Particle swarm optimization of a non-collocated MIMO PPF active vibration control of a composite sandwich plate

Abstract

The Particle Swarm Optimization (PSO) algorithm is employed to tune the Multiple-Input Multiple-Output (MIMO) Positive Position Feedback (PPF) controller to reduce vibrations on a composite sandwich plate with free boundary conditions, equipped with piezoelectric actuators and sensors. The PSO algorithm determines the value of forty-four parameters describing the transfer functions of the MIMO PPF controllers. The solution found by the PSO code depends on the cost-function employed. In this work, two cost functions are investigated. The first function minimizes the time necessary for free vibrations to decay to zero, while the second function minimizes the area under the resonance peaks of the frequency response of the controlled plate. For both cost-function cases, the parameters found by the PSO algorithm are determined through simulations based on a reduced-order model of the structure. Once the parameters are determined, they are loaded on a real-time prototyping board and employed to dampen the vibrations of the sandwich plate tested in laboratory experiments. The compatibility of simulations and experiments was verified at various points on the surface of the plate, and vibration reductions up to 90% of the uncontrolled amplitude were achieved experimentally for the first eight modes. The absence of spillover on following modes was verified in simulations and experiments.