Scholarlink

<p>Structural health monitoring (SHM) in civil engineering structures is essential for ensuring structural integrity and safety. The current study presents an integration of particle swarm optimization (PSO) with a support vector machine (SVM) model for SHM of joints in steel frame structures with statistical features of vibration data. In the study, the PSO is employed to optimize the SVM hyperparameters (penalty parameters and Gaussian kernel function) to enhance accuracy and robustness. For that purpose, a five-story 3D steel frame structure is considered. An impact excitation is used to excite the structure and record the time-history acceleration data for both damaged and undamaged cases. From the data, the statistical features were extracted and used as input to the PSO-based SVM model. The training and testing results show that the model is effective in distinguishing between undamaged and damaged cases. This study creates a robust model for SHM applications, advancing the development of autonomous structural evaluation systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimized SVM-based model for health monitoring of joints in a multi-story 3D steel frame structure

  • 期刊论文

  • 发表日期：

    2025年3月19日

查看全文
  • Maloth Naresh,
  • Maloth Ramesh,
  • Ashish Balavant Jadhav

摘要

Structural health monitoring (SHM) in civil engineering structures is essential for ensuring structural integrity and safety. The current study presents an integration of particle swarm optimization (PSO) with a support vector machine (SVM) model for SHM of joints in steel frame structures with statistical features of vibration data. In the study, the PSO is employed to optimize the SVM hyperparameters (penalty parameters and Gaussian kernel function) to enhance accuracy and robustness. For that purpose, a five-story 3D steel frame structure is considered. An impact excitation is used to excite the structure and record the time-history acceleration data for both damaged and undamaged cases. From the data, the statistical features were extracted and used as input to the PSO-based SVM model. The training and testing results show that the model is effective in distinguishing between undamaged and damaged cases. This study creates a robust model for SHM applications, advancing the development of autonomous structural evaluation systems.