| With the improvement of design theories, the application of the advanced analysis technique, the progress in high-strength building materials and the perfection of construction technology, there are more and more long-span bridges and high-rise buildings. Such structures are important to not only the people's livelihoods but also the country's economic lifelines. The serious damage or collapse of these huge structures will cause great economic loss and social impact due to natural disasters or man-made accidents. As the sizes of structure become bigger, the performance of structures will change accordingly. In some cases, because the stiffness and damping decease dramatically, the dynamic response of structures become greater under strong earthquake and high speed wind, thus the safety and reliability are reduced. In this way, it is not only feasible but also necessary to apply structural control techniques to effectively promote the safety and reliability of the structure against natural hazards. As the vibration control system of large structures is a high dimensional system, the control system would be rather complicated. The decentralized control techniques show their advantages in addressing these high dimensional issues, which is an innovative point in this thesis.Main research contents of this dissertation are as follows:Firstly, the hierarchical decentralized and structural perturbation decentralized control methods are used for building vibration control. The structure model is a building with two adjacent benchmark buildings, one with 9 stories and the other 20. The results show this two decentralized methods have good control effects.Secondly, a new design method based on the optimization of control parameters and actuators placement is proposed with regards to the characteristics of the hierarchical decentralized control method. By using genetic algorithm, the optimized control parameters and actuators placement are obtained based on the minimization of active control force of the close loop system. Meanwhile, the control performance indexes are satisfied.Thirdly, robust control theory is applied to the control of buildings with uncertainty; Hierarchical decentralized control is analyzed based on linear matrix inequality; Hierarchical decentralized robust H∞control and decentralized robust H∞control design methods are proposed. The results show these two methods can be effectively used in the control of buildings with uncertainty.Fourthly, based on Lyapunov stability method, the fault tolerate control methods of standard building and building with uncertainty are both investigated. Hierarchical decentralized fault tolerate control algorithms are proposed by using switch matrix when sensors and actuators of the control system fail simultaneously. The results demonstrate the proposed fault tolerate control method can guarantee the asymptotically stability and robustness of the system, and has good control effect.Fifthly, the damage detection method under hierarchical decentralized control is analyzed and the one based on sensitivity of frequency square is proposed. The results show the proposed method can improve the sensitivity of frequency to damage, and reliable and accurate damage detection is obtained, by comparing the frequency square before and after damage.Finally, hierarchical decentralized control experiments are implemented based on MATLAB/xPC real time control method. A cantilever active control experiment using piezoelectric material is set up. The vibration control of hierarchical decentralized robust control and hierarchical decentralized fault tolerate control are analyzed, and control effects of vibration are compared with each other with two cases of initial disturbance and continue excitation, respectively. The results demonstrate the hierarchical decentralized robust control has much better control effect and robustness even when there exist uncertain parameters and failure of actuators and sensors in the control system.
|
PDF Full Text Request