In this paper, the robust control of a real high-rise tower is studied, using a newly proposed, in the structural control field, Robust Model Predictive Control scheme (RMPC). Two RMPC controllers were designed considering either displacement mitigation (RMPC 1 ) or power consumption efficiency (RMPC 2 ).The two controllers were compared to the benchmark, robustness-wise, H ∞ control scheme to demonstrate their relative performance. A number of stiffness and damping uncertainty scenarios were designed based on a broad study of the relevant literature, in order to estimate the robustness of each of the three controllers. In all scenarios, variable actuator uncertainty of AE5% was introduced. It was found that all controllers are effective in controlling the tower and demonstrate robustness against parametric and actuator uncertainties with different relative merits over each other. Indicatively, when considering root-mean-square (RMS) and peak displacement and acceleration reduction, the H ∞ had an average performance reduction of 24%, the RMPC 1 31% and the RMPC 2 28% against their uncontrolled equivalent.

No abstract

This paper considers different vibration control options for a real high-rise tower subjected to real wind loading. To mitigate excessive responses, the tower utilizes a hybrid passive -active control system with a relatively small actuator capacity. Firstly, the methodology for establishing a reduced order numerical model of the tower from a finite element realization is presented. Subsequently, to deal with physical constraints, a Model Predictive Control (MPC) algorithm is employed and compared to a more conventional Linear Quadratic Regulator (LQR). In both cases, the algorithms managed to keep the dynamic displacements within the set desired limit. It was concluded that, the MPC had a better performance when compared to the LQR in terms of peak amplitudes at the expense though of energy.