{"title":"基于高斯过程的工程结构动态可靠性评估方法","authors":"Jianbao Wei, Kaiqing Qiao, Dameng Zhu, Zhiyi Feng, Xiaobang Wang, Zhijie Liu","doi":"10.1016/j.istruc.2025.109643","DOIUrl":null,"url":null,"abstract":"<div><div>This study presents a novel dynamic reliability assessment method based on active learning Gaussian processes (GP), designed for engineering structural systems subjected to random dynamic loads. In this method, the evaluation of the extreme value distribution of structural responses is used to represent the dynamic reliability level over a specific time period. The GP model is used to establish a strong nonlinear relationship between the influencing factors and output responses of the structure. Furthermore, an active learning strategy is proposed, which includes an optimal sampling domain localization method called 6<em>σ</em>-<em>β</em> sampling domain and a new learning function <em>RIGS</em>(<em><strong>x</strong></em>) for selecting the best samples. The proposed active learning strategy enables the constructed GP model to efficiently and accurately approximate the complex failure limit state function of the structure, and achieve adaptive update of the GP model. Numerical examples and the structural reliability assessment of the crane mounted on the Xuelong 2 polar ship are used to verify the applicability and effectiveness of the proposed method, and the results are compared with other approaches. The proposed method achieves a synergistic optimization of computational efficiency and accuracy in dynamic reliability analysis through the integrated use of intelligent sampling strategies and adaptive modeling techniques.</div></div>","PeriodicalId":48642,"journal":{"name":"Structures","volume":"80 ","pages":"Article 109643"},"PeriodicalIF":4.3000,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Dynamic reliability assessment method based on Gaussian process for engineering structures\",\"authors\":\"Jianbao Wei, Kaiqing Qiao, Dameng Zhu, Zhiyi Feng, Xiaobang Wang, Zhijie Liu\",\"doi\":\"10.1016/j.istruc.2025.109643\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>This study presents a novel dynamic reliability assessment method based on active learning Gaussian processes (GP), designed for engineering structural systems subjected to random dynamic loads. In this method, the evaluation of the extreme value distribution of structural responses is used to represent the dynamic reliability level over a specific time period. The GP model is used to establish a strong nonlinear relationship between the influencing factors and output responses of the structure. Furthermore, an active learning strategy is proposed, which includes an optimal sampling domain localization method called 6<em>σ</em>-<em>β</em> sampling domain and a new learning function <em>RIGS</em>(<em><strong>x</strong></em>) for selecting the best samples. The proposed active learning strategy enables the constructed GP model to efficiently and accurately approximate the complex failure limit state function of the structure, and achieve adaptive update of the GP model. Numerical examples and the structural reliability assessment of the crane mounted on the Xuelong 2 polar ship are used to verify the applicability and effectiveness of the proposed method, and the results are compared with other approaches. The proposed method achieves a synergistic optimization of computational efficiency and accuracy in dynamic reliability analysis through the integrated use of intelligent sampling strategies and adaptive modeling techniques.</div></div>\",\"PeriodicalId\":48642,\"journal\":{\"name\":\"Structures\",\"volume\":\"80 \",\"pages\":\"Article 109643\"},\"PeriodicalIF\":4.3000,\"publicationDate\":\"2025-07-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Structures\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2352012425014584\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Structures","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352012425014584","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Dynamic reliability assessment method based on Gaussian process for engineering structures
This study presents a novel dynamic reliability assessment method based on active learning Gaussian processes (GP), designed for engineering structural systems subjected to random dynamic loads. In this method, the evaluation of the extreme value distribution of structural responses is used to represent the dynamic reliability level over a specific time period. The GP model is used to establish a strong nonlinear relationship between the influencing factors and output responses of the structure. Furthermore, an active learning strategy is proposed, which includes an optimal sampling domain localization method called 6σ-β sampling domain and a new learning function RIGS(x) for selecting the best samples. The proposed active learning strategy enables the constructed GP model to efficiently and accurately approximate the complex failure limit state function of the structure, and achieve adaptive update of the GP model. Numerical examples and the structural reliability assessment of the crane mounted on the Xuelong 2 polar ship are used to verify the applicability and effectiveness of the proposed method, and the results are compared with other approaches. The proposed method achieves a synergistic optimization of computational efficiency and accuracy in dynamic reliability analysis through the integrated use of intelligent sampling strategies and adaptive modeling techniques.
期刊介绍:
Structures aims to publish internationally-leading research across the full breadth of structural engineering. Papers for Structures are particularly welcome in which high-quality research will benefit from wide readership of academics and practitioners such that not only high citation rates but also tangible industrial-related pathways to impact are achieved.