摘要
为准确表征月度船舶交通流量的发展趋势,向水上智能交通系统及港口水域的合理布局等提供基础性数据.基于季节性自回归移动平均(seasonal auto regressive integrated moving average,SARIMA)模型、马尔科夫(Markov)模型、粒子群算法(particle swarm optimization,PSO)建立SARIMA-Markov船舶交通流量预测模型,该模型消除了季节成分、趋势性以及因经济、政策等因素导致的前后背景不一致的干扰,充分考虑了近期状况对预测值的影响,运用具有全局搜索能力的粒子群算法求取模型中的最佳白化系数.以通过赤壁长江公路大桥船舶交通流量月度统计数据为样本进行模型训练和预测,通过计算和、仿真,结果表明,SARIMA-Markov模型的拟合精度及预测精度分别为92.084 9%和95.786 1%,提高了船舶交通流量的预测精度.
In order to accurately represent the development trend of the monthly vessel traffic flow,the basic data for rational distribution of waterborne intelligent transportation systems and harbor waters was provided.Based on the Seasonal Auto Regressive Integrated Moving Average(SARIMA)model,Markov model and Particle Swarm Optimization(PSO),the SARIMA-Markov vessel traffic flow prediction model was established.This model eliminated the disturbance of seasonal components and trends,as well as the influence of inconsistent backgrounds due to factors such as economics and policies,which fully considers the impact of recent conditions on predicted values.And a global search capability particle swarm optimization algorithm was used to obtain the optimal whitening factor in the model.Based on the monthly statistical data of vessel traffic flow through the Chibi Yangtze River Highway Bridge to carry out model training and prediction.After the calculation based on Eviews and Matlab,the results show that the fitting accuracy and prediction accuracy of the SARIMA-Markov model are 92.0849 and 95.7861% respectively,which improves the prediction accuracy of the vessel traffic.
作者
江福才
范庆波
马全党
张帆
马勇
JIANG Fucai1,2, FAN Qingbo1,2, MA Quandang1,2, ZHANG Fan1,2, MA Yong1,2(1.School of Navigation ,Wuhan University of Technology, Wuhan 40063, China;2.Hubei Key Laboratory of Inland Navigation Technology, Wuhan University of Technology, Wuhan 40063, Chin)
出处
《武汉理工大学学报(交通科学与工程版)》
2018年第4期609-615,共7页
Journal of Wuhan University of Technology(Transportation Science & Engineering)
基金
国家自然科学基金项目(51579202
51709218)
国家自然科学青年基金项目(51309186)资助