This study proposed optimization procedures to design an LQR controller for an active tuned mass damper on a 10-story structure. For the optimization, two multiobjective-function problems were formulated. The number of objective functions in both problems was equal to the number of stories, and they measured the ratio of controlled to uncontrolled drift. Optimizations of the ATMD have been realized by utilizing 28 near-field earthquake records with pulse ground motion. The performance of the resulting controller was assessed using five performance indices by utilizing 96 earthquakes comprised of near field with a pulse, near field without a pulse, and far-field records. The first optimization problem has no bounds on the magnitude of the applied force. Frequency analysis has been used along with time domain analysis to assess and figure out the characteristics of the controlled structure. The results indicate that a high amount of active force is needed. Several methods were tested to find the most effective way to decrease the needed DFWXDWRU¶V force while keeping a good performance index similar to the original model. In the second optimization problem, a limitation was considered for the applied force. In this problem, the time simulation and frequency analysis have been used as in the first one. The force limitation in this problem triggers a bang-bang action issue. Several low-pass FIR filters have been tested against the issue, resulting in a better understanding of the originating reason for the bang-bang action and the filters' effect on the controller. To decrease the number of sensors used for the feedback system Kalman filter has been used. The output of Kalman filter was the same as the original system. The robustness of the controller was assessed by changing the characteristics of the uncontrolled structure and comparing it with the original model. It turns out that the optimized LQR-ATMD is robust.