Optimizing the Durability of Buildings Against Earthquake-Induced Collapse by the Implementation of Deep Learning-Based Control Strategy
Normaisharah Mamat,Rawad Abdulghafor,Sherzod Turaev,Fitri Yakub,Mohd Fauzi Bin Othman
DOI: https://doi.org/10.1109/access.2024.3452763
IF: 3.9
2024-09-10
IEEE Access
Abstract:Natural disasters have the potential to inflict damage to building structures, which prompting the structural engineers striving to build buildings that are both safer and durable. This research endeavor aims to enhance the stability of a structure by using control strategy to reduce the hazards of earthquake-induced damage. As a consequence, a hybrid control device incorporates control strategies to improve structural integrity has been investigated. This study proposes control system comprising integrated nonsingular terminal sliding mode control (TSMC), nonsingular TSMC, and TSMC. Integrated nonsingular TSMC merges deep learning methods, specifically long short-term memory (LSTM). LSTM has the capability to forecast subsequently values by analyzing previous sequential data. This character enhances precision in systems, leading to better decision-making. This attribute enhances the precision of any system, contributing to better decision-making. The integration controller empowers the system to adjust its operation to achieve the most optimal mode of operation, hence improving control performance regardless of significant uncertainty. The integrated controller is invented using both the control law and sliding surface. A two-degree-of-freedom (DOF) system structure is constructed in Matlab and validated by experimental work associated with LMS Test.Lab software. In addition, the integrated controller is employed in a 10-DOF system represents a high-level structure. The performance of integrated nonsingular TSMC is compared to nonsingular TSMC and TSMC in terms of displacement response, sliding surface, and damage probability. The outcomes indicated that integrated nonsingular TSMC could mitigate vibration by up to 46% and had a 15% possible risk of inflicting complete damage to the controlled structure of a low-rise building. Its installation in high-rise structures resulted in the suppression of 26%. Consequently, these findings are crucial in enhancing the safety of building structures and residents, and minimizing disaster damage costs.
computer science, information systems,telecommunications,engineering, electrical & electronic