Real-time hybrid simulation (RTHS) is increasingly being recognized as a powerful cyber-physical technique that offers the opportunity for system evaluation of civil structures subject to extreme dynamic loading. Advances in this field are enabling researchers to evaluate new structural components/systems in cost-effective and efficient ways, under more realistic conditions. For RTHS, performance metric clearly needs to be developed to predict and evaluate the accuracy of various partitioning choices while incorporating the dynamics of the transfer system, and computational/communication delays. In addition, because of the dynamics of the transfer system, communication delays, and computation delays, the RTHS equilibrium force at the interface between numerical and physical substructures is subject to phase discrepancy. Thus, the transfer system dynamics must be accommodated by appropriate actuator controllers. In this paper, a new performance indicator, predictive performance indicator (PPI), is proposed to assess the sensitivity of an RTHS configuration to any phase discrepancy resulting from transfer system dynamics and computational/communication delays. The predictive performance indicator provides a structural engineer with two sets of information as follows: (i) in the absence of a reference response, what is the level of fidelity of the RTHS response? and (ii) if needed, what partitioning adjustments can be made to effectively enhance the fidelity of the response? Moreover, along with the RTHS stability switch criterion, this performance metric may be used as an acceptance criteria for conducting single-degree-of-freedom RTHS.
In a real-time hybrid simulation, a transfer system is used to enforce the interface interaction between computational and physical substructures. A model-based, multilayer nonlinear control system is developed to accommodate extensive performance variations and uncertainties in a physical substructure. The aim of this work is to extend the application of real-time hybrid simulation to investigating failure, nonlinearity, and nonstationary behavior. This Self-tuning Robust Control System (SRCSys) consists of two layers: robustness and adaptation. The robustness layer synthesizes a nonlinear control law such that the closed-loop dynamics perform as intended under a broad range of parametric and nonparametric uncertainties. Sliding mode control is employed as the control scheme in this layer. Then, the adaptation layer reduces uncertainties at run time through slow and controlled learning of the control plant. The tracking performance of the SRCSys is evaluated in two experiments that have highly uncertain physical specimens.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.