VIBRATION CONTROLLER DESIGN METHOD OF MDOF SYSTEM USING OPTIMIZATION ALGORITHM

Abstract

This paper discusses on designing of vibration control system using evolutionary algorithms based on Particle Swarm Optimization (PSO). First, a new modified PSO strategy by adding a mutation rule (MPSO) is shown. Search efficiencies with Genetic Algorithms, original PSO and proposed MPSO are compared by solving standard benchmark functions. MPSO shows the excellent performance to optimize parameter of these functions, even it has multi-modal or parameters with epistasis. Second, structural system identification problems using MPSO are discussed. Structural system represented with product of biquad transfer functions can be identified by MPSO strategy. Finally, MPSO is applied to design MDOF vibration controller. The control performance by MPSO controller is described in both simulation and experiment to reduce floor vibration using an active mass damper.