Automated Substructuring of Adaptive High-Rise Structures for Decentralized Control

Adaptive structures offer a way to reduce material consumption in construction by actively counteracting external disturbances. They comprise a number of sensors, a control unit and actuators, resulting in a strongly coupled multiple-input-multiple-output system. The implementation of a centralized controller is computationally expensive and leads to high wiring effort and latencies in the control loop. Thus, this paper presents an algorithm to automatically substructure adaptive high-rise buildings by evaluating the input/output (I/0)-couplings of the system. The number, size, and composition of the substructures are investigated based on two interaction measures: the Dynamic Relative Gain Array, which focuses on the I/O-coupling in the frequency domain and the Participation Matrix, which investigates I/O-interactions in the time domain. The effectiveness of the different sub-structurings is compared to a state-of-the-art configuration by designing a decentralized control strategy based on Model Predictive Control.