A Comparative Study of Selective Breeding Strategies in a Multiobjective Genetic Algorithm

The design of Pressurized Water Reactor (PWR) reload cores is a difficult combinatorial optimization problem with multiple competing objectives. This paper describes the use of a Genetic Algorithm (GA) to perform true multiobjective optimization on the PWR reload core design problem and improvements made to its performance in identifying nondominated solutions to represent the trade-off surface between competing objectives. The use of different pairing strategies for combining parents is investigated and found to produce promising results in some cases.