集装箱船主尺度全局最优化的混沌算法

Global Optimization of Container Ships' Principal Parameters based on Chaos Optimization Algorithm

提出了应用混沌优化方法 ( Chaos Optimization Algorithm,COA)进行集装箱船船型主尺度要素全局优化的新策略。混沌优化方法利用混沌变量的随机性、规律性、遍历性寻优 ,能够克服经典优化方法如直接法、梯度法、Hessian法等方法容易陷入局部极小点的不足 ,方法简单、快速、易于掌握 ,其效率比一些目前广泛应用的随机优化方法如模拟退火法 ( SAA)、遗传算法 ( GA)等高得多。实际算例的结果证实了混沌优化方法用于集装箱船船型优化的有效性。

A new method for the global optimization of container ships' principal parameters is proposed by applying the Chaos Optimization Algorithm (COA), which is of ergodicity, stochastic property and regularity of chaotic motion. It calculates the value of objective function of each candidate solution to evaluate their optimum without any special requests, so can avoid the local optimal points. Moreover, this method is simple and fast. The efficiency of COA is much higher than some stochastic research algorithms such as Simulation Anneal Algorithm (SAA) and Genetic Algorithm (GA). Instance results show that chaotic optimization is effective for the global optimization of container ships' principal parameters, which is a complex nonlinear optimal problem. The minimum deviation method is adopted to establish the multi objective optical mathematical models for container ships' principal paraneters. The method and the COA are united to carry out respectively the single objective and multi objective optimization. The optimum results of COA vs GA are presented, which show that COA has less calculation, higher velocity and higher precision, and confirms that COA is effective and reliable.