Nonlinear Robust Framework for Real-time Hybrid Simulation of Structural Systems: Design, Implementation, and Validation

Real-time hybrid simulation (RTHS) is a cyber-physical, time-efficient and cost-effective technique that integrates physical testing with computational simulation to offer powerful methods to examine structural behavior and seismic resilience at multiple scales. To enforce interface conditions between computational and physical substructures, usually, servo-hydraulic actuators serve as transfer system. In implementation of any RTHS, global stability and performance must be given appropriate attention. In this dissertation, the overarching research objective is to establish and validate a systematic approach to design a safe, stable and accurate RTHS considering the significance of partitioning configuration, transfer system dynamics, nonlinearity and uncertainty in the physical substructure. Transfer and sensing systems introduce de-synchronization at the interface. These added dynamics in the feedback loop can result in instabilities and losses in performance. In RTHS, each partitioning choice has a different set of stability requirements and as the partitioning choice becomes more complex, set of stability requirements can become extremely narrow. Thus, predictive indicators are developed which provide a relatively easy and useful method to examine various partitioning choices. To enhance understanding of transfer system dynamics, a nonlinear dynamical model is proposed for a servo-hydraulic actuator coupled with a nonlinear physical specimen. The nonlinear dynamical model is transformed into controllable canonical form for further tracking control purposes. Based on the dynamical model, a high-precision nonlinear control system is developed to achieve the interface conditions. Self-tuning Robust Control System is a multi-layer nonlinear control system designed to accommodate extensive performance variations in the physical substructure due to structural component failure, complexity, and nonstationary behavior