| 注册
首页|期刊导航|东南大学学报(自然科学版)|基于改进Jaya和DREAM算法的变分贝叶斯学习的输电塔结构损伤识别

基于改进Jaya和DREAM算法的变分贝叶斯学习的输电塔结构损伤识别

胡琴 林昭生 周南飞

东南大学学报(自然科学版)2026,Vol.56Issue(1):142-153,12.
✕
东南大学学报(自然科学版)2026,Vol.56Issue(1):142-153,12.DOI:10.3969/j.issn.1001-0505.2026.01.015

基于改进Jaya和DREAM算法的变分贝叶斯学习的输电塔结构损伤识别

Damage identification of a transmission tower via variational Bayesian learning based on enhanced Jaya and DREAM algorithms

胡琴 1林昭生 1周南飞1

作者信息

  • 1. 华中科技大学土木与水利工程学院,武汉 430074
  • 折叠

摘要

Abstract

To address the challenges of insufficient accuracy in damage identification and computational diffi-culties in posterior distribution estimation for complex spatial transmission tower structures with a large num-ber of uncertain parameters,this paper proposes a sparse Bayesian learning method integrated with the varia-tional Bayesian inference,a multi-chain differential evolution adaptive Metropolis(DREAM)and an en-hanced Jaya algorithm.The method resolves the high-dimensional integration problem in estimating the poste-rior distribution of uncertain parameters through variational Bayesian learning integrated with the DREAM al-gorithm.Simultaneously,the enhanced Jaya algorithm incorporates adaptive chaotic perturbation and im-proved evolutionary strategies to mitigate the inherent local optima entrapment in the original Jaya framework,thereby aiming to balance the trade-off between the identification accuracy and computational efficiency.The proposed approach is validated by using a scaled laboratory transmission tower model.Experimental results demonstrate that both the damage location and severity can be accurately identified through iteratively updating of the model parameters and hyperparameters,and the associated uncertainties can be estimated by the poste-rior distributions.

关键词

损伤识别/变分贝叶斯推断/改进Jaya算法/多链差分自适应进化算法/输电塔

Key words

damage identification/variational Bayesian inference/enhanced Jaya algorithm/multi-chain dif-ferential evolution adaptive Metropolis/transmission tower

分类

建筑与水利

引用本文复制引用

胡琴,林昭生,周南飞..基于改进Jaya和DREAM算法的变分贝叶斯学习的输电塔结构损伤识别[J].东南大学学报(自然科学版),2026,56(1):142-153,12.

基金项目

国家自然科学基金资助项目(52178287). (52178287)

东南大学学报(自然科学版)

1001-0505

访问量0
|
下载量0
段落导航相关论文