船舶交通量的BP神经网络-马尔科夫预测模型

发布时间：2018-03-24 16:27

本文选题：船舶交通量　切入点：BP神经网络　出处：《上海海事大学学报》2017年02期

【摘要】：为提高船舶交通量的预测精度,在BP神经网络的基础上结合马尔科夫预测模型建立一个新的预测模型.采用通过长江九江大桥的月度船舶交通量数据进行模型训练、验证和预测,求出相对残差值,将相对残差的前8项归一化后划分为3个状态,利用马尔科夫预测模型修正BP神经网络的预测值.该新模型将BP神经网络的相对残差值区间从[-12.9%,12.3%]降低至[-9.9%,5.4%].该模型能提高船舶交通量的预测精度,用于预测船舶交通量是可行的.
[Abstract]:In order to improve the prediction accuracy of ship traffic volume, a new prediction model is established based on BP neural network combined with Markov forecasting model. The model is trained by the monthly ship traffic volume data from Jiujiang Bridge of the Yangtze River. The first eight items of the relative residuals are normalized and divided into three states. The prediction value of BP neural network is modified by Markov forecasting model. The new model reduces the relative residual value of BP neural network from [-12.9%] to [-9.9%]. The model can improve the prediction accuracy of ship traffic volume and is feasible to predict ship traffic volume.
【作者单位】：武汉理工大学航运学院;武汉理工大学内河航运技术湖北省重点实验室;
【基金】：国家科技支撑计划(2015BAG20B01)
【分类号】：TP183;U692

【相似文献】