Using vector operations to identify niches for particle swarm optimization
Isabella Schoeman, Andries Petrus Engelbrecht · 2005
In some problems described by objective functions, it is important to find all optimal solutions in the problem space. The particle swarm optimizer has originally been designed to locate a single optimum and avoid premature convergence on suboptimal solutions. To locate all optimal solutions, niches have to be identified containing good candidate solutions. In this paper a technique is introduced using vector operations to demarcate the boundaries of those niches and maintain subswarms in order to find optimal solutions.