船舶交通流量预测的研究为水道的规划、设计和船舶通航管理提供基础性依据。将智能融合算法应用于船舶交通流量预测系统,较好地解决了现有船舶预测算法中存在的预测精度不高,依赖于经验等不足。以长江江阴大桥2007年船舶流量观测数据为例进行分析,实验结果表明,融合预测能够对多个数据源进行预测,并可以减缓单种预测方法单独预测的不确定性,从而增加了预测的准确性和整个预测系统的鲁棒性。
The prediction of ship traffic flow is an important fundamental preparation for layout and design of channels as well as management of ship navigation. An intelligent fusion algorithm is applied to ship traffic flow forecasting to remedy the shortcomings in existing ship flow prediction systems, such as low degree of forecasting accuracy and the dependence on experience. Through analyzing the ship flux data gained during the survey in 2007 at Jiagying Bridge over the Changjiang River, the experiment shows that the fusion prediction can forecast multi source data and also reduce the forecast uncertainty so as to increase the accuracy and the robustness of prediction.