Abstract:The optimization design of fast reactors faces two major problems: insufficient design or operation experience and more complicated optimization objectives, compared with pressurized water reactor (PWR). An optimization method which is independent of application banckground and experience is required. Intelligent optimization algorithm is a strategy for the search of optimal or satisfactory solution for given design objectives and satisfies above requirements. In the process of search, multiple design objectives and constraints are considered simultaneously and the times of search and calculation are limited. Genetic algorithm is a classical and representative method in numerous of optimization algorithms, which performs well in solving multi-objective problems with constraints. Genetic algorithm has inherent parallel characteristics as well for which the searching efficiency is able to be increased. Based on a genetic algorithm module in an open source toolkit DAKOTA and a self-developed neutron transport program SARAX, an optimal design frame for fast reactors with multivariable, multi-objective and multi-constraint problems was constructed in this work. Both genetic algorithm module and neutron transport program performed as a black-box. A coding system that could transform a decimal integer and a loading pattern to each other was suggested and applied to this optimal frame. This coding strategy realized the coverage of the multi-dimensional search space, so that the optimal frame can deal with the problems of continuous and discrete variables. For the constrained multi-objective problem, the direct ranking method and the weighted sum method were both applied and compared. Two cases were analyzed to test this frame: 1) Rearrange the subassemblies of ABTR to find a scheme whose keff and power peak factor meet the given expectations, the searching result was compared with the enumeration results; 2) Optimizing the design of the loading position of the 237Np-containing subassemblies for China Experimental Fast Reactor (CEFR). The objective of this case was obtaining more 238Pu after a cycle of burnup, ensuring safety constraints. The first calculation example was used to verify the feasibility of the new coding scheme. The second case demonstrated the specific application of the optimal frame in engineering practice. According to the numerical results, the fast reactor optimal frame proposed in this work achieves the research objectives of finding the best solutions and meeting the requirements under a limited number of searches with multiple variables, multiple design objectives and multiple constraints. What’s more, the coding of variables, the optimization objectives setting and the given of constraints are all of strong flexibility. To sum up, this frame can cope with the complex requirements of fast reactor engineering design.

Development of Optimal Framework Sarax/Dakota for Multiple Objective Optimization of Fast Reactor and Its Application

Research on Multi-objective Optimization of Fast Reactor and Development of SARAX/DAKOTA Optimal Frame

Multi-objective Optimization of Operating Parameters Based on Neural Network and Genetic Algorithm in the Blast Furnace

Application of genetic algorithms in optimization of SFR nuclear reactor design

Application Of Dual-Adaptive Niched Genetic Algorithm In Optimal Design Of Nuclear Power Components

Study on multi-objective optimization for nuclear reactor radiation shield design coupling genetic algorithm with paralleling discrete ordinate code

SARAX: A New Code for Fast Reactor Analysis Part I: Methods

MOF: A Modular Framework for Rapid Application of Optimization Methodologies to General Engineering Design Problems

Design Optimization for Pressurized Water Reactor Using Improved Quantum Fish Swarm Algorithm and Intuitionistic Linguistic Decision-Making

Integrating the Deep Learning and Multi-Objective Genetic Algorithm to the Reloading Pattern Optimization of HPR1000 Reactor Core

Development of an Optimized Transport Solver in SARAX for Fast Reactor Analysis

Optimization of Moderator Materials by NSGA II Based on Macroscopic Cross-Sections: Applications in Accelerator Neutron Sources

Optimization of physical design for small long life Lead Cooled fast Reactor based on SARAX

Neutronics Design Optimization of a Small Modular Fast Reactor Based on Response Surface Methodology

Developments of the sodium fast reactor analysis code SARAX: methods and verification

Nsga-Ii Algorithm with A Local Search Strategy for Multiobjective Optimal Design of Dry-Type Air-Core Reactor

Optimization and Improvement of Sodium Heated Once-through Steam Generator Transient Analysis Code Based on the JFNK Algorithm

Multi-objective Optimization Strategies for Radiation Shielding Design with Genetic Algorithm.

Efficient optimization of an accelerator neutron source for neutron capture therapy using genetic algorithms

Multi-objective optimization of a steam generator

Population-based metaheuristic optimization in neutron optics and shielding design