港口船舶交通流量预测.pdfVIP

35 3 大 连 海 事 大 学 学 报 Vol.35 No.3 2009 8 Journal of Dalian Maritime University Aug., 2009 : 1006-7736(2009) 03-0040-03 X 1 1 2 李红喜 ,付玉慧, 张仁初 ( 1. , 116026; 2., 315200) :为更精确地对港口或航道内船舶交通流量进行预测, . 分别建立BP 神经网络预测模 和RBF 神经网络预测模 , 进行仿真, 并以宁波港船舶交通流量为例进行验证. 结果表 . , 明,在宁波港现有发展基础和港口设施状况下, RBF 神经网 . 络用于宁波港船舶交通流量预测误差较小,预测值与实际值 相近. 1 神经网络预测模 :船舶交通流量; 预测模 ;神经网络 , :U691 :A Ship traffic volume forecasting in port , 1 1 2 [5] LI ong-xi ,FU Yu-hui , Z ANG Ren-chu . [6] . ( 1. N avigati on Coll ege , Dalian Maritime University, Dalian 116026, China; 2. Ningbo M aritime Safety Administration, Ningbo 315200, China) , . , Abstract: To forecast ship traffic volume in port or waterway precisely, back propagation ( BP) and radial basis function BP RBF . ( RBF) neural network forecasting models were established and 1. 1 BP simulated. Tests in Ningbo port show that RBF neural network BP ( ) has the smaller error in forecasting ship traffic volume, and the . , predictive value is close to actual oneunder the current basis and y y , port facilities. ., Key words: ship traffic volume; forecasting