ABSTRACT
In this work, particle swarm optimization and the maximum dominant period approach are used to improve on the conventional linear quadratic Gaussian (LQG) controller. To obtain an appropriate command signal, the LQG controller's control weighting matrix is varied. This command signal is sent to the magnetorheological (MR) damper, which generates the appropriate amount of counter-control force for seismic vibration mitigation. The proposed controller is evaluated using a benchmark three-story structure with an MR damper at the base. On this structure, the proposed controller is tested under three different earthquake time histories, different soil conditions, and a situation wherein power is lost at the peak of the earthquake. According to the outcomes and discussion, the proposed control strategy outperforms the conventional LQG controller.
