Particle swarm optimisation algorithm with iterative improvement strategy for multi-dimensional function optimisation problems
Yiwen Zhong, Xian Liu, Lijin Wang, Changying Wang · International Journal of Innovative Computing and Applications · 2012
Classical particle swarm optimisation algorithms update the velocity and position of all dimensions at each iteration step, and accept the updated velocity and position unconditionally. For multi-dimensional function optimisation problems, this strategy deteriorates the intensification ability of algorithm because different dimensions may interfere with each other. To deal with this shortage, this paper presents a particle swarm optimisation algorithm with iterative improvement strategy for multi-dimensional function optimisation problems. In the process of search, each particle updates velocity and position dimension by dimension, and evaluates position dimension by dimension. In each dimension, if the updated value can improve the solution, it will be accepted. Otherwise, the original value is retained. In order to keep algorithm from premature stagnation, particle will accept the velocity and position, which are calculated using classical methods, if there is no improvement found in any dimension. The experiments, which were carried on benchmark functions, showed that the iterative improvement strategy can improve the performance of particle swarm optimisation algorithm remarkably.