Improved particle swarm optimization with wavelet-based mutation operation

Abstract

An improved wavelet-based mutation particle swarm optimization (IWMPSO) algorithm is proposed in this paper in order to overcome the classic PSO's drawbacks such as the premature convergence and the low convergence speed. The IWMPSO introduces a wavelet-based mutation operator first and then the mutated particle replaces a selected particle with a small probability. The numerical experimental results on benchmark test functions show that the performance of the IWMPSO algorithm is superior to that of the other PSOs in references in terms of the convergence precision, convergence rate and stability. Moreover, a pattern synthesis of linear antennas array is implemented successfully using the algorithm. It further demonstrates the effectiveness of the IWMPSO algorithm. © 2012 Springer-Verlag.