Multiple attempts of differential evolution Bayesian finite element model revision based on Kriging model

PDF(2980 KB)

Journal of Vibration and Shock ›› 2024, Vol. 43 ›› Issue (9) : 204-213.

QIN Shiqiang, LI Ning, SONG Renxian

Author information +

History +

Abstract

The standard Differential Evolution Adaptive Metropolis (DREAM) algorithm requires parallel computing with multiple Markov chains, which suffers from low convergence efficiency and high computational costs. To address these challenges, this study proposes a novel framework called MT-DREAM(ZS), based on the Kriging model, for multiple-try differential evolution Bayesian finite element model updating. This framework extends the DREAM algorithm by introducing historical vector differential sampling, Snooker update, and multiple-try Metropolis sampling. Furthermore, it utilizes the Kriging model as a surrogate for the finite element model to enable efficient random sampling, allowing for rapid exploration of high-dimensional posterior distributions of correction parameters using a minimal number of parallel chains. To validate the effectiveness of the proposed MT-DREAM(ZS) method, this study conducts the experiment on a consolidated steel plate beam model and compare the updating performance with the standard DREAM algorithm. The results demonstrate that MT-DREAM(ZS) achieves fast convergence of the Markov chains, improving the convergence efficiency by a significant factor of 3.42 compared to DREAM. Moreover, the updating accuracy and stability are enhanced. Additionally, the utilization of the Kriging model substantially reduces computational costs. The proposed framework provides a new approach to address the challenges of low convergence efficiency and high computational costs in the updating of multi-parameter uncertain models.

Key words

Finite element model updating / Bayesian estimation / Multiple-try Metropolis sampling / Differential Evolution Adaptive Metropolis (DREAM) / Kriging model

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

QIN Shiqiang, LI Ning, SONG Renxian. Multiple attempts of differential evolution Bayesian finite element model revision based on Kriging model[J]. Journal of Vibration and Shock, 2024, 43(9): 204-213

References

[1] 彭珍瑞, 郑捷, 白钰, 等. 一种基于改进MCMC算法的模型修正方法[J]. 振动与冲击, 2020, 39(04):236-245. Peng Zhenrui, Zheng Jie, Bai Yu, et al. A model updating method based on an improved MCMC algorithm [J]. Journal of Vibration and Shock, 2020, 39 (04) : 236-245. [2] 秦世强, 胡佳, 曹鸿猷, 等. 基于试验数据的大跨度拱桥有限元模型修正[J]. 中国公路学报, 2019, 32(7);66-76. Qin Shiqiang, Hu Jia, Cao Hongyou, et al. Finite element model updating of long-span arch bridge based on experimental data [J]. China Journal of Highway and Transport, 2019, 32 (7) ; 66-76. [3] Beck J L , Au S K . Bayesian Updating of Structural Models and Reliability using Markov Chain Monte Carlo Simulation[J]. Journal of Engineering Mechanics, 2002, 128(4):380-391. [4] 韩芳,钟冬望,汪君.基于贝叶斯法的复杂有限元模型修正研究[J].振动与冲击,2012,31(01):39-43. Han Fang, Zhong Dongwang, Wang Jun. Complicated finite element model updating based on Bayesian method [J]. Journal of Vibration and Shock, 2012,31 ( 01 ) : 39-43. [5] 易伟建, 周云, 李浩, 等. 基于贝叶斯统计推断的框架结构损伤诊断研究[J]. 工程力学, 2009, 26(5): 121-129. Yi Weijian, Zhou Yun, Li Hao, et al. Damage assessment research of frame structures based on Bayesian statistical inference [J]. Engineering Mechanics, 2009, 26 (5) : 121-129. [6] 邓振鸿, 张保强, 苏国强, 等. 基于近似似然的频响函数不确定性模型修正[J]. 振动、测试与诊断, 2020, 40(3):548-554. Deng Zhenhong, Zhang Baoqiang, Su Guoqiang, et al. Uncertainty modle updating of frequency response function based on approximate likelihood function [J]. Vibration, test and diagnosis, 2020, 40 (3) : 548-554. [7] 蒋伟, 刘纲. 基于多链差分进化的贝叶斯有限元模型修正方法[J]. 工程力学, 2019, 36(06):101-108. Jiang Wei, LIU Gang. Bayesian finite element model updating method based on multi-chain differential evolution [J]. Engineering Mechanics, 2019, 36 (06) : 101-108. [8] Braak C J F T . A Markov Chain Monte Carlo version of the genetic algorithm Differential Evolution: easy Bayesian computing for real parameter spaces[J]. Statistics & Computing, 2006, 16(3):239-249. [9] Vrugt J A, Ter Braak C J F, Clark M P, et al. Treatment of input uncertainty in hydrologic modeling: Doing hydrology backward with Markov chain Monte Carlo simulation[J]. Water Resources Research, 2008, 44(12). [10] 仉文岗, 顾鑫, 刘汉龙, 等. 基于贝叶斯更新的非饱和土坡参数概率反演及变形预测[J]. 岩土力学, 2022, 43(04):1112-1122. Zhang Wengang, Gu Xin, Liu Hanlong, et al.Probabilistic back analysis of soil paraments and displacement prediction of unsaturated slopes using Bayesian updating[J]. Rock and Soil Mechanics, 2022,43 (04) : 1112-1122. [11] 叶凯. 基于改进马尔科夫链算法的贝叶斯有限元模型修正研究[D].重庆大学,2017. Ye Kai.Research on Bayesian finite element model updating based on improved Markov chain algorithm [D].Chongqing University, 2017. [12] 茆诗松, 程依明, 濮晓龙, 等. 概率论与数理统计教程[M]. 北京: 高等教育出版社, 2011: 107-108. Mao Shisong, Cheng Yiming, Pu Xiaolong, et al. Probability and statistics [ M ]. Beijing : Higher Education Press, 2011 : 107-108. [13] 方圣恩, 陈杉. 结合近似贝叶斯计算和改进群体蒙特卡洛抽样的结构损伤识别[J]. 振动与冲击, 2020, 39(5):143-149. FANG Shengen, CHEN Shan. Structural damage identification using ABC combined with improved PMC sampling [J]. Journal of Vibration and Shock, 2020, 39 (5) : 143-149. [14] Vrugt J A, Ter BraakC J F, Diks C G H, et al. Accelerating Markov chain Monte Carlo simulation by differential evolution with self-adaptive randomized subspace sampling[J]. International journal of nonlinear sciences and numerical simulation, 2009, 10(3): 273-290. [15] Gelman A, Rubin D B. Inference from iterative simulation using multiple sequences [J]. Statistical Science, 1992, 7(4): 457- 472. [16] Zeng J, Kim Y H, Qin S. Bayesian Model Updating for Structural Dynamic Applications Combing Differential Evolution Adaptive Metropolis and Kriging Model[J]. Journal of Structural Engineering, 2023, 149(6): 04023070. [17] Haario H, Laine M, Mira A, et al. DRAM: efficient adaptive MCMC[J]. Statistics and computing, 2006, 16(4): 339-354. [18] Braak C , Vrugt J A . Differential Evolution Markov Chain with snooker updater and fewer chains[J]. Statistics & Computing, 2008, 18(4):435-446. [19] Liu J S , Liang F , Wong W H . The Multiple-Try Method and Local Optimization in Metropolis Sampling[J]. Publications of the American Statistical Association, 2000, 95(449):121-134. [20] Laloy E, Vrugt J A. High‐dimensional posterior exploration of hydrologic models using multiple-try DREAM (ZS) and high‐performance computing[J]. Water Resources Research, 2012, 48(1). [21] 薛国峰, 王伟, 赵威. 基于多点 Metropolis 的子集模拟方法[J]. 哈尔滨工业大学学报, 2011 (S1): 120-125. Xue Guofeng, Wang Wei, Zhao Wei.Subset simulation methods based on Multiple-Try Metropolis [J].Journal of Harbin Institute of Technology, 2011 (S1) : 120-125. [22] 曹彤彤. 基于贝叶斯理论的地下水模型评价[D].南京大学,2020. Cao Tongtong. Groundwater model evaluation based on Bayesian theory [D]. Nanjing University, 2020. [23] 周云, 贾凡丁, 奚树杭. 基于贝叶斯理论的多模型结构识别的试验研究[J]. 湖南大学学报(自然科学版), 2018, 45(5):36-45. Zhou Yun, Jia Fanding, Xi Shuhang.Experimental study on multi-model structure recognition based on Bayesian theory [J].Journal of Hunan University (Natural Science), 2018,45 (5) : 36-45. [24] Sadegh M, Vrugt J A. Approximate bayesian computation using markov chain monte carlo simulation: Dream (abc)[J]. Water Resources Research, 2014, 50(8): 6767-6787.