桥梁风工程机器学习研究进展  

Research Progressof Machine Learning in Bridge Wind Engineering

作　　者：赖马树金[1,2,3] 李文杰 冯辉[1,2] 周旭曦 张泽宇 陈文礼 李惠[1,2,4] LAIMA Shu-jin;LI Wen-jie;FENG Hui;ZHOU Xu-xi;ZHANG Ze-yu;CHEN Wen-li;LI Hui(Key Lab of Smart Prevention and Mitigation of Civil Engineering Disasters of the Ministry of Industry and Information Technology,Harbin Institute of Technology,Harbin 150090,Heilongjiang,China;Key Lab of Structures Dynamic Behavior and Control of the Ministry of Education,Harbin Institute of Technology,Harbin 150090,Heilongjiang,China;State Key Laboratory for Health and Safety of Bridge Structures,Wuhan,430034,Hubei,China;Guangdong-Hong Kong-Macao Joint Laboratory for Data-driven Fluid Mechanics and Engineering Applications,Harbin Institute of Technology(Shenzhen),Shenzhen 518055,Guangdong,China)

机构地区：[1]哈尔滨工业大学土木工程智能防灾减灾工业与信息化部重点实验室,黑龙江哈尔滨150090 [2]哈尔滨工业大学结构工程灾变与控制教育部重点实验室,黑龙江哈尔滨150090 [3]桥梁结构健康与安全国家重点实验室,湖北武汉430034 [4]哈尔滨工业大学(深圳)粤港澳数据驱动下的流体力学与工程应用联合实验室,广东深圳518055

出　　处：《中国公路学报》2023年第8期1-13,共13页China Journal of Highway and Transport

基　　金：国家重点研发计划项目(2022YFC3005303);国家自然科学基金项目(52178470);桥梁结构健康与安全国家重点实验室开放课题项目(BHSKL20-01-KF)。

摘　　要：自然风场复杂风况参数、非线性气动力与流固耦合作用等是桥梁风工程面临的系列主要挑战。在过去几十年中,风洞试验、CFD数值模拟和结构健康监测(SHM)积累了大量桥梁结构风与风效应数据,这为风工程的研究提供了宝贵的数据资源。由于具有良好的非线性表征能力、强大的优化算法、优异的泛化性能和灵活的网络结构,近年来机器学习(尤其是深度学习)在非线性科学和工程问题中取得了巨大成功。融合机器学习算法和大数据的数据驱动方法有助于克服桥梁风工程所面临的挑战,并从海量数据中挖掘物理新知识。目前机器学习方法已应用于风工程研究的各个领域,例如风环境、空气动力学、风致振动、气动优化和控制以及风灾害评估等。为系统总结梳理该领域的最新研究成果,从桥梁风环境机器学习预测、结构风压和气动力机器学习预测、桥梁风振机器学习预测3个方面重点介绍机器学习方法在桥梁风工程中的应用进展。Complex wind environments, nonlinear aerodynamics, and wind-structure interactions are the main challenges in wind-engineering research. During the past several decades, the number of data accumulated from wind tunnel tests, numerical simulations based on computational fluid dynamics, and structural health monitoring has become massive, providing valuable resources for addressing these challenges. With the development of deep-learning technology, machine learning (ML) has achieved great success in nonlinear science and engineering problems owing to its nonlinear representation capabilities, powerful optimization algorithms, excellent generalization performance, and flexible network architecture. Emerging data-driven approaches based on ML algorithms have helped address these challenges in wind engineering and increased physical and engineering knowledge based on the available wind-engineering data. The application of ML in wind engineering involves all aspects of the wind-engineering field, such as the wind environment, aerodynamic and aeroelastic forces, wind-induced vibrations, aerodynamic optimization and control, and wind disaster assessment. The purpose of this paper is to introduce the research progress and state-of-the-art technologies in ML and artificial-intelligence applications for bridge wind engineering.

关 键 词：桥梁工程 风工程 综述 机器学习 风致振动 

分 类 号：U441[建筑科学—桥梁与隧道工程]