Quantum Artificial Bee Colony Algorithm for Knapsack Problem

Abstract

The traditional quantum evolutionary algorithm takes a long time to converge and can be easy trap into local optima. In order to overcome and accelerate the speed of the convergence, a new quantum evolutionary algorithm is proposed in the paper. The proposed new algorithm named discrete quantum bee colony algorithm incorporate the basic idea of the artificial bee colony algorithm. The initial population can be initialized randomly using quantum encoded and the population can be formed by there parts and every subpopulation can evolve cooperatively. In the end, the individual will rated according to the multi-granularity mechanism and also rated according to the evolution condition. Simulation results of knapsack problems show that the proposed algorithm performs better than other algorithm.