DECENTRALIZED SEMI-ACTIVE CONTROL FOR MULTI- PERFORMANCE-BASED DESIGN

摘要

Traditional performance-based design (PBD) that has a single performance level has been widely researched by changing section sizes of structural members or material properties to resist single hazard levels. However, this approach has limitations in terms of achieving performance and alternative design options for the owner. To overcome these limitations of the traditional PBD method, a multi-performance-based control design (MPBCD) methodology is newly proposed. The MPBCD integrates a decentralized semi-active control algorithm with semi-active smart damping devices and an advanced multi-objective optimization method. The multi-objective optimization is used to achieve various sets of performance-based control designs. The control designs satisfy multiple performance levels under multiple hazard levels without changing cross-section sizes or material properties of structural members. This MPBCD provides multiple sets of control designs (i.e., control device layouts with control design variables) to minimize design costs and maximize control effectiveness. The multiple sets of designs offer optimal performance-based control design covering a broad range of hazard levels with various performance levels. This numerical study uses an advanced decentralized semi-active controller and large-scale 200-kN magnetorheological (MR) dampers installed in a nine-story moment-resisting frame (MRF) building. From the multi-objective optimization technique, multiple layouts of control devices and controller parameters for multiple performance levels under multiple hazard levels are investigated.