Predictive Structural Vibration Control Using Neural Network

It has been proved by theoretical analyses and experiments that structural control as a new technology is a valid method to reduce the influence of seismic disaster. Domestic and international researchers have proposed many theoretical analyses of control algorithms which appear to be the cases. However, due to the specialities of different structures, it is difficult to obtain the mathematical model of controller accurately. The study of intelligent material and intelligent control method corrects the disadvantage of traditional control theory and opens up a new field for structure control technology.In this paper a system identification method based on the technology of the Artificial Neural Network(ANN) is proposed, which establishes the identification model reflecting the dynamic characteristics of structure accurately. Based on the peculiarities in setting up the model of predictive control and the technology of ANN, a new kind controller is designed which avoids the disadvantage of traditional calculating methods such as LQG; Time-delay can result in negative influence to structural controlling, the Neural Network estimates the structure's response and sends out control signals in advance, then the structure's response is reduced. Because of the main disadvantage of current vibration control devices, a new kind of damper made from intelligent Magnetorheological (MR) is designed to control the structure semi-actively. A simple and effective control method is proposed, which is based on the predictive control theory. The results of computer simulation indicate that the method proposed in this paper is valid and practicable.