A new approach for heuristics-guided search in the In-Core Fuel Management Optimization

摘要

The In-Core Fuel Management Optimization (ICFMO) is a well-known problem of Nuclear Engineering whose features are complexity, high number of feasible solutions, and a complex evaluation with high computational cost, which makes it prohibitive to have a great number of evaluations during an optimization process. The use of optimization metaheuristics such as Genetic Algorithms, Particle Swarm and Ant Colonies Optimization has been successful in ICFMO. In this paper, we propose a new approach for the use of Relational Heuristics to guide the metaheuristic search of the ICFMO using approximations with the Reactive Neighborhood Acceptance Heuristic (RNAH). The RNAH is applied to Random Search and the Particle Swarm Optimization and compared to previous results in the literature. Results demonstrate that it is possible to reduce the computational cost using this approach and therefore to evaluate more interesting candidate solutions on the long run.