CAJ | 학술논문

船体结构之最佳化设计是一个复杂非线性的混和离散问题,并且要搜寻到全域的最佳值并不容易.在复杂的设计环境下基因演算法(Genetic Algorithm;GA)却可以搜寻到近似的全域最佳值.本文主要是应用基因演算法对T加强板架(Tee stiffened panel)、平板加强板架(flat-bat stiffened Panel)等常用且最具代表性之船体结构件进行最佳化设计,使结构在满足终极破坏限制(ultimate failure constraints)与耐用破坏限制(serviceability failure constraints)等所有设定限制条件下,求得最佳目标函数值中各设计变数之最佳组合.在分析过程中并考量不同族群大小、变换机率、突变机率等因素对最佳化结果的影响.文中是以制造成本为目标函数,其中同时考量材料成本及劳工成本,且所得之结果与连续性线性规则(Sequential Linear Programming;SLP)最佳化结果作了比较.计算的结果显示基因演算法可以有效地与快速地获得最小重量和最低成本的目标.
선체결구지최가화설계시일개복잡비선성적혼화리산문제,병차요수심도전역적최가치병불용역.재복잡적설계배경하기인연산법(Genetic Algorithm;GA)각가이수심도근사적전역최가치.본문주요시응용기인연산법대T가강판가(Tee stiffened panel)、평판가강판가(flat-bat stiffened Panel)등상용차최구대표성지선체결구건진행최가화설계,사결구재만족종겁파배한제(ultimate failure constraints)여내용파배한제(serviceability failure constraints)등소유설정한제조건하,구득최가목표함수치중각설계변수지최가조합.재분석과정중병고량불동족군대소、변환궤솔、돌변궤솔등인소대최가화결과적영향.문중시이제조성본위목표함수,기중동시고량재료성본급노공성본,차소득지결과여련속성선성규칙(Sequential Linear Programming;SLP)최가화결과작료비교.계산적결과현시기인연산법가이유효지여쾌속지획득최소중량화최저성본적목표.
Optimal design of ship structures is a complicated nonlinear mixed-discrete problem and a difficult challenge in searching for its global optimal values. Under complicated design environment, the stateof-art genetic algorithm (GA) can rapidly search for the approximate global optimum while the optimization problem involves discrete design variables. The purpose of this study is to obtain an optimal design by means of GA approach for two stiffened panel forms,namely tee stiffened and flat-bar stiffened panels that are often used in ship structures. The genetic parameters, such as population sizes, crossover probability and mutation probability, in the optimal design will carefully be examined for their influence. The least cost on panel production, including costs of materials and labors, is defined as an objective function, that is required to be minimized. The optimal value obtained for each design variable must satisfy the pre-described constraints and the fabrication restrictions.A comparison between the genetic algorithms and the sequential linear programming optimization algorithms is also made under similar constrains in this research. Computer simulation results reveal that the GA approach can achieve the goal of obtaining minimum weight and minimum cost efficiently and promptly.