Various semi-active control methods have been proposed for vibration control of civil structures. In contrast to active vibration control systems, all semi-active control systems are essentially asymptotically stable because of the stability of the structural systems themselves (with structural damping) and the energy dissipating nature of the semi-active control law. In this study, by utilizing the above property on the stability of semi-active control systems, a reduced-order structural model and a semi-active control law are simultaneously obtained so that the performance of the resulting semi-active control system becomes good. Based on the above fact any semi-active control laws derived from some models stabilize all real-existing structural systems that have structural damping. It means that the difference of dynamic behaviors between the real structural system and the reduced-order mathematical model in the sense of the open-loop response is no longer an important issue. In other words, we do not have to consider the closed-loop stability, which is one of the most important constraints in active control, in the process of the reduced-order structural modeling and the semi-active control design. We can only focus on the control performance of the closed-loop system with the real structure with the (model-based) semi-active control law in obtaining the reduced-order model. The semi-active control law in the present study is based on the one step ahead prediction of the structural response. The Genetic Algorithm (GA) is adopted to obtain the reduced-order model and the semi-active control law based on the reduced order model.
An Auto-Pilot System Design Using Reduced Order Model Reference Adaptive Control
