Numerical and Experimental Study of Seismically Excited Scaled Structure with Active Mass Damper

In recent years, the development and implementation of artificial intelligence (AI) have attracted tremendous attention. The implementation of active control systems for building structures can be improved by using an AI controller. Non-AI controllers such as the Linear Quadratic Regulator (LQR) controller require full state variables of the structure to be measured, which is rarely feasible. To address this problem, two AI models, namely, artificial neural network (ANN) and fuzzy logic (FL), have been tried as AI-based controller in various studies. In the present study, both AI models were investigated to see their practicality and effectiveness. The AI models were implemented to control an active mass damper (AMD) in a three-story prototype-sized building. The simulation results from the structure with an LQR controller were used as benchmark and training data for the AI models. The results of the study demonstrated that although both AI models could reduce the structure responses, ANN was more practical and effective compared to FL as an AI-based controller for the given structure. Furthermore, the effectiveness of an ANN-based AMD was also shown by the experimental results.