摘要
潮汐在海洋生物、岸侵蚀与管理以及科学研究中都扮演重要的角色,而传统的潮汐预测方法存在精度不高、效率低的缺点。为了克服这些缺陷,本文提出使用白鲸优化算法对双向长短时神经网络的超参数进行寻优。与原始的BILSTM网络训练的结果相比,优化后的BILSTM网络在MSE、MAPE上下降了0.08、1.59,在关联系数R上提升了0.06,证明了本文所提出的改进在潮位预测上的有效性。
Tides play an important role in marine life,coastal erosion and management,and scientific research,while traditional tide prediction methods suffer from the shortcomings of low accuracy and efficiency.To overcome these shortcomings,this paper proposes to use the beluga optimization algorithm to optimize the hyperparameters of the bidirectional long short-term neural network.Compared with the results of the original BILSTM network training,the optimized BILSTM network decreases by 0.08 and 1.59 in MSE and MAPE and increases by 0.06 in the correlation coefficient R,which proves the effectiveness of the proposed improvement in tide level prediction.
作者
吴际
周玉岑
WU Ji;ZHOU Yucen(Jiangxi Jianghui Geological Engineering Survey Institute Co.,Ltd.,Shangrao 334000,China;Jiangxi Shangrao Urban Rural Planning Research Center,Shangrao 334000,China)
出处
《测绘与空间地理信息》
2025年第2期174-176,共3页
Geomatics & Spatial Information Technology
关键词
潮位预测
白鲸优化算法
双向长短时网络
tide level prediction
beluga whale optimization
Bi-directional Long Short-Term Memory
