Hierarchical model predictive vibration control for high‐rise structures using semi‐active tuned liquid column damper

Abstract

Vibration control for high-rise structures has become more important as buildings become more lightweight. Tuned liquid column dampers are an established approach for this problem, as they do not consume high amounts of energy while achieving high damping forces. The natural frequency of tuned liquid column dampers (TLCDs) can be changed semi-actively by modifying their column geometry. As shown by previous studies, this approach increases the control performance of the damper and allows a larger application field. However, this modification simultaneously provokes strong nonlinearities, which need to be addressed for controller design. In this contribution a model-based controller for such TLCDs is proposed, which explicitly accounts for the nonlinear dynamics of the coupled system consisting of a multi-storey structure with a semi-active TLCD. To account for various system dynamics of the structure, a hierarchical nonlinear model predictive controller (NMPC) in combination with a multi-start strategy is formulated. In addition, an extrapolation algorithm is developed to determine an approximation of the future trajectory of the structure's induced excitation load. The excitation trajectory is then incorporated in the NMPC framework. For validation, a frequency response analysis and a case study are conducted on a 20-storey benchmark structure, which is equipped with a semi-active TLCD. The study is carried out in simulation, and the results are compared with passive vibration control. With the proposed hierarchical control, a significant improvement is shown in the reduction of structural vibrations over the relevant spectrum of excitation frequencies up to a mean of 50%.