B-Spline Curve Modelling Based on Nature Inspired Algorithms

Norazwin Buang, Siti Aishah Hanawi, Hazura Mohamed, Ruzzakiah Jenal
dx.doi.org/10.17576/apjitm-2016-0501-03

Abstract

Curve modelling is an essential aspect in the field of geometric modelling. Conventional curve modelling generates large errors due to the number of data points which is more than the number of control polygon vertices. This study focuses on modelling the B-spline curve. Knot vector optimization method based on nature inspired algorithms is used to model B-spline curve with minimum error. The three nature inspired algorithms are particle swarm optimization, gravity search algorithm and harmony search algorithm. The result shows that the particle swarms optimization model the curve more accurate than the gravity search algorithm and harmony search algorithm. The study also demonstrates that by increasing the number of control polygon vertices, the error will be reduced. Hence the B-spline curve can be modelled accurately by minimum error.

keyword

B-spline, curve, particle swarm optimization, gravity search algorithm and harmony search algorithm.

Area

Industrial Informatics