MODIFIED PREDICTIVE OPTIMAL LINEAR CONTROL OF STRUCTURES IN SEISMIC REGION*

A modified predictive optimal linear control (MPOLC) algorithm is proposed for controlling the seismic response of elastic structures. This algorithm compensates for the time delay that occurs in real control applications by predicting the structural response in the modified optimal linear control equation. Since the environmental loads and disturbances are not measured during real-time control, they are not involved in the derivation of the control algorithm. Therefore the predictive optimal linear controller (POLC) is a proportional feedback of the only predicted current state. In the modified control algorithm (MPOLC), using a logical assumption, the immeasurable disturbances are considered in the state space equation and also in the derivation of the control algorithm, so that the controller is a combination of the control force in the last step and the proportional feedback of the predicted states in the last two steps. Hence, the control performance of the modified control algorithm is superior to that of the original one. The feasibility and effectiveness of the proposed control algorithm is verified through frequency- domain and time-domain analyses, and compared with the original one. The tendon control system of a three-degree-of-freedom structure is illustrated to demonstrate the control effectiveness of the modified predictive control algorithm.