传算法对反应堆屏蔽结构开展多目标优化方法研究,并开发了反应堆辐射屏蔽多目标优化计算分析程序;利用典型反应堆辐射屏蔽结构模型对此优化方法和计算程序开展了初步验证。结果表明,非支配遗传算法可正确有效地用于辐射屏蔽结构的设计,优化效果显著。      

Optimization of the core configuration design using a hybrid artificial intelligence algorithm for research reactors

To successfully carry out material irradiation experiments and radioisotope productions, a high thermal neutron flux at irradiation box over a desired life time of a core configuration is needed. On the other hand, reactor safety and operational constraints must be preserved during core configuration selection. Two main objectives and two safety and operational constraints are suggested to optimize reactor core configuration design. Suggested parameters and conditions are considered as two separate fitness functions composed of two main objectives and two penalty functions. This is a constrained and combinatorial type of a multi-objective optimization problem. In this paper, a fast and effective hybrid artificial intelligence algorithm is introduced and developed to reach a Pareto optimal set. The hybrid algorithm is composed of a fast and elitist multi-objective genetic algorithm (GA) and a fast fitness function evaluating system based on the cascade feed forward artificial neural networks (ANNs). A specific GA representation of core configuration and also special GA operators are introduced and used to overcome the combinatorial constraints of this optimization problem. A software package (Core Pattern Calculator 1) is developed to prepare and reform required data for ANNs training and also to revise the optimization results. Some practical test parameters and conditions are suggested to adjust main parameters of the hybrid algorithm. Results show that introduced ANNs can be trained and estimate selected core parameters of a research reactor very quickly. It improves effectively optimization process. Final optimization results show that a uniform and dense diversity of Pareto fronts are gained over a wide range of fitness function values. To take a more careful selection of Pareto optimal solutions, a revision system is introduced and used. The revision of gained Pareto optimal set is performed by using developed software package. Also some secondary operational and safety terms are suggested to help for final trade-off. Results show that the selected benchmark case study is dominated by gained Pareto fronts according to the main objectives while safety and operational constraints are preserved.     

基于改进MOGA的供应商选择方法以及在核电设备采购中的应用

由于核电站建设周期长、投入高,且对运营过程的安全性和经济性有很高要求.因此,可用于供应商选择的决策支撑方法较少,供应商评价流程客观性相对不高.针对这一问题,本文提出了一种客观可行的、基于改进多目标遗传算法(MOGA)的供应商选择方法.仿真试验结果表明了该方法在核电设备采购实践中的有效性和实用性.     

基于不同图像质量评价标准的自适应辐射图像增强

实现了基于非完全Beta函数的遗传算法以及基于此遗传算法的自适应灰度变换,在此基础上,使用3种图像质量评价标准实现了辐射图像的自适应灰度变换,比较了3种方法的处理时长、稳定性、遗传代数的影响等,将其应用到集装箱DR/CT检测系统图像处理模块中,实现了有效的自适应图像增强。     

遗传算法在高通量工程试验堆燃料管理优化中的应用

建立了基于遗传算法的高通量工程试验堆(HFETR)堆内燃料管理优化模型。根据HFETR的实际情况，提出了基于组件位置的二进制编码/解码技术。研究了不同选择策略对算法性能的影响，探讨了遗传算法和专家经验的结合，吸收了自适应遗传算法的思想，得到了可直接应用于HFETR的优化换料方案。     

船用堆辐射屏蔽优化设计平台开发与验证

船用堆对核反应堆屏蔽设计提出了更高的要求,传统辐射屏蔽设计方法及设计软件已不能满足要求。为了得到更加精确的辐射屏蔽设计,本文基于开源的SALOME框架建立了一套集“几何建模-材料建模-屏蔽优化-结果可视化”功能为一体的船用堆辐射屏蔽多目标优化平台——MOSRT。MOSRT平台可实现屏蔽结构三维CAD实体建模、基于遗传算法的辐射屏蔽多目标优化以及屏蔽计算结果剂量场三维可视化。基于Savannah和MRX船用堆模型对MOSRT平台进行了辐射屏蔽优化验证,优化方案与初始方案相比,在剂量、质量、体积方面均得到了良好的优化效果,证明了MOSRT平台初步具备辐射屏蔽优化设计功能,可为船用堆工程及概念屏蔽设计提供辅助设计手段。      

多目标辐射屏蔽优化设计方法

由于复杂核装置的屏蔽设计目标的多样化,同时屏蔽设计过程的不确定因素众多,因此有必要开发一种智能屏蔽优化设计的方法,实现屏蔽方案选择的自动化,减少人因等不确定因素的影响。本工作结合遗传算法与离散纵标方法,同时考虑造价、体积、重量等的最小化,开发了遗传算法多目标屏蔽优化程序,实现了经济可行的辐射屏蔽设计方案的自动化获取。该工作对优化屏蔽设计方案的获取有一定的现实意义。     

Fuzzy identification of T-S model for beam stability control for electron gun

In this paper, Takagi-Sugeno(T-S) fuzzy control is proposed for stabilizing the output beam of accelerators. To model the nonlinear system, we proposed a hybrid optimization algorithm based on quantum-inspired differential evolution and genetic algorithm. Based on the T-S model identified, the corresponding statefeedback fuzzy controller is designed. The method is applied to the La B6 electron gun system in the industrial radiation accelerator and the simulation results show its effectiveness.     

Optimization of zigzag flow channels of a printed circuit heat exchanger for nuclear power plant application

Printed circuit heat exchanger (PCHE) is recently considered as a recuperator for the high-temperature gas cooled reactor. In this study, shape optimization of zigzag flow channels in a PCHE has been performed to enhance heat transfer performance and reduce the friction loss based on three-dimensional Reynolds-averaged Navier–Stokes analysis with the Shear Stress Transport Turbulence model. A multi-objective genetic algorithm is used for the multi-objective optimization. Two non-dimensional objective functions related to heat transfer performance and friction loss are employed. The shape of a flow channel is defined by two geometric design variables, viz. the cold channel angle and the ellipse aspect ratio of the cold channel. The experimental points within the design space are selected using Latin hypercube sampling as the design of the experiment. The response surface approximation model is used to approximate the Pareto-optimal front. Five optimal designs on the Pareto-optimal front have been selected using k-means clustering. The flow and heat transfer characteristics, as well as the objective function values, of these designs have been compared with those of the reference design.     

Regulation of nuclear power plants a multi objective approach

The regulation of nuclear power plant (NPP) is evolving in a direction to harmonize probabilistic safety criteria in the near future. The utilities will not only have to demonstrate that they are operating below a target risk level but also to demonstrate that the unavailability of some of the critical safety systems are below a specified level. In order to satisfy the Technical Specification and Maintenance (TS&M) requirements in a cost effective manner multi-objective optimization of TS&M requirements is of profound interest. The constrained multi-objective optimization of the TS&M requirements of a nuclear power plant (NPP) based on risk and cost gives the pareto-optimal solutions, from which the utility can pick suitable decision variables. The paper presents a multi objective genetic algorithm (GA) technique to investigate a trade-off between risk and cost both at the system and the plant level for Loss of Coolant Accident (LOCA) and Main Steam Line Break (MSLB) as initiating events in a NPP.     

Optimal design of the first stage of the platefin heat exchanger for the EAST cryogenic system

The size of the heat exchanger is an important factor determining the dimensions of the cold box in helium cryogenic systems. In this paper, a counter-flow multi-stream plate-fin heat exchanger is optimized by means of a spatial interpolation method coupled with a hybrid genetic algorithm.Compared with empirical correlations, this spatial interpolation algorithm based on a kriging model can be adopted to more precisely predict the Colburn heat transfer factors and Fanning friction factors of offset-strip fins. Moreover, strict computational fluid dynamics simulations can be carried out to predict the heat transfer and friction performance in the absence of reliable experimental data. Within the constraints of heat exchange requirements, maximum allowable pressure drop, existing manufacturing techniques and structural strength, a mathematical model of an optimized design with discrete and continuous variables based on a hybrid genetic algorithm is established in order to minimize the volume. The results show that for the first-stage heat exchanger in the EAST refrigerator, the structural size could be decreased from the original2.200?×?0.600?×?0.627(m~3) to the optimized 1.854?×?0.420?×?0.340(m3), with a large reduction in volume. The current work demonstrates that the proposed method could be a useful tool to achieve optimization in an actual engineering project during the practical design process.     

Genetic algorithm optimization of a model-free fuzzy control system

In this work, we present an approach, based on a genetic algorithm optimization search, for determining the values of the parameters of an adaptive stable fuzzy control system suitable to drive a given plant to a desired reference trajectory. The method is demonstrated on a naturally unstable system taken from the literature and then applied to the well known Chernick’s model of a nuclear reactor.     

Particle swarm evolutionary computation-based framework for optimizing the risk and cost of low-demand systems of nuclear power plants

In this paper, an adapted multi-objective multi-swarm co-evolutionary particle swarm optimization (PSO) framework is developed to simultaneously optimize the risk and cost of low-demand systems of nuclear power plants (NPPs). In the built framework, multi-swarm co-evolutionary strategy is introduced to handle the fitness assignment puzzle of multi-objective optimization problems. Besides, to deal with the mixed-integer problem of the decision variables vector, a sub-interval covering-based nearest boundary method is also adopted. To illustrate the effectiveness and efficiencies of the proposed method, a typical high-pressurized injection system (HPIS) is analyzed. The results indicate that, compared with the classic non-dominated sorting genetic algorithm (NSGA)-II approach, the proposed method is more simple and easier to be convergent, besides, of which the Pareto front is better distributed.     

基于非支配排序遗传算法的核动力堆中子-γ混合射线屏蔽智能优化

以萨瓦纳船用核动力堆为原型,等比构建了中子-γ混合辐射场多目标优化模型,使用非支配排序遗传算法（NSGA-Ⅱ）与神经网络相结合的屏蔽智能优化方法,将屏蔽层总重量和屏蔽后的剂量率作为优化目标,进行多目标寻优,得到了pareto最优解;选取其中1组最优解分别利用蒙特卡罗方法计算和神经网络预测进行可行性对比验证,在神经网络预测误差允许的范围内,得到的剂量率均满足寻优时设置的约束限值。研究结果表明,该屏蔽智能优化方法对反应堆中子 γ混合射线的屏蔽参数优化是可行的,相比于传统的纯蒙特卡罗方法而言,能在计算准确的前提下极大减少计算时间。     

基于SARAX/DAKOTA的快堆设计多目标优化框架开发与验证

与具有大量运行经验、设计目的相对单一的压水堆相比,快堆设计经验少、设计目的更复杂,难以直接使用经验方法实现考虑多目标、多约束的堆芯设计以及布料方案的优化。本文将遗传算法用于快堆设计的多目标优化问题,对传统的遗传算法在编码方案和多目标的处理上进行了改进,使用中子学分析软件SARAX与优化工具箱DAKOTA搭建了针对快堆设计与布料方案优化的框架,并进行了功能上的验证和实际问题的应用。