基于GA优化的灰色神经网络船舶交通流量预测方法研究  

Prediction Method of Ship Flow Based on Grey-neural Network Improved by GA

作　　者：李俊[1] 徐志京[1] 唐贝贝[1] 

机构地区：[1]上海海事大学信息工程学院,上海201306

出　　处：《船海工程》2013年第5期135-137,共3页Ship & Ocean Engineering

基　　金：上海海事大学校研基金项目(20120108)

摘　　要：结合灰色模型和BP神经网络模型的特点,对两种模型进行有机地组合,构建一种改进的灰色神经网络预测船舶流量方法。以实际船舶交通流量和主要影响因素为数据,运用遗传算法改进的灰色神经网络模型对上海洋山港的船舶交通流量进行预测,计算和Matlab仿真结果表明,改进的灰色神经网络模型预测不仅精度较高,而且能准确预测船舶交通流量的变化规律。Combine with the characteristics of grey model and BP neural network model, an improved method of grey neural network was proposed to the predicted of ship flow. In terms of actual ship flow and the main influence factors, the ship flow of Yangshan port can be predicted by the grey neural network model improved by the genetic algorithm. The calculation and Matlab simulation results showed that the improved grey neural network prediction model can predict the variation of vessel traffic flow with high precision.

关 键 词：船舶流量 灰色神经网络 遗传进化算法 预测系统 

分 类 号：U69[交通运输工程—港口、海岸及近海工程]