基于遗传算法的船体主甲板外展程度寻优

Main deck extending degree optimization based on GA method

阐述了利用遗传算法思想解决某些具有特殊甲板尺度要求的船舶甲板主尺度(长度、宽度)与船体设计水线长、宽比值的优化设计问题。在舯横剖面和中纵剖面上分别定义甲板宽度与设计水线宽之比α、甲板长度与设计水线长之比β,以该问题最为突出的大型舰船为例,建立最小偏差数学模型,对α、β优化确定问题进行计算,根据约束条件及优化对象的相对权重编制Matlab程序进行寻优,最终获得了最优化的甲板外展程度。整个过程采用二进制编码以增大搜索随机性,通过设定父辈交叉及基因突变概率,防止搜索过程出现局部最优。

The measure was analyzed about how to use GA method to solve the problem of rate between the deck scale （length and width） and design waterplane scale （length and width） on ship which has special demand. The rate between width of deck and width of design waterplane on section was defined as α, and the rate between length of deck and length of design waterplane on buttock as β. A big vessel was taken as an example in this aspect, and the least warp function was constituted,α and β were optimized under the constraints and relatively weights with Matlab, and the best deck extending degree was got at last. The binary system code was used to enlarge random search scope and the crossover and mutation rate were set to prevent local optimization during the whole process.