摘要
GESA（GuidedEvolutionarySimulatedAnnealing）算法是一种采用全局优化策略的人工神经网络的并行技术，它以一种新颖的方式综合了遗传算法、模拟淬火、模拟进化的思想；它的另一个特点是在解空间实施区域引导（regionalguidance）。本文用GESA算法进行船舶性能优化计算，这种方法速度快捷，由于是全局优化策略，避免了陷入局部最小值，利用该方法，可向设计部门提供咨询服务。
GESA (Guided Evolutionary Simulated Annealing)algorithm is a artificial neural network technique which adopts global optirnal strategy , and it combines the ideals of genetic algorithm, simulated annealing and simulated evolution in a novel way. In the solution space,it carries on search of regional guidance,we demonstrate GESA technique using the optindzation of the performance of ship S60-AU5,the metliod takes much less time,because of global optimal strategy, it keeps out off trapping into local minimum, this technique can provide consult service for design units.
出处
《船舶力学》
EI
1997年第2期8-13,共6页
Journal of Ship Mechanics
关键词
人工神经网络
船舶
性能
设计
GESA算法
artificial neural network
parallel algorithm
genetic algorithm
simulated annealing
simulated evolution
regional guidance
