针对面板波浪力预测问题，利用Matlab软件开展了神经网络预测模型的比较和优化。首先，选取GA-BP、SSA-BP、PSO-RBF和WOA-RBF四种神经网络模型分别进行波浪力预测，通过平均绝对误差和均方根误差等评价指标，分析得到WOA-RBF神经网络预测模型表现较优。其次，针对WOA-RBF模型在应用中存在的问题，提出一种优化的IWOA-RBF神经网络，并进行该神经网络的效果研究，表明IWOA-RBF模型预测误差小，精度更准。最后，分析IWOA-RBF神经网络的主要影响因素，给出应用建议。研究成果将对面板波浪力预测技术有所提升，可为船舶和海岸工程设计、施工及防护等工作提供参考。
This paper focuses on the issue of predicting panel wave forces, employing Matlab software to conduct comparative and optimization research on neural network prediction models. Initially, four neural network models, namely GA-BP, SSA-BP, PSO-RBF, and WOA-RBF, were selected for the prediction of wave forces. Through evaluation metrics such as mean absolute error and root mean square error, it was determined that the WOA-RBF neural network model exhibits superior performance. Subsequently, addressing the shortcomings of the WOA-RBF model in practical applications, an optimized IWOA-RBF neural network is proposed, and its effectiveness is studied, demonstrating that the IWOA-RBF model offers reduced prediction errors and enhanced accuracy. Lastly, the study analyzes the main factors influencing the IWOA-RBF neural network and provides application recommendations. The research findings contribute to the advancement of panel wave force prediction technology, offering valuable insights for the design, construction, and protection of ships and coastal engineering projects.
2024,46(22): 17-22 收稿日期：2024-1-30
DOI：10.3404/j.issn.1672-7649.2024.22.003
分类号：U661；TV139.2
基金项目：长沙理工大学水沙科学与水灾害防治湖南省重点实验室开放基金资助项目（2020SS03）
作者简介：金凤(1980-),女,博士,副教授,研究方向为波浪与结构物相互作用
参考文献：
[1] 张红玉. 基于遗传算法的神经网络在波浪冲击力计算中的应用[D]. 大连: 大连理工大学, 2005.
[2] FORMENTIN S M, ZANUTTIGH B, MEER J W V D. A neural network tool for predicting wave reflection, overtopping and transmission[J]. Coastal Engineering Journal, 2017, 59(1): 13.
[3] WEI Zhangping, DAVISON A. A convolutional neural network based model to predict nearshore waves and hydrodynamics[J]. Coastal Engineering. 2022: 171-183.
[4] XU Guoji, JI Chengjie, WEI Huan, et al. A novel ensemble model using artificial neural network for predicting wave-induced forces on coastal bridge decks.[J]. Engineering with Computers, 2022, 39(5): 47-53.
[5] 刘春艳, 凌建春, 寇林元, 等. GA-BP神经网络与BP神经网络性能比较[J]. 中国卫生统计, 2013, 30(2): 173-176.
[6] 孙全, 孙渊. 基于麻雀搜索算法的BP神经网络优化技术[J]. 上海电机学院学报, 2022, 25(1): 12-16.
[7] 邵宁宁, 王英. 基于Adam优化的改进PSO-RBF牵引变压器故障诊断研究[J]. 电气工程学报, 2024(12): 209-216.
[8] 李其操, 董自健. 基于WOA-RBF的温室温度预测研究[J]. 现代计算机, 2023, 29(11): 41-51.
[9] TENG Zhijun, LYU Jinling, GUO Liwen, et al. An improved hybrid grey wolf optimization algorithm based on Tent mapping[J]. Journal of Hrbin Istitute of Technology, 2018, 50(11): 40-49.
[10] 龙文, 伍铁斌, 唐明珠, 等. 基于透镜成像学习策略的灰狼优化算法[J]. 自动化学报, 2020, 46(10): 2148-2164.
[11] ZHANG Xinming, WANG Xia, KANG Qiang. Improved grey wolf optimizer and its application to high- dimensional functionand FCM optimization[J]. Control and Dcisione, 2019, 34(10): 2073-2084.
