Optimal control of structures using neural networks

Time delay in structural active control is one of the most important problems that face the active control in civilstructures. This thesis introduces a new technique in overcommg the time delay in active control cycle. Thistechnique uses an Artificial Neural Network to estimate the future earthquake record for a number of ahead steps online.By knowing in prior the coming forces and starting from thecurrent state of the controled structure, the future response iscalculated and the required control force is estimated. In thisway the control force will be applied at nearly the same statefrom whice it was calculated. This algorithm can be joinedwith any control law and any control device to overcome itsinherent time delay. In this work optimal control withtendon controller is used. Different structures and differentearthquakes were used to study the effect of time delay andto prove the efficiency of the proposed technique in compensating it.