Inteligência computacional para sistemas self-healing: um estudo de caso em redes elétricas inteligentes

Abstract

This work employs computational intelligence techniques in optimization problems found in smart grid systems. A feature of the smart grid is the auto-recovery capability of a system, for example, restore normalcy of a power distribution network after occurrence of a fault. When a system has the capability of auto-recovery after failure, this is called self-healing. An emerg-ing alternative to a self-healing system is the application of population-based techniques. Optimization algorithms by particle swarms are alternatives to find topologies able to restore a system. To this the power distribution network was modeled using graph theory, as well used the sum of the currents method for calculating the power flow. The recovery of network after fault is performed using bio-inspired algorithms adapted to the optimization problem in smart grids. The results evaluate the different factors that influence the algorithm execution and its resolution capability of the problem addressed.