Active vibration control of building structure using active disturbance rejection control

Abstract

This article presents an active vibration control of seismically excited building structures. The control scheme is based on active disturbance rejection control, which is an attractive alternative technique for structural vibration suppression and practical motion control solution in the presence of parametric uncertainties and disturbances. The proposed active disturbance rejection control scheme uses a generalized proportional integral observer, which allows us to estimate in real time the unknown dynamics and disturbances in the building structure to cancel their effect using a part of the control signal. First, the active disturbance rejection control provides a proportional derivative controller with robustness to external disturbances and uncertainties, and its structure is expressed in a compact error-based form. An important advantage with respect to other methods is that the proposed scheme does not need the system parameters. Moreover, supposing that displacement and velocity cannot be measured directly, an online robust adaptive observer is introduced to estimate both data, required for the proportional derivative controller. The adaptive observer removes constant disturbance and attenuates measurement noise in acceleration data. Under this line, a second active disturbance rejection control scheme is introduced based on a proportional controller that, unlike proportional derivative, it only needs the velocities that can be directly estimated by integrating the acceleration signals and does not require the adaptive observer. An advantage of this scheme is its simplicity to be implemented because it only needs to tune the proportional gain. Furthermore, this scheme has a similar performance of the proportional derivative controller. The effectiveness of the proposed active disturbance rejection control schemes is demonstrated through experimental results of a reduced scale five-story building structure. The results are found to be a good step in that direction, confirming that the proposed method is promising for practical applications.