Record Details

Estimating transmitted waves of floating breakwater using support vector regression model

DRS at CSIR-National Institute of Oceanography

View Archive Info
 
 
Field Value
 
Title Estimating transmitted waves of floating breakwater using support vector regression model
 
Creator Mandal, S.
Hegde, A.V.
Kumar, V.
Patil, S.G.
 
Subject floating breakwater
vector regression model
wave energy
 
Description Planning and design of coastal protection works like floating breakwaters requires information about the performance characteristics of the structure in reducing the wave energy. They can be used as a temporary protection for offshore activities in hostile environment during construction, drilling works, salvage operation etc. The design involves hydraulic as well as structural aspects. The hydraulic design consists of suitable type, cross section and dimension of the breakwater to achieve the desired degree of wave attenuation. The structural design involves the design of various elements of the breakwater and checking the stability of the structure against possible modes of failure. The prime factor in the construction of floating breakwater is to obtain the optimum width of the breakwater (relative breakwater width) along the direction of wave propagation. In this present work the performance of support vector regression model for predicting the wave transmission coefficient of floating pipe breakwater is investigated. The data used in the model are the experimental parameters obtained from the work conducted in the wave flume at Marine Structure laboratory, NIT Karnataka, India. Based on the concept of energy dissipation through the pipe network, a floating breakwater of simple geometry with the arrangements of the structure in its horizontal plane and with simple geometry is selected for studying the wave transmission characteristics. The input non-dimensional parameters that significantly influence the wave transmission of the floating breakwater are considered as wave steepness (Hi/L, Hi- wave height, L-wave length), relative wave height (Hi/d, d- water depth), relative width of the breakwater (W/L, W- breakwater width) and ratio of spacing to diameter of pipes (S/D). The radial basis functions performed well than the polynomial function in the regressive support vector machine as the kernel function for the given set of data. The support vector regression model gives the correlation coefficients of 0.968 for training and 0.961 for testing.
 
Date 2010-05-11T10:05:43Z
2010-05-11T10:05:43Z
2009
 
Type Conference Article
 
Identifier Proceedings of 3rd International Congress on Computational Mechanics and Simulation (ICCMS - 09). eds. by: Kant, T.; Eldho, T.I.; Banerjee, S. Indian Institute of Technology Bombay; Powai, Mumbai; India; 2009; 173-174
http://drs.nio.org/drs/handle/2264/3613
 
Language en
 
Rights Copyright [2009]. All efforts have been made to respect the copyright to the best of our
knowledge. Inadvertent omissions, if brought to our notice, stand for correction and withdrawal of document from this repository.
 
Publisher IIT Bombay