The Particle Swarm Optimization Algorithm Applied to Nuclear Systems Surveillance Test Planning

This work presents Particle Swarm Optimization (PSO) as an alternative method for optimizing surveillance test policies in nuclear power plant (NPP) electromechanical systems, which has been successfully handled by the use of Genetic Algorithms (GA). The main idea is to find the optimum interval between test interventions, for each component of the system, considering as main objective, the system’s average availability, during a given time period. Computational experiments demonstrated that PSO was able to find optimized surveillance test policies. In the case study used in this work, PSO has outperformed the GA, achieving slightly better results, with lower computational efforts