Parameter optimization for enhanced modeling of dynamic systems

To ensure safety and continuous operation of dynamic systems, including civil structures, mechanical and aerospace systems, structural health monitoring and control systems aim to study and optimize parameters that govern their behavior. The development of any intelligent infrastructure systems must take into account the specific challenges of each system and provide customized solutions to address those challenges effectively. In this dissertation, a structural health monitoring system and a passive control system are explored. The structural health monitoring system is developed to estimate the behavior of a special group of systems experiencing large accelerations, higher than 100 $g$, for durations less than 100 ms called high-rate systems. The high-rate systems are characterized by 1) large uncertainties in their external loads, 2) high levels of nonstationarity and heavy disturbances, and 3) unmodeled dynamics from changes in system configuration. A neural-network structural health monitoring system is proposed and customized to address the challenges of such systems. The passive control system assesses the performance of a novel passive variable friction damper (PVFD) in achieving multi-objective vibration mitigation. A probabilistic performance-based design framework is presented to select the optimum parameters of the system. In order to effectively implement the systems of interest of this study, methods for effective system parameter selection is investigated. The algorithms presented allow for informed parameter selection in a large space of possible values. The methods have been verified on a number of case studies and have shown promise for effective monitoring and control