集成栈式自编码器与XGBoost的深度学习海浪有效波高预报模型

计算机与现代化 ›› 2021, Vol. 0 ›› Issue (04): 32-36.

集成栈式自编码器与XGBoost的深度学习海浪有效波高预报模型

(1.河海大学海岸灾害及防护教育部重点实验室,江苏南京210098;2.河海大学计算机与信息学院,江苏南京211100;
3.福建省海洋预报台,福建福州350003;4.河海大学港口海岸与近海工程学院,江苏南京210098)

出版日期:2021-04-22 发布日期:2021-04-25
作者简介:陆小敏(1995—)，女，安徽合肥人，硕士研究生，研究方向：数据挖掘，E-mail: 1846072423@qq.com; 刘凡(1988—)，男，江苏宿迁人，教授，博士，研究方向：计算机视觉，机器学习，多媒体分析与理解，E-mail: fanliu@hhu.edu.cn; 通信作者：李雪丁(1982—)，男，江西新余人，副研究员，硕士，研究方向：海洋预报，海洋遥感，E-mail: lxd007@xmu.edu.cn。
基金资助:
河海大学海岸灾害及防护教育部重点实验室开放基金资助项目(20150009); 福建省科技计划项目(2018Y0001)

Ensemble Stacked Autoencoders and XGBoost Based Deep Learning Model for Significant Wave Height Forecasting

(1. Key Laboratory of Coastal Disaster and Protection of Ministry of Education, Hohai University, Nanjing 210098, China;
2. College of Computer and Information, Hohai University, Nanjing 211100, China;
3. Fujian Marine Forecasts, Fuzhou 350003, China;
4. College of Harbour, Coastal and Offshore Engineering, Hohai University, Nanjing 210098, China)

Online:2021-04-22 Published:2021-04-25

摘要/Abstract

摘要： 有效波高预报对人类海上活动和海洋工程都至关重要。人工神经网络在有效波高预报中得到广泛的应用，并取得了良好的效果。但是，它作为一种浅层的网络架构，表达能力有限，这使得预报准确性在不同区域中波动。因此，为了提高有效波高的总体预报准确性，本文提出一种集成栈式自编码器（SAE）和XGBoost的深度学习海浪有效波高预报模型。首先，利用SAE算法强大的特征表征能力处理海浪数据，实现数据的扩维表达。其次，将SAE深层的特征作为XGBoost算法的输入，进行有效波高预测。本文重点研究有效波高预报方法，并根据台湾海峡中部2号大浮标2017年全年的实测波浪资料进行研究。实验结果表明，本文方法在确定性系数（R^2）和均方误差（MSE）方面均优于现有方法。

关键词: 有效波高, 栈式自编码器, XGBoost, 深度学习

Abstract: The significant wave height forecast is crucial for both human marine activities and marine engineering. The artificial neural network has been widely used in significant wave height prediction and achieved good results. However, as a shallow network architecture, it has limited expressive ability, making the forecast accuracy fluctuate in different regions. Therefore, to improve the overall forecast accuracy of the significant wave height, this paper proposes a deep learning model of significant wave height forecasting by integrating stacked autoencoders (SAE) and XGBoost. First, the powerful feature representation capabilities of the SAE algorithm are used to process ocean wave data to realize the extended dimension expression of the data. Secondly, the deep feature expression of SAE is used as the input of the XGBoost algorithm to predict effective wave heights. This paper focuses on the significant wave height prediction method and uses the measured wave data of Buoy 2 in the central Taiwan Strait in 2017. The experimental results show that our approach is superior to existing methods in terms of deterministic coefficient (R^2) and mean square error (MSE).

Key words: significant wave height, stacked autoencoders, XGBoost, deep learning

陆小敏, 刘凡, 蔡丽华, 李雪丁, 徐啸. 集成栈式自编码器与XGBoost的深度学习海浪有效波高预报模型[J]. 计算机与现代化, 2021, 0(04): 32-36.

LU Xiao-min, LIU Fan, CAI Li-hua, LI Xue-ding, XU Xiao. Ensemble Stacked Autoencoders and XGBoost Based Deep Learning Model for Significant Wave Height Forecasting[J]. Computer and Modernization, 2021, 0(04): 32-36.

参考文献

［1］邢闯,李本霞. 中国近海2012年灾害性海浪分析及2013年预测［J］. 海洋预报, 2013,30(3):1-8.
［2］ The Wamdi Group. The WAM model: A third generation ocean wave prediction model［J］. Journal of Physical Oceanography, 1988,18(12):1775-1810.
［3］ SAAD A M E. Wave and Wind Conditions in the Red Sea: A Numerical Study Using a Third Generation Wave Model［D］. The University of Bergen, 2010.
［4］ TUOMI L, BJORKQVIST J V. Wave forecasting in coastal archipelagos［C］// Proceedings of the 2014 IEEE/OES Baltic International Symposium (BALTIC). 2014. DOI: 10.1109/BALTIC.2014.6887855.
［5］ BOOIJ N, RIS R C, HOLTHUIJSEN L H. A third-generation wave model for coastal regions: 1. Model description and validation［J］. Journal of Geophysical Research, 1999,104(C4):7649-7666.
［6］ LIANG B C, GAO H J, SHAO Z X. Characteristics of global waves based on the third-generation wave model SWAN［J］. Marine Structures, 2019,64:35-53.
［7］ AKPINAR A, VAN VLEDDER G P, KOMURCU M I, et al. Evaluation of the numerical wave model (SWAN) for wave simulation in the Black Sea［J］. Continental Shelf Research, 2012,50-51:80-99.
［8］ LUO W M, FLATHER R. Nesting a nearshore wave model (SWAN) into an ocean wave model (WAM) with application to the southern North Sea［C］// Proceedings of the 1997 International Conference on Computer Modelling of Seas and Coastal Regions. 1997:253-264.
［9］ TOLMAN H L, BALASUBRAMANIYAN B, BURROUGHS L D, et al. Development and implementation of wind-generated ocean surface wave modelsat NCEP［J］. Weather and Forecasting, 2002,17(2):311-333.
［10］SWAIN J, UMESH P A. Prediction of uncertainty using the third generation wave model WAVEWATCH III driven by ERA-40 and blended winds in the North Indian Ocean［J］. Journal of Oceanography and Marine Research, 2018,6(1). DOI: 10.4172/2572-3103.1000173.
［11］BARBARIOL F, ALVES J H G M, BENETAZZO A, et al. Numerical modeling of space-time wave extremes using WAVEWATCH III［J］. Ocean Dynamics, 2017,67(3-4):535-549.
［12］SHARIFI F S, EZAM M, KARAMI K A. Evaluating the results of Hormuz strait wave simulations using WAVEWATCH-III and MIKE21-SW［J］. International Journal of Marine Science and Engineering, 2012,2(2):163-170.
［13］郑崇伟,潘静,黄刚. 利用WW3模式实现中国海击水概率数值预报［J］. 北京航空航天大学学报, 2014,40(3):314-320.
［14］TSAI C C, WEI C C, HOU T H, et al. Artificial neural network for forecasting wave heights along a ship’s route during hurricanes［J］. Journal of Waterway, Port, Coastal and Ocean Engineering, 2018,144(2). DOI: 10.1061/(ASCE)WW.1943-5460.0000427.
［15］陈希,李妍,沙文钰,等. 人工神经网络技术在台风浪预报中的应用［J］. 海洋科学, 2003,27(2):63-67.
［16］齐义泉,张志旭,李志伟,等. 人工神经网络在海浪数值预报中的应用［J］. 水科学进展, 2005,16(1):32-35.
［17］王燕,钟建,张志远. 支持向量回归的机器学习方法在海浪预测中的应用［J］. 海洋预报, 2020,37(3):29-34.
［18］金权,华锋,杨永增. 基于SVM的海浪要素预测试验研究［J］. 海洋科学进展, 2019,37(2):199-209.
［19］MAHJOOBI J, MOSABBEB E A. Prediction of significant wave height using regressive support vector machines［J］. Ocean Engineering, 2009,36(5):339-347.
［20］朱智慧,曹庆,徐杰. 神经网络方法在上海沿海海浪预报中的应用［J］. 海洋预报, 2018,35(5):25-33.
［21］陈德花,刘铭,苏卫东,等. BP人工神经网络在福建MM5模式风场预报释用研究中的应用［J］. 福建气象, 2010(2):29-32.
［22］GOPINATH D I, DWARAKISH G S. Wave prediction using neural networks at new Mangalore port along west coast of India［J］. Aquatic Procedia, 2015,4:143-150.
［23］MANDAL S, PRABAHARAN N. Ocean wave forecasting using recurrent neural networks［J］. Ocean Engineering, 2006,33(10):1401-1410.
［24］顾兴健,赵璐,金明,等. 基于LSTM神经网络的我国典型试航海域环境短期预报方法研究［J］. 中国造船, 2017,58(4):100-107.
［25］LIU F, XU F, YANG S. A flood forecasting model based on deep learning algorithm via integrating stacked autoencoders with BP neural network［C］// Proceedings of the 2017 IEEE 3rd International Conference on Multimedia Big Data (BigMM). 2017:58-61.
［26］杨洪富,贾晓亮. 基于SAE的航空发动机气路故障诊断［J］. 航空计算技术, 2018,48(6):47-51.

[1]	祁贤, 刘大铭, 常佳鑫. 基于改进自注意力机制的多视图三维重建[J]. 计算机与现代化, 2024, 0(11): 106-112.
[2]	陈凯1, 李宜汀1, 2, 全华凤1 . 基于改进YOLOv8的河道废弃瓶检测方法[J]. 计算机与现代化, 2024, 0(11): 113-120.
[3]	杨骏1, 胡为1, 朱文福2. 基于改进MobileNetV3的视觉SLAM回环检测算法[J]. 计算机与现代化, 2024, 0(10): 21-26.
[4]	王莹莹, 郝潇. 基于Res2Net和递归门控卷积的细粒度图像分类[J]. 计算机与现代化, 2024, 0(10): 74-79.
[5]	史星宇1, 李强2, 庄莉3, 梁懿3, 王秋琳3, 陈锴3, 伍臣周3, 常胜1. 一种面向工业部署的目标检测模型蒸馏技术[J]. 计算机与现代化, 2024, 0(10): 93-99.
[6]	张泽1, 张建权2, 3, 周国鹏2, 3. 基于改进YOLOv8s的摄像头模组缺陷检测[J]. 计算机与现代化, 2024, 0(09): 107-113.
[7]	程亚子1, 雷亮1, 2, 陈瀚1, 赵毅然1. 基于转置注意力的多尺度深度融合单目深度估计[J]. 计算机与现代化, 2024, 0(09): 121-126.
[8]	程萌, 李浩. 改进YOLOv5s的落叶树鸟巢检测方法[J]. 计算机与现代化, 2024, 0(08): 24-29.
[9]	王梦溪, 李峻. 老年人跌倒检测技术研究综述[J]. 计算机与现代化, 2024, 0(08): 30-36.
[10]	时现伟1, 范鑫2. 基于轻量化的视频帧场景语义分割方法[J]. 计算机与现代化, 2024, 0(08): 49-53.
[11]	徐新爱, 李钢. 基于DCGAN的课堂表情图像生成方法[J]. 计算机与现代化, 2024, 0(08): 88-91.
[12]	高帅鹏, 王怡凡. 基于图像的群体情绪识别综述[J]. 计算机与现代化, 2024, 0(08): 98-107.
[13]	黄文栋, 王怡凡. 基于模态类别的多模态信息处理与融合综述[J]. 计算机与现代化, 2024, 0(07): 47-62.
[14]	武丽1, 张征浩2, 葛彩成2, 俞俊2. 基于改进SCNN网络的车道线检测算法[J]. 计算机与现代化, 2024, 0(07): 87-92.
[15]	苏凯旋. 基于改进XGBoost模型的网络入侵检测研究[J]. 计算机与现代化, 2024, 0(06): 109-114.