Journal of Shanghai Jiaotong University ›› 2018, Vol. 52 ›› Issue (2): 169-175.doi: 10.16183/j.cnki.jsjtu.2018.02.007

Previous Articles     Next Articles

Lightweight Approximation of Hull Surface Based on Non-Uniform Rational B-Spline

ZHANG Yanrua,LIN Yana,b,LU Conghonga,JI Zhuoshanga   

  1. a. School of Naval Architecture and Ocean Engineering; b. State Key Laboratory of Structural Analysis for Industrial Equipment, Dalian University of Technology, Dalian 116085, Liaoning, China
  • Online:2018-03-01 Published:2018-03-01

Abstract: During the processes of digital ship design, there are too many control points on the hull surface based on interpolation section curves, which are disadvantageous to the hull surface fairing and modification. The shape of hull surface, based on non-uniform rational B-spline section curves, is uncontrollable. To solve the forementioned problems, a design of lightweight approach of hull surface approximation is proposed. The hull surface is generated by two steps of section curves approximation. In step one, the equivalent chord deviation curve discretization method is applied to the curve approximation. In step two, the optimization approximation model is constructed with the knots of section curves as the design vari-ables. The objective function is set to minimize the sum of the approximate error of all section curves in the least-squares sense. According to the character of this problem, the genetic algorithm with adaptively alterable chromosome length is improved and applied to solve this optimization problem. The instances of the hull surface approximation and design for full-scale ship indicate that it is feasible and can satisfy the requirements of the engineering design for hull surface approximation. Applying this method, the data used for hull surface design and approximation can be reduced. And it can provide a reference for the lightweight design of other similar complex surface.

Key words: hull surface, rational section curves, non-uniform rational B-spline (NURBS), equivalent chord deviation, genetic algorithm

CLC Number: