摘要
远海航道船舶流量受到多种因素的综合作用,具有一定的周期性,同时具有强烈的非线性,传统线性建模方法无法对远海航道船舶流量进行高精度的拟合,使得远海航道船舶流量预测偏差大。为了克服当前远海航道船舶流量预测系统存在的局限性,设计了基于神经网络的远海航道船舶流量预测系统。首先分析当前远海航道船舶流量预测系统的研究现状,指出各种系统存在的缺陷,然后利用神经网络的非线性建模性能设计了性能良好的远海航道船舶流量预测系统,最后对该远海航道船舶流量预测系统的有效性进行了测试。本文系统可以高精度实现远海航道船舶流量预测,远海航道船舶流量预测误差远远小于实际应用要求的临界要求,并与其他系统进行对比分析,本文远海航道船舶流量预测系统的预测效果明显更优。
Vessel flow in long-distance waterway is affected by many factors,which has periodicity and strong non-linearity.Traditional linear modeling method cannot accurately fit vessel flow in long-distance waterway,which makes the deviation of vessel flow prediction in long-distance waterway large.In order to overcome the limitation of current long-distance waterway vessel flow prediction system,a forecasting system of ship flow in distant sea channel based on neural network is designed.Firstly,this paper analyses the current research status of the ship flow forecasting system in the distant sea channel,points out the shortcomings of various systems,then designs the ship flow forecasting system in the distant sea channel with good performance by using the non-linear modeling performance of the neural network,and finally tests the effectiveness of the ship flow forecasting system in the distant sea channel.The system in this paper can realize the ship flow forecasting system in the distant sea channel with high precision.Flow forecasting shows that the error of ship flow forecasting is far less than the critical requirement of practical application.Compared with other systems,the forecasting effect of this system is obviously better.
作者
徐健清
XU Jian-qing(Chongqing Creation Vocational College,Chongqing 402160,China)
出处
《舰船科学技术》
北大核心
2019年第4期37-39,共3页
Ship Science and Technology
关键词
远海航道
船舶流量
预测系统
神经网络
offshore channel
ship flow
prediction system
neural network