Optimal control of nonlinear frames by Newmark and distributed genetic algorithms

Active control of multi-storey frame structures with nonlinear hysteretic response has been studied in this paper. A new nonlinear optimal control algorithm based on nonlinear Newmark integration method and distributed genetic algorithm (DGA) has been developed. The objective has been to reduce the response to below any desired level. In this study, DGA has been used to determine the weights in the control law corresponding to displacement, velocity and acceleration. The capabilities of the method has been assessed through simulation where an eight-storey frame with bilinear hysteretic behaviour has been subjected to a white noise ground acceleration and controlled by the controllers designed by the new algorithm. An active tendon control system comprised of prestressed tendons and either a single actuator, which could apply a force to one of the floors, or eight actuators, which could fully control the frame, has been used. Also, the optimum actuator placement in the single actuator case has been studied. The designed controllers have been tested under both near and far-field earthquakes and have performed successfully. Copyright (C) 2009 John Wiley & Sons, Ltd.