A particle swarm optimization for high-dimensional function optimization

Particle swarm optimization (PSO) has received increasing interest from the optimization community due to its simplicity in implementation and its inexpensive computational cost. However, PSO face a common problem of premature convergence or stagnation in high-dimensional functions or complex multimodal functions. This paper proposes a modified PSO with two techniques: a mutation operator to increase swarm diversity for high-dimensionality; and an improved mechanism to detect and resolve the stagnation once it is found. The effectiveness of the proposed schemes is investigated on two widely-used PSO models: constriction factor and time-varying coefficients. The experimentation is performed using six wellknown benchmark functions of 30- and 100-dimensions with asymmetric initialization which is widely known to be difficult for most PSO variants.