Identification of nonlinear wave forces. Part 1: Time domainanalysis
Abstract
The Nonlinear Autoregressive Moving Average model with exoneous inputs (NARMAX) system identification technique is applied to a number of fluid loading data sets. The aim is to determine if there exists a simple extension of Morison's equation to predict forces with improved accuracy. In part 1, concerned with time domain results, a number of NARMAX time series models are obtained for a range of situations of wave loading on slender cylinders. An introduction to system identification using the NARMAX nonlinear time series analysis procedures is provided. The basic theory required for any parameter estimation problem is summarized. The problem of determining the appropriate form for the system model is discussed. The model validation is presented. A simulation of a Morison type system is used to demonstrate the utility of the NARMAX procedures in analyzing fluid loading data. The results of applying the techniques to experimental data measured in a Utube are discussed. Unidirectional random waves in a large flume are analyzed. Data from a directional random sea is discussed.
 Publication:

NASA STI/Recon Technical Report N
 Pub Date:
 October 1991
 Bibcode:
 1991STIN...9318510W
 Keywords:

 Autoregressive Processes;
 Computational Fluid Dynamics;
 Data Sampling;
 Fluid Pressure;
 Nonlinear Systems;
 Parameter Identification;
 Wave Propagation;
 Algorithms;
 Fluid Flow;
 Frequency Ranges;
 Hydrodynamics;
 Orthogonal Functions;
 Time Functions;
 Fluid Mechanics and Heat Transfer