Quantum inspired particle swarm optimization (QPSO) is widely used global convergence algorithm for complex design problems. But it may trap into local optima due to premature convergence because of insufficient diversity at the later stage of search process. In this regard, to intensify the QPSO performance in preventing premature convergence to local optima. This work presents a novel QPSO approach using student t probability distribution method with mutation operator on particle with global best position. In addition, a new dynamic control parameter is proposed to tradeoff between the exploration and exploitation searches. The proposed method will intensify the improvement in its convergence behavior and solution quality. The proposed improve QPSO called IQPSO is tested on an electromagnetic design problem namely, the TEAM workshop benchmark problem 22. The experimental results showcase the merit and efficiency of the proposed method.
展开▼
