Comparative study of Model Order Reduction using combination of PSO with conventional reduction techniques

Recently, Particle Swarm Optimization has evolved as an effective global optimization algorithm whose dynamics has been inspired from swarming or collaborative behavior of biological populations. In this paper, two combinations of PSO and conventional methods of order reduction namely; PSO with Pole clustering and PSO with Dominant pole; have been applied for Model Order Reduction to find reduced order model by minimization of error between the step responses of higher and reduced order model. MOR using PSO algorithm is advantageous due to ease in implementation, higher accuracy and decreased time of computation. The third order reduced transfer functions of Triple Link Inverted Pendulum system evaluated by different methods have been compared.