This paper focuses on the design of adaptive-neural interval type-2 fuzzy logic controller (AN-IT2FLC) in an active tuned mass damper to reduce the response of building under seismic excitation. One of the main shortcomings of interval type-2 fuzzy logic controller (IT2FLC) is its need to adjust in any earthquake. This is whilst the AN-IT2FLC can solve this problem using the training process. In this research, four inputs are used for designing and training of AN-IT2FLC as controlled displacement and velocity of roof level with IT2FLC, acceleration of the implemented earthquakes and the control force of IT2FLC. AN-IT2FLC training performed based on the eight earthquake records of El Centro, Hachinohe, Kobe, Northridge, Loma Prieta, Tabas, Morgan Hill and Erizkan with various seismic characteristics. In order to investigate the effectiveness of the proposed controller, an 11-story building with ATMD on its top floor analyzed under another four ground accelerations of Chi-Chi, Kern-county, Coalinga and Coyote-lake records. The results revealed that ATMD with AN-IT2FLC is able to achieve more response reduction with higher speed and accuracy rather than that of the IT2FLC.

Keywords