Robust output feedback-based neuro-fuzzy controller for seismically excited tall buildings with ATMD accounting for variations in the type of supporting soil

Abstract

Recently, developing the control algorithms for tuning the voltage/force command of smart control devices for application in seismic-excited buildings and bridges has been well addressed considering soil-structure interaction (SSI) effects. A controller is often designed for a soil pre-set condition. Moreover, considering an exact model of soil in the controller design process is a tedious task due to high computational costs. Consequently, the seismic performance and robustness of the adopted controller are questionable against the variations in soil conditions. The motivation of the present paper is to propose a robust controller for seismic-excited tall buildings against the variations of soil conditions than the pre-set condition. Taking advantage of the optimal output feedback controller (OOFC), a simple and practical controller based on OOFC is proposed for application in tall buildings equipped with an active tuned mass damper (ATMD) accounting for SSI effects. Considering different soil types, a dataset is generated from the proposed OOFC and an adaptive neuro-fuzzy inference system (ANFIS) is trained to fit them in a robust dataset. Then, a robust output feedback-based ANFIS controller, namely ROFBANFISC, is proposed. The robustness of the OOFC and ROFBANFISC are validated for different soil types. Considering six performance criteria, the numerical studies are conducted on a 40-story structure subjected to five real ground motions. The simulation results carried out for different soil conditions show that the proposed ROFBANFISC is more robust than the OOFC. Also, it can maintain its appropriate seismic performance in different soil types.