Acceleration Feedback Strategies for Active and Semiactive Control Systems: Modeling, Algorithm Development, and Experimental Verification
Licensed under
Category
Published on
Abstract
Most of the current research on active structural control for aseismic protection has been based on either full state feedback strategies or velocity feedback strategies. However, accurate measurement of the necessary displacements and velocities of the structure is difficult to achieve directly, particularly during seismic activity. Because accelerometers are inexpensive and can readily provide reliable measurement of the structural accelerations, development of control methods based on acceleration feedback is an ideal solution to this problem.
The focus of this dissertation is the development and experimental verification of acceleration feedback strategies for seismically excited structures. Both active and semiactive control systems are considered. Three different active control configurations are considered, including an active bracing system, an active tendon system, and an active mass driver system. The system identification procedure used in these studies is presented, and the effects of control-structure interaction are incorporated into the models. H2/LQG control strategies are applied to design the control systems. In addition to the active control experiments, one semi-active system employing a promising new semi-active device known as a magnetorheological (MR) damper is studied. The device is employed to control a three-story test structure. A phenomenological model that is based on a Bouc-Wen hysteresis model is proposed and shown to effectively portray the behavior of a typical MR damper. System identification techniques and acceleration feedback control strategies which are appropriate for semi-active control systems are developed and applied. The studies discussed herein demonstrate that acceleration feedback control strategies are effective and practically implementable for active and semi-active structural control applications.