Akademik Veri Yönetim Sistemi
International Conference on Applied Sciences and Engineering: ICASE 2025, İstanbul, Türkiye, 01 Ekim 2025, ss.211-219, (Tam Metin Bildiri)
Tuned liquid damper (TLD) systems are simple and economical designs used to reduce the response of structures exposed to various vibrations. These devices dampen vibration by utilizing the oscillation of their own solid mass through the sloshing of the liquid mass they contain. Their design is effective in ensuring structural control. The parameters contained in the structure and control systems have different effects on the structural response. Predicting the behavior of a structure with a control system in response to various vibrations is possible using artificial intelligence methods. This study aims to predict the structural response based on the system characteristics and mass ratio of a TLD to be added to the structure for the control of structures with different periods and natural damping. Models predicting the structure's response to seismic vibrations were developed using machine learning methods. For this purpose, the seismic response of the structure was recorded using combinations of structures and dampers with different characteristics under a far-fault earthquake. The results obtained were converted into a dataset and models were produced using ensemble algorithms for use in machine learning. The performance metrics of the models were compared to determine the best algorithm. The final prediction model created explained the variance in the data with an R2 score exceeding 99% level and showed high prediction performance with a mean deviation error of less than 5% from the actual value.