Real-time hybrid testing for active mass driver system based on an improved control strategy against control-structure-interaction effects

Abstract

General real-time hybrid testing (RTHT) for active mass driver (AMD) system requires shake table to achieve boundary conditions between the numerical and experimental substructures, which may be unavailable in some situations. Simplified RTHT is free of shake table, but it cannot consider control-structure-interaction (CSI) effects, and may result in an inaccurate performance evaluation of AMD system due to a significant influence of CSI on control performance. In this study, to eliminate the CSI effects, an improved control strategy is proposed, by incorporating the acceleration of control floor into the AMD controller and transforming all the control commands and feedbacks from the geocentric coordinate system to the relative coordinate system. Based on this strategy, the implementations of general and simplified RTHTs for AMD system are described. Experimental investigations are conducted on a two-story structure equipped with an AMD. It is confirmed that the general RTHT can be used to evaluate the actual performances of different control strategies. The CSI effects are essentially eliminated by the improved control strategy, but adversely affect the performance of traditional control strategy. In addition, the simplified RTHT is inappropriate for the performance evaluation of the traditional control strategy, but is available to that of the improved control strategy. Finally, an AMD control scheme is presented for a seismically excited 20-story benchmark building, and is evaluated through the simplified RTHT. Experimental results demonstrate the effectiveness of the proposed AMD control scheme. The maximum roof displacement and acceleration decrease from 0.175 m and 3.059 m/s2 to 0.116 m and 1.356 m/s2, respectively, when subjected to the El Centro earthquake.