System identification of distributed-parameter marine riser models
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
Modeling engineering problems of interest often requires some type of discretization model of the physical system and quite naturally leads to a mathematical description involving partial differential equations whose coefficients are dependent on both time and spatial location. In this study a reverse system identification approach is presented that utilizes generalized coordinate and force functions to recover the value of the key system parameters for each mode of vibration. To illustrate the analysis procedures, a single marine riser with general damping-restoring types of non-linearities subject to random wave excitation is considered. Analytical expressions as functions of the modes for bending stiffness and tension are derived and used for comparison with the results obtained using system identification. Numerical simulations including band-limited white noise and random wave excitation are used to explore the adequacy of the methodology and the benefits of using modal analysis in the system identification procedure. Finally, the use of and comparison with experimental data is presented and the frequency variation of parameters obtained resulting from system identification procedures discussed. Collectively, the examples demonstrate that this system identification methodology accurately identifies system parameters over portions of the frequency range of interest. 2003 Elsevier Science Ltd. All rights reserved.
