Yuanzineng kexue jishu (Feb 2023)
Optimizing Design of Accelerator Boron Neutron Capture Therapy Beam Shaping Assembly Using Genetic Algorithm
Abstract
The accelerator-based boron neutron capture therapy (AB-BNCT) facility is a cancer treatment assembly based on epithermal neutrons generated by an accelerator, which can be built in hospitals in densely populated cities. BNCT has strict requirements for the neutron fluence rate and various contaminations. In order to meet these requirements, the neutron beam shaping assembly (BSA) needs to be designed. The traditional design method for BSA needs to know the performance of various materials in advance in terms of moderation, alignment, shielding, etc. Many researchers have done a lot of work in this area, but the neutron energy changes due to complex nuclear reactions between neutrons and materials. Since the nuclear reaction cross section is closely related to energy, the judgments on the properties of these materials are mostly qualitative and has only little use for the design of BSA. The traditional method first gives a preliminary plan, and then adjusts the material type and material size according to the theoretical calculation results, it is easy to fall into the local optimum, and it is difficult to obtain the global optimal solution. This process takes a lot of time. Traditional design methods rely heavily on the designer’s experience, which is not a good guide for design work. Genetic algorithm is a random global search and optimization method developed by imitating the biological evolution mechanism. It draws on some ideas in evolutionary biology such as heredity, mutation, natural selection and hybridization, and is a general algorithm for solving search problems. The algorithm has little requirements for the problem to be solved, and only needs the fitness function. The genetic algorithm can deal with multiple individuals in the group at the same time, perform parallel search according to multiple routes, and use the roulette method to make selections. The advantage of genetic algorithm is that the search range is large, which is conducive to global selection. Genetic algorithm also has the characteristics of robustness, parallelism and portability. Genetic algorithm has applications in nuclear power plant fuel management, shielding design and reactor BNCT BSA design.In order to solve the problems existing in traditional methods, based on the 14 MeV cyclotron, an optimal design method of BSA based on genetic algorithm was studied in this paper. The program mainly included genetic algorithm module, fitness calculation module, input file generation module and output file processing module. The results show that the method can realize the multi-objective optimization design and obtain good results efficiently. The program can be modified for design in other related fields of nuclear engineering.
