Stellarator Optimization Using a Distributed Swarm Intelligence-Based Algorithm

keywords: Large-scale optimization problems, swarm intelligence, distributed break computing
The design of enhanced fusion devices constitutes a key element for the development of fusion as a commercial source of energy. Stellarator optimization presents high computational requirements because of the complexity of the numerical methods needed as well as the size of the solution space regarding all the possible configurations satisfying the characteristics of a feasible reactor. The size of the solution space does not allow to explore every single feasible configuration. Hence, a metaheuristic approach is used to achieve optimized configurations without evaluating the whole solution space. In this paper we present a distributed algorithm that mimics the foraging behaviour of bees. This behaviour has manifested its efficiency in dealing with complex problems.
reference: Vol. 31, 2012, No. 1, pp. 31–43