Wave Force Analysis Using System Identification

摘要

A review of work on improving the predictive capabilities of the Morison equationfor wave force analysis in both the time and frequency domain using system identification is presented. System identification requires samples of time data for force and flow velocity; data from a U tube, the De Voorst wave flume, and the Cristchurch Bay tower, is considered. The results obtained in the time domain include an improved classification of wave forces and excellent predictions of measured forces although the model structures are case specific. A series expansion for the force spectrum is proposed via the Morison-Duffing equation which promises more accurate modeling of high frequency components in the force histories for general application. This has implications for fatigue analysis. The use of neural networks to model dynamical systems is discussed.