一种高精度的短期潮汐预报模型  

A high-accuracy short-term tide prediction model

作　　者：柳成[1] 尹建川[1] LIU Cheng YIN Jianchuan(Navigation College, Dalian Maritime University, Dalian 116026, Liaoning, China)

机构地区：[1]大连海事大学航海学院,辽宁大连116026

出　　处：《上海海事大学学报》2016年第3期74-80,共7页Journal of Shanghai Maritime University

基　　金：中央高校基本科研业务经费(3132016116);交通运输部应用基础研究项目(2014329225010);辽宁省教育厅科学研究一般项目(L2014214)

摘　　要：为提高潮汐预报的精度,提出一种基于支持向量机(Support Vector Machine,SVM)的模块化潮汐实时预报模型.将潮汐分为受天体引潮力影响的天文潮和受环境因素和其他因素影响的非天文潮,分别使用调和分析法和改进的SVM对天文潮和非天文潮进行预报,结合两种方法的输出构造最终的潮汐预报结果.在对非天文潮的预测中,将SVM与灰色模型相结合,并利用粒子群优化(Particle Swarm Optimization,PSO)算法对SVM的参数进行优化以提高预报精度.利用火奴鲁鲁港口的实测潮汐数据进行实时潮汐预报仿真.仿真结果表明该方法具有较高的短期预报精度.In order to improve the accuracy of tide prediction,a modular real-time tide prediction model is proposed based on the Support Vector Machine（ SVM）. The tides are divided into astronomical tides and non-astronomical tides,where astronomical tides are caused by tide-generating force of celestial bodies and non-astronomical tides are caused by environmental and other factors; astronomical tides are predicted by the harmonic analysis method while non-astronomical tides are predicted by the improved SVM. The final prediction result is achieved by combining the outputs of the two methods. In the prediction of non-astronomical tides,in order to improve the prediction accuracy,SVM is combined with the grey model and the parameters of SVM are optimized by Particle Swarm Optimization（ PSO） algorithm.The measured tide data of Honolulu Port are used to carry out the real-time tide prediction simulation.Simulation results show that the model is of high accuracy for short-term tide prediction.

关 键 词：潮汐预报 模块化方法 调和分析法 支持向量机(SVM) 灰色模型 粒子群优化(PSO) 

分 类 号：U675.12[交通运输工程—船舶及航道工程]