Rule-Based Control Algorithm for Active Tuned Mass Dampers

Abstract

A rule-based control algorithm for active tuned mass dampers (TMDs) is proposed. First, perturbation solutions of the linear quadratic regulator (LQR) feedback gains for the active TMD system are derived. Using these solutions, interaction of the TMD and the actuator force is discussed in detail. Then, an intelligent rule-based control scheme is constructed based on this discussion along with a previously developed semiactive control method. Fuzzy logic is incorporated with the rule bases to implement analytical and/or intuitive knowledge about the TMD into control rules. The algorithm consists of two parts: (1) a variable gain displacement feedback control; and (2) a variable TMD damping control. The first one is applied when the TMD amplitude is small to make the TMD more effective, and the second one is applied when the TMD amplitude is large to dissipate the energy. Treatment of constraints on control force capacity and TMD displacement is emphasized in construction of the controller. Performance of the proposed method is demonstrated by numerical studies of continuous structures with widely spaced natural frequencies subject to seismic excitation.