System identification
For the most of design tasks of civil structures dynamic system properties must be identified. The field of system identification offers possibilities to determine these properties for existing structures via vibration measurements at the relevant degrees of freedom. Assuming a time invariant structural behavior, i.e. linear material behavior, the system parameters can be detected in a certain timeframe. There are many different possible methods, which can be applied for this purpose. The Kalman Filter (KF) is also one of these methods.
The application of KF on the measurement signal yields the linear system properties of the system. However, especially during strong dynamic excitations, such as storms and earthquakes, most of the structures response nonlinearly.
In this context, the chair focuses on the time variant system parameter identification for structures with semi-active dampers. For this purpose, we are investigating a modified version of KF, the unscented Kalman Filter (UKF), which covers also the nonlinear structural behavior.
It is necessary for semi-active damper systems to have a continuous parameter update of the damping values and the natural frequencies to guarantee the best performance. Otherwise, even small deviations of the natural frequencies will decrease the damping efficiency significantly. Due to continuous parameter updating the developed system identification method aims a higher efficiency of the damper.