@inproceedings{Ge:2016:10.1109/CONTROL.2016.7737594,
author = {Ge, M and Kerrigan, EC},
doi = {10.1109/CONTROL.2016.7737594},
publisher = {IEEE},
title = {Short-term ocean wave forecasting using an autoregressive moving average model},
url = {http://dx.doi.org/10.1109/CONTROL.2016.7737594},
year = {2016}
}

Download

TY  - CPAPER
AB  - In order to predict future observations of a noisedrivensystem, we have to find a model that exactly or atleast approximately describes the behavior of the system sothat the current system state can be recovered from pastobservations. However, sometimes it is very difficult to modela system accurately, such as real ocean waves. It is thereforeparticularly interesting to analyze ocean wave properties inthe time-domain using autoregressive moving average (ARMA)models. Two ARMA/AR based models and their equivalent statespace representations will be used for predicting future oceanwave elevations, where unknown parameters will be determinedusing linear least squares and auto-covariance least squaresalgorithms. Compared to existing wave prediction methods, inthis paper (i) an ARMA model is used to enhance the predictionperformance, (ii) noise covariances in the ARMA/AR model arecomputed rather than guessed and (iii) we show that, in practice,low pass filtering of historical wave data does not improve theforecasting results.
AU  - Ge,M
AU  - Kerrigan,EC
DO  - 10.1109/CONTROL.2016.7737594
PB  - IEEE
PY  - 2016///
TI  - Short-term ocean wave forecasting using an autoregressive moving average model
UR  - http://dx.doi.org/10.1109/CONTROL.2016.7737594
UR  - http://hdl.handle.net/10044/1/40562
ER  - 

Download