摘要
将人工神经网络理论引入了船舶油污损失赔偿额的估算中,通过从历史船舶溢油事故案例库中选取了十组样本数,以污染损失赔偿额与影响因子集作为训练样本,建立了船舶溢油事故污染损失额估算的人工神经网络模型。并对人工神经网络模型进行仿真训练学习,并将训练好的神经网络应用于对相关案例的估算。研究结果表明应用人工神经网络方法进行溢油损失估算结果客观、可靠,为船舶油污索赔和环境保护提供了有力的理论依据。
An artificial neural network approach was proposed to estimate the loss of oil pollution because of ship＇s accidents. The artificial neural network model founded based on training patterns including damage value and its influenced factors by selecting ten cases from historical database. Then the emulation and training for the network were proceed with, and the parameters derived from the trained network could be used to evaluate relevant cases. Results show that the artificial neural network approach for the damage estimation of ship＇s oil pollution has some advantages such as objective and reliability, and they can provide important proofs for claiming for ship＇s oil pollution compensation and protecting environment.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2006年第2期475-477,481,共4页
Journal of System Simulation
基金
国家自然科学基金(70273019)
上海市教委基金(2004118)
关键词
船舶油污损失
人工神经网络
仿真训练
应用估算
ship＇s oil pollution damage
artificial neural network
emulation training
estimation
