Multi-objective optimization strategies for radiation shielding design with genetic algorithm?

The radiation shielding design for advanced nuclear facilities is a typical complicated multi-objective and multi-parameter optimization problem in the nuclear engineering. To obtain an optimal solution of the shielding design is of significance in developing high-performance advanced nuclear facilities, especially for compact and mobile devices. The classical method of shielding design is a brute force trial-and-error procedure subjecting to human preferences and expectations, which is of failure to meet the requirements in radiation shielding optimization applications. Two multi-objective optimization strategies were developed to optimize the shielding structures and materials aiming at lightweight, compactness and low radiation dose under a set of constraints. The strategies employed an evolutionary algorithm, genetic algorithm, to perform radiation shielding design optimization efficiently and automatically. The most advantage of the strategies is that multiple optimal shield solutions could be achieved in one single simulation run, which will make the radiation shielding design procedure more efficient and flexible. The strategies were verified fully with a realistic multi-objective radiation shielding design problem. The numerical results showed that the strategies could balance well the shielding quality against the weight and the volume of the shield. It is confirmed that the strategies are applicable and effective for multi-objective and multi-parameter radiation shielding design optimization applications. (C) 2020 Elsevier B.V. All rights reserved.

Automated summary

This study developed two multi-objective optimization strategies using genetic algorithm to efficiently and automatically optimize radiation shielding design, successfully balancing the relationship between shielding quality against the weight and volume of the shield, improving the efficiency and flexibility of radiation shielding design.