Optimal control of linear buildings under seismic excitations

Abstract

SUMMARYConsiderable e ort has been devoted to develop optimal control methods for reducing structural response under seismic forces. In this study analytical solution of the linear regulator problem applied widely to the control of earthquake-excited structures is obtained by using the su cient conditions of optimality even though almost all of the optimal controls proposed previously for structural control are based on the necessary conditions of optimality. Since the resulting optimal closed-open-loop control cannot be implemented for civil structures exposed to earthquake forces, the solution of the optimal closed-openloop control is carried out approximately based on the prediction of the seismic acceleration values in the near future. Upon obtaining the relation between the exact optimal solution and future values of seismic accelerations, it is shown numerically that the solution of the optimal closed-open-loop control problem can be performed approximately by using only the ÿrst few predicted seismic acceleration values if a given norm criteria is satisÿed. Calculated performance measures indicate that the suggested approximate solution is better than the closed-loop control and as we predict the future values of the excitation more accurately, it will approach the optimal solution.

Predictive instantaneous optimal control of inelastic structures based on ground velocity

“…It is noted that recent research has shown that earthquake ground acceleration can be accurately predicted for up to six time steps (Aldemir et al, 2001). Since the ground velocity is not of high frequency as compared to the ground acceleration, it is highly possible to predict the ground velocity for more than six time steps when an appropriate prediction algorithm is used.…”

Active Tendon Control of Torsionally Irregular Structures under Near‐Fault Ground Motion Excitation

“…Aldemir et al. () developed an approximately optimal closed‐open‐loop control based on the prediction of near‐future excitations provided that a given norm criteria is satisfied. Aldemir and Bakioglu () analytically solved the modified linear quadratic regulator (MLQR) problem including a parameter representing system stability order in the presence of unknown seismic excitations.…”

Absolute-energy-based active control strategies for linear seismic isolation systems

“…Optimal control methods may be classified into several categories, i.e. classical optimal control [24][25][26][27][28][29], H 2 and H N control [30][31][32] and instantaneous optimal control [33][34][35]. The classical optimal control minimizes a performance index defined as the integration of a quadratic expression with respect to the state vector and the control vector.…”