模拟退火算法具有强大的全局搜索能力,通过改进可用于求解多目标优化问题的Pareto前沿。文章针对核动力航母早期设计阶段方案生成问题,其中包含16个主尺度要素和14个目标函数,以效能最高和费用最低作为两个总的优化目标,建立了效能-费用双目标优化模型,并应用改进的多目标模拟退火算法进行迭代求解。结果表明,本方法快捷可靠,可在短时间内生成大量备选设计方案,为最终设计方案的确定提供充足的选择余地。
Simulated annealing algorithm can be easily adopted to calculate the Pareto frontier. There are 16 principal parameters and 14 objective functions and the 2 overall objects in the conceptual design model of a nuclear aircraft carrier (CVN). The 2-objective optimization model with efficiency and cost is established and the multi-objective simulated annealing algorithm is applied to calculating the solutions. The result shows that this method can generate large number of alternative designs and provides abundant choice for the determination of the final design.