基于优化RBF网络的港口船舶交通流量预测

发布时间：2018-01-20 11:02

本文关键词： 水路运输船舶交通流量 RBF神经网络遗传算法港口预测　出处：《中国航海》2014年02期 　论文类型：期刊论文

【摘要】：港口船舶交通流量预测能为港口规划、交通管理提供决策支持。RBF神经网络在交通流预测领域有着广泛的应用,但其在网络权值等参数的选取算法上存在缺陷。遗传算法具有全局搜索速度快的优点,利用该算法对RBF神经网络的权值进行遗传操作,可获得具有一定遍历性的初始权值。文章尝试将基于遗传算法优化的RBF神经网络应用到港口船舶交通流量预测领域并以芜湖港为例进行验证。结果显示,优化后的RBF神经网络的预测误差比普通的RBF神经网络小5%左右,表明优化后的RBF神经网络计算量更小、识别速度更快、预测误差更小,在港口船舶交通流量预测领域具有广阔的应用前景。
[Abstract]:Port vessel traffic flow forecasting can provide decision support for port planning and traffic management. RBF neural network has a wide range of applications in the field of traffic flow forecasting. The genetic algorithm has the advantage of fast global search speed. The genetic algorithm is used to operate the weight of RBF neural network. The initial weights with certain ergodicity can be obtained. This paper attempts to apply the RBF neural network based on genetic algorithm optimization to the field of port ship traffic flow prediction and take Wuhu Port as an example to verify the results. The prediction error of the optimized RBF neural network is about 5% less than that of the ordinary RBF neural network, which indicates that the optimized RBF neural network has less computation, faster recognition speed and smaller prediction error. It has a broad application prospect in the field of port ship traffic flow prediction.
【作者单位】：武汉理工大学航运学院;内河航运技术湖北省重点实验室;
【分类号】：U692
【正文快照】： 港口是水陆交通的集结点和枢纽,对区域经济发展具有十分重要的作用。港口船舶交通流量预测能够为港口规划、交通管理提供决策支持。回归预测法、时间序列预测法、灰色理论预测法在船舶交通流量预测方面存在局限性。[1]神经网络因其简单的网络结构、快速的学习方法、较好的推广

【相似文献】