Distribution network reconfiguration using an efficient evolutionary algorithm

Abstract

Network reconfiguration is an important tool to optimize the operating conditions of a distribution system. This is accomplished modifying the network structure of distribution feeders by changing the open/close status of sectionalizing switches. This not only reduces the power losses, but also relieves the overloading of the network components. Network reconfiguration belongs to a complex family of problems because of their combinatorial nature and multiple constraints. This paper proposes a solution to this problem, using a specialized evolutionary algorithm, with a novel codification, and a brand new way of implement the genetic operators considering the problem characteristics. The algorithm is presented and tested in a real distribution system, showing excellent results and computational efficiency. © 2007 IEEE.