QACO_Alpha applied to the nuclear reactor core fuel reload optimization

Abstract The nuclear reload optimization is an important issue to nuclear engineering. It consists on maximizing the length of the operation cycle of the power plant. The aim is to find a configuration of the fresh fuel assemblies and remnants in order to keep the power plant running at full power by the largest time as possible. Quantum-inspired evolutionary algorithms are optimization tools based in artificial intelligence developed to simulate the quantum processing in classical computers. In this work is introduced one of these tools, which adds quantum concepts to the biological metaphor of collective learning of the real ants, named QACO_Alpha. It uses mechanisms developed in order to avoid premature convergence problems like a pheromone evaporation step besides a new updating method. To show its effectiveness, QACO_Alpha was applied to the optimization of 7th cycle of Angra 1. Experimental results were confronted to that obtained with other optimization methods, qualifying QACO_Alpha as a valid optimization tool for this kind of problem.