The Inverse System Predictive Controller For Structure Vibration

Vibration control is different from traditional seismic design, which resists external power by changing structure characteristic itself. Instead, it suppresses earthquake and wind responses by the active or passive control systems. Although modern control theory can solve many complex problems of system, such as stability, controllability, observability and so on, the accurate mathematical model is needed. However, most civil engineering structures are such systems, whose accurate mathematical models cannot be found easily. Therefore, we need a control strategy without accurate mathematical models.This dissertation focuses on the research of inverse system and neural network, as well as its application in structural vibration control. By this way, an inverse system predictive controller was built, which is the control strategy we need. Afterwards, the controller was used in the structural vibration controlling of a high-rise construction. The numerical simulation was used to simulate the earthquake responses of construction model controlled by TMD and HMD. Finally, the inverse system predictive controller was contrast with three common active control algorithms on several different points.The simulation results showed that the inverse system predictive controller could play good roles on suppressing earthquake responses of such a building. At the same time, the contrasting of different active control algorithms may be useful for similar control problems of other high-rise structures.