Quantum-inspired evolutionary algorithms and its application to numerical optimization problems

Abstract

This work proposes a new kind of evolutionary algorithm inspired in the principles of quantum computing. This algorithm is an extension of a proposed model for combinatorial optimization problems which uses a binary representation for the chromosome. This extension uses probability distributions for each free variable of the problem, in order to simulate the superposition of solutions, which is intrinsic in the quantum computing methodology. A set of mathematical operations is used as implicit genetic operators over those probability distributions. The efficiency and the applicability of the algorithm are demonstrated through experimental results using the F6 function. © Springer-Verlag Berlin Heidelberg 2004.