Optimal structural control using neural networks

An optimal control algorithm using neural networks is proposed. The control ler neural network is trained by a training rule developed to minimize cost function. Both the linear structure and the nonlinear structure can be con trolled by the proposed neurocontroller. A bilinear hysteretic model is use d to simulate nonlinear structural behavior. Three main advantages of the n eurocontroller can be summarized as follows. First, it can control a struct ure with unknown dynamics. Second, it can easily be applied to nonlinear st ructural control. Third, external disturbances can be considered in the opt imal control. Examples show that structural vibration can be controlled suc cessfully.