State Rebuilding And Robust Controller Design Of AMD Control System

In the high-rise building AMD control system design, how to reduce the impact of computing time consuming, how to obtain more easily measured feedback signal, how to design AMD control system which can be considered structured uncertainties are still difficult engineering applications. AMD control system has some following problems: Firstly, When the controlled structural model’ dimensionality is reduced, structural information which is corresponded to system’s small characteristic values is directly abandoned, as a resulting, dimensional reduction processes’ precision will be reduced; secondly, acceleration signal of the structure is more easily to get than displacement and velocity, how to ensure the accuracy of the controller based measured acceleration signal feedback; lastly, structural uncertainties, such as structural rigidity, mass, etc., will reduce the control effect. In response to those problems, the following work has been done:In order to solve the problem of reduced accuracy in the process of construction dimensionality reduction, by balanced truncation method and pole assignment algorithm, the paper propose to a new low-dimensional controller design method to consider the structure discarded higher-order information, including higher order frequency, high order vibration damping ratio and higher order modes of the original structure. Using ten-story frame structure as an example, analyze the reduced structural model how to effect on the properties of AMD control system. Finally, set up a four-frame model of AMD control test system to prove the correctness of the numerical analysis results.For the control system dependent deeply on the number of sensors, this paper derived the system state estimation matrix, and then by the estimated acceleration response as feedback, the new control method is based on parts of floors acceleration response. Based on this method, analyze the number of sensors and their arrangement how to affect the performance of AMD control system. Finally, ten-story frame structure’s numerical simulation and four layers framework model test verifies the correctness of the conclusions.For the robust control problem that structure has uncertain parameters, combining the LMI technology and regional pole placement algorithm, derive to LMI considered the structural parameter uncertainty, and find a suitable area for closed-loop system poles configuration, the paper propose to a robust controller design method based on regional pole assignment algorithm. Ten-story frame structure’s numerical simulation and four layers framework model test verifies the robust controller has better effectiveness and robustness than the accurate pole placement controller.Using AMD low-dimensional controller design method considered higher modes of information, state observer design method based on the parts of floors acceleration response and robust controller design method based on regional pole assignment to improve KK100 AMD control system’s performance. Through analysis identified some important parameters include Kk100 building suitable dimensionality, the number of the acceleration sensors and their arrangement, etc., and finish the robust controller design for KK100.