Part 2 of 2 Parts (Please read Part 1 first)
In their research, the authors developed improved crossover and mutation operators. An LP can be considered as a two-dimensional array containing materials of different types such as fuel, absorber and reflector. Each LP is represented by a “chromosome” whose “genes” represent the different locations and types of FAs in the core. The chromosome representation is chosen to be a permutation of the core structure. This is done in order to preserve the predetermined quantities of the different materials and elements in the core. This representation is selected because it gives researchers simple and intuitive physical meaning to the “genetic” variation of the of the population. Variations that are similar have similar LPs.
Another contribution of the authors is that they are taking geometrical aspects into account by considering the physical spatial structure of the core. They have developed a new geometric crossover operator based on the layout of the core. Their tests indicate that this approach yields excellent results for optimization. Crossover is the genetic operator that is responsible for creating new entities based on two or more parents. The operator swaps gene segments between the parents to create offspring with a mixture of the parents’ genes. This done by swapping rectangular segments of neighboring FAs between two selected LP parents.
The researchers from BGU also came up with highly adaptive mutation techniques. These are based on the instantaneous genetic variance of the population. The algorithm tracks the genetic diversity of the population in real time. Then it automatically alters the mutation rate based on the level of homogeneity of the population. As the population becomes more homogenous, the mutation rate is raised. This approach results in enhanced algorithmic performances.
The BGU researchers challenge the traditional assumptions about symmetrical core design which have dominated the field of LP design up to this point. This assumption is based on the fact that the different primary coolant loops in the nuclear reactor have to maintain similar thermal-hydraulic conditions during normal operation. This imposes symmetry on the power distribution of the reactor core. Symmetrical LP designs are much more intuitive than other layouts and engineers who design nuclear reactors often include their intuitions and experiences in reactor core design. Such symmetry constraints are not necessarily honored in the design of research reactors and small modular reactors. The BGU researchers came to the conclusion that the best LPs are not necessarily symmetrical.
This study was an excellent example of an interdisciplinary project. In order to carry it out, the researchers had to have major expertise in both GAs and nuclear reactor physics. This is obviously an important area of research, but it is just in the beginning stages. Any tool that improves the operation of nuclear power plants is certainly welcome as the nuclear industry competes with cheap natural gas and renewable sources with falling prices. Many nuclear reactors are in danger of being shut down because they are too uneconomical to maintain and operate.