An Improved Quantum-behaved Particle Swarm Optimization Algorithm Based on Chaos Theory Exerting to Local Optimal Position
Dazhi Pan · Journal of Information and Computational Science · 2013
In this paper, we introduce chaos theory into QPSO and propose an improved Quantum-behaved Particle Swarm Optimization (QPSO), in which the logistic map is exerted to every particle’s local optimal position P (t) at a certain probability. In this improved QPSO, the logistic map is used to generate a set of chaotic offsets and produce multiple positions around P (t). According to their fitness values, the particle’s position X(t) and P (t) are updated. In order to further enhance the diversity of particles, mutation operation is introduced into and acts on one dimension of the particle position. In improved QPSO, the chaos and mutation probabilities are carefully selected. Through several typical function experiments, its results show that the convergence accuracy of the improved QPSO is better than those of QPSO, so it is feasible and effective to introduce chaos theory and mutation operation into QPSO.