{"title":"Physics Informed Machine Learning (PIML) for Design, Management and Resilience-Development of Urban Infrastructures: A Review","authors":"Alvin Wei Ze Chew, Renfei He, Limao Zhang","doi":"10.1007/s11831-024-10145-z","DOIUrl":null,"url":null,"abstract":"<div><p>Building resilient and sustainable urban infrastructures is imperative to prepare future generations against new pandemics and climate change uncertainties. In general, modelling of urban infrastructures requires modelers to carefully consider their initial design phase, subsequent life-span management, and long-term resilience development. With the continual development of machine learning (ML) and artificial intelligence (AI) approaches, significant opportunities are available to civil engineers to improve the existing computing systems of urban infrastructures to contribute to their overall design, management, and resilience-development. Often, an important requirement for the successful adoption of ML/AI techniques is to ensure sufficient field data for training effective predictive models for the above objectives. However, this requirement may be difficult to achieve for all infrastructure engineering applications in the practical field context due to sensor constraints (e.g., limited sensor deployment), coupled with other computational challenges. To address the multiple challenges, this review paper evaluates the important and relevant physics informed machine learning (PIML) publications from 1992 to 2022 for various critical infrastructure engineering applications, namely: (1) PIML for Infrastructures Design and Analysis, (2) PIML for Infrastructure Built-Environment Modelling, (3) PIML for Infrastructures Health Monitoring, and (4) PIML for Infrastructures Resilience Management/Development. In each application, we discuss on the key modelling objectives involved for the specific infrastructure systems, and their associated advantages and/or likely limitations obtained from the PIML implementation. Finally, we then summarize the key research trends and their associated challenges to continue leveraging on PIML techniques to better benefit the overall design, management, and resilience-development of urban infrastructures.</p></div>","PeriodicalId":55473,"journal":{"name":"Archives of Computational Methods in Engineering","volume":"32 1","pages":"399 - 439"},"PeriodicalIF":9.7000,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Archives of Computational Methods in Engineering","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s11831-024-10145-z","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS","Score":null,"Total":0}
引用次数: 0
Abstract
Building resilient and sustainable urban infrastructures is imperative to prepare future generations against new pandemics and climate change uncertainties. In general, modelling of urban infrastructures requires modelers to carefully consider their initial design phase, subsequent life-span management, and long-term resilience development. With the continual development of machine learning (ML) and artificial intelligence (AI) approaches, significant opportunities are available to civil engineers to improve the existing computing systems of urban infrastructures to contribute to their overall design, management, and resilience-development. Often, an important requirement for the successful adoption of ML/AI techniques is to ensure sufficient field data for training effective predictive models for the above objectives. However, this requirement may be difficult to achieve for all infrastructure engineering applications in the practical field context due to sensor constraints (e.g., limited sensor deployment), coupled with other computational challenges. To address the multiple challenges, this review paper evaluates the important and relevant physics informed machine learning (PIML) publications from 1992 to 2022 for various critical infrastructure engineering applications, namely: (1) PIML for Infrastructures Design and Analysis, (2) PIML for Infrastructure Built-Environment Modelling, (3) PIML for Infrastructures Health Monitoring, and (4) PIML for Infrastructures Resilience Management/Development. In each application, we discuss on the key modelling objectives involved for the specific infrastructure systems, and their associated advantages and/or likely limitations obtained from the PIML implementation. Finally, we then summarize the key research trends and their associated challenges to continue leveraging on PIML techniques to better benefit the overall design, management, and resilience-development of urban infrastructures.
期刊介绍:
Archives of Computational Methods in Engineering
Aim and Scope:
Archives of Computational Methods in Engineering serves as an active forum for disseminating research and advanced practices in computational engineering, particularly focusing on mechanics and related fields. The journal emphasizes extended state-of-the-art reviews in selected areas, a unique feature of its publication.
Review Format:
Reviews published in the journal offer:
A survey of current literature
Critical exposition of topics in their full complexity
By organizing the information in this manner, readers can quickly grasp the focus, coverage, and unique features of the Archives of Computational Methods in Engineering.