Performance and robustness of hybrid model predictive control for controllable dampers in building models

Erik A. Johnson; Wael M. Elhaddad; Steven F. Wojtkiewicz

doi:10.1117/12.2218673

Translator Disclaimer

20 April 2016 Performance and robustness of hybrid model predictive control for controllable dampers in building models

Erik A. Johnson, Wael M. Elhaddad, Steven F. Wojtkiewicz

Author Affiliations +

Proceedings Volume 9803, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2016; 98032C (2016) https://doi.org/10.1117/12.2218673
Event: SPIE Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring, 2016, Las Vegas, Nevada, United States

Abstract

A variety of strategies have been developed over the past few decades to determine controllable damping device forces to mitigate the response of structures and mechanical systems to natural hazards and other excitations. These “smart” damping devices produce forces through passive means but have properties that can be controlled in real time, based on sensor measurements of response across the structure, to dramatically reduce structural motion by exploiting more than the local “information” that is available to purely passive devices. A common strategy is to design optimal damping forces using active control approaches and then try to reproduce those forces with the smart damper. However, these design forces, for some structures and performance objectives, may achieve high performance by selectively adding energy, which cannot be replicated by a controllable damping device, causing the smart damper performance to fall far short of what an active system would provide. The authors have recently demonstrated that a model predictive control strategy using hybrid system models, which utilize both continuous and binary states (the latter to capture the switching behavior between dissipative and non-dissipative forces), can provide reductions in structural response on the order of 50% relative to the conventional clipped-optimal design strategy. This paper explores the robustness of this newly proposed control strategy through evaluating controllable damper performance when the structure model differs from the nominal one used to design the damping strategy. Results from the application to a two-degree-of-freedom structure model confirms the robustness of the proposed strategy.

Citation Download Citation

Erik A. Johnson, Wael M. Elhaddad, and Steven F. Wojtkiewicz "Performance and robustness of hybrid model predictive control for controllable dampers in building models", Proc. SPIE 9803, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2016, 98032C (20 April 2016); https://doi.org/10.1117/12.2218673

ACCESS THE FULL ARTICLE