集装箱船舶配载方案评价的Hopfield神经网络模型  

Hopfield neural network model of evaluating for container ship stowage plans

作　　者：张维英[1] 林焰[1] 纪卓尚[1] 邓林义[1] 

机构地区：[1]大连理工大学船舶CAD工程中心

出　　处：《大连海事大学学报》2005年第3期13-16,共4页Journal of Dalian Maritime University

基　　金：高等学校博士学科点专项科研基金资助项目(2000014125)

摘　　要：将人工神经网络技术应用于集装箱船舶配载方案评价,建立了集装箱船舶配载方案评价的Hopfield神经网络模型.该模型首先建立配载方案评价指标与标准类别体系(即配载方案的级别,如优,良,…),以标准类别为记忆模式,将待评价方案作为联想回忆的噪声模式,当网络达到平衡点时其lyapunov函数达到极小值,对应待评价方案的类别等级.此方法为集装箱船舶配载方案优选提供了一种实用的模型.The stowage is the pivotal link in container transportation. And optimization of loading plan plays an important role in guarding the security of ships and cargoes and in improving the efficiency of transportation. This paper applied the Hopfield neural network to evaluating the stowage plan for container ship, and built the evaluation model. The model built a system of evaluation index and standard category and using standard category as anamnestic pattern. The system took schemes of candidate as its noisy pattern, the lyapunov function gets infinitesimal value when the network reached its balance point, which corresponded to the classification of scheme candidated. This paper presented an application model in the optimization of the scheme of stowage.

关 键 词：集装箱船 配载方案评价模型 HOPFIELD神经网络 

分 类 号：U695.22[交通运输工程—港口、海岸及近海工程] TP391[交通运输工程—船舶与海洋工程]