Record Details

Wave Prediction Using Genetic Programming and Model Trees

DSpace at IIT Bombay

View Archive Info
 
 
Field Value
 
Title Wave Prediction Using Genetic Programming and Model Trees
 
Creator RAMBEKAR, AR
DEO, MC
 
Subject Wave prediction
genetic programming
model trees
wave buoys
meteorological buoys
ARTIFICIAL NEURAL-NETWORKS
WIND
PARAMETERS
 
Description KAMBEKAR, A.R. and DEO, MC., 2012. Wave prediction using genetic programming and model trees. Journal of Coastal Research, 28(1), 43-50. West Palm Beach (Florida), ISSN 0749-0208. Wave predictions at specified locations can some times be conveniently made using empirical methods in preference to numerical ones in view of the empirical method's computational efficiency and lack of requirement for any exogenous information. This article describes one such attempt in which the time series prediction of waves was made through two alternative data driven methods: genetic programming and model trees. The significant wave height, H-s, and average zero-cross wave period, T-z, over different time steps in the future have been modeled as functions of a preceding sequence of the causal wind vector. The unknown functions have been captured in computer programs yielded by the genetic programming technique and also alternatively in the set of linear models produced by the technique of model trees. The measurements made by wave rider buoys at two locations along the Indian coastline have been analyzed to build and validate the models. Predictions have been made over the intervals of 1 to 4 days in the future based on a sequence of preceding wind vectors going back in time by the similar period as the prediction horizon. Applicability of such predictions at a nearby observation station is also explored. Although the overall performance of genetic programming was more attractive than that of model trees, both approaches performed well during the model testing exercise as evidenced by the high values of the correlation coefficient and the coefficient of efficiency and the low values of the mean absolute error, root mean square error, and scatter index.
 
Publisher COASTAL EDUCATION & RESEARCH FOUNDATION
 
Date 2014-10-15T12:11:03Z
2014-10-15T12:11:03Z
2012
 
Type Article
 
Identifier JOURNAL OF COASTAL RESEARCH, 28(1)43-50
http://dx.doi.org/10.2112/JCOASTRES-D-10-00052.1
http://dspace.library.iitb.ac.in/jspui/handle/100/14871
 
Language en