基于BAS-BP模型的南海北部波高预测  

Prediction of wave height based on BAS-BP model in the northern part of the South China Sea

作　　者：夏天亮 李醒飞[1,2,3] 杨少波 XIA Tian-liang;LI Xing-fei;YANG Shao-bo(State Key Laboratory of Precision Measuring Technology and Instruments,Tianjin University,Tianjin 300072,China;Qingdao Institute for Ocean Technology of Tianjin University,Qingdao 266200,China;Pilot National Laboratory for Marine Science and Technology(Qingdao),Qingdao 266003,China)

机构地区：[1]天津大学精密测试技术及仪器国家重点实验室,天津300072 [2]天津大学青岛海洋技术研究院,山东青岛266200 [3]青岛海洋科学与技术试点国家实验室,山东青岛266003

出　　处：《海洋湖沼通报》2021年第5期9-16,共8页Transactions of Oceanology and Limnology

基　　金：青年教师科研启动基金资助项目(Pilq1702);“问海计划”(ZR2016WH01)。

摘　　要：在海洋探测与预警领域,海浪高度预测是一项非常重要的工程。为了准确、快速地预测海浪高度,本文采用天牛须搜索算法优化的BP神经网络模型(BAS-BP),该模型结合了BP神经网络的局部搜索能力和BAS的全局搜索能力,避免了过早收敛和预测效果差的问题。本研究的数据集为WAVEWATCH-III模型模拟的2015—2017年南海北部波高数据。以南海北部的8个地点作为预测点,它们涵盖了广泛的地理位置和不同的天气条件。研究结果表明,相比于BP、GA-BP和PSO-BP模型,BAS-BP模型的预测效果最好。In ocean detection and early warning,the wave height prediction is a very important project.In order to predict wave height accurately and quickly,this paper uses the beetle antennae search algorithm to optimize the BP neural network model(BAS-BP).This model combines the local search ability of BP neural network and the global search ability of BAS,avoiding the problems of premature convergence and poor prediction effect.The wave height data simulated by the WAVEWATCH-III model in the northern part of the South China Sea in 2015—2017 was applied in this study.We selected eight sites in the northern South China Sea as prediction points,covering a wide range of geographical locations and different weather conditions.The results show that compared with the BP,GA-BP and PSO-BP models,the BAS-BP model has the best prediction effect.

关 键 词：南海北部 BAS-BP模型 波高预测 

分 类 号：P743.2[天文地球—海洋科学]