摘要
本文采用多目标遗传算法求解船舶出坞拖轮配置优化模型的近似Pareto最优解集。针对遗传算法只能优化单目标、无约束问题,作者通过引入罚函数除去了船舶出坞中多变量非线性有约束问题的约束限制。文中利用加权和方法处理船舶出坞拖轮配置多目标优化问题,并按专家决策方法确定各目标函数的权系数。拖轮配置优化结果表明了多目标遗传算法的有效性,并为船坞主管进行船舶出坞拖轮配置决策提供了充分的依据。
A Pareto aggregate of tugs configuration is resolved by mutli-objective optimization GA in this paper. Authors make use of penalty function to remove restricitve condition of mutli-variable nonlinear constraint problem. The object function is treated with adding balance coefficient and summing. And these balance coefficients are confirmed by expert decision-making. The results of tugs configuration optimization express validity of multi-objective optimization GA. The optimization results also provide a foundation for dock governor decision-making of tugs configuration.
出处
《中国造船》
EI
CSCD
北大核心
2005年第2期69-74,共6页
Shipbuilding of China
