Study On Control Algorithm And Simulation For Smart Base Isolated Benchmark Building

Combining active or semi-active control devices with base isolation system results in smart isolated system, which can improve the overall performance of passive isolated structure effectively. Control algorithm works as the theoretic basis of smart isolated control and determines the seismic mitigation effect to a great extent. Therefore, evaluating various kinds of control algorithm through a normal, acceptable smart isolated model is with great theoretic value and operation significance. In view of this, ASCE, Gavin of Duke University and Johnson of South California University presents the smart base isolated Benchmark building.Taking the Benchmark structure as object of study, characteristics of this model are analyzed, and motion equations of this structure are derived considering bidirectional earthquake excitation in this paper. Deriving methods of several optimal control algorithms, i.e., Classical Optimal Control, Linear Quadratic Gaussian Control, Instantaneous Optimal Control and Sequential Optimal Control are presented and their features are discussed. This paper also sets up a simulation model of the smart base isolated Benchmark building with Simulink software, meanwhile, Simulink simulations are realized for Linear Quadratic Gaussian and Sequential Optimal Control algorithms. Simulation results of the two algorithms mentioned above are analyzed by employing the simulation model of the Benchmark building as verifying platform. Some response indices defined for this Benchmark problem are adopted as evaluation criteria to evaluate the control performance, and thus the following conclusions can be obtained: both kinds of control algorithm can suppress the response of isolated building when vibration control is performed, and thus means that smart base isolated system is more effective than passive isolation system;under comparison with Linear Quadratic Gaussian Control algorithm, the reduction of structural response for Sequential Optimal algorithm is more obvious and the control forces calculated by the both algorithms are almost on the same level, so it can be shown that Sequential Optimal Control is more advantageous in terms of control efficiency.