Improvement the performance of a simulated earthquake table using evolutionary adaptive filtering and PID controller

Shaking simulation table is one of effective tools for analysis of the structures' behavior against earthquake input. The most important part of this table is a control system that tracks displacement profile, velocity and acceleration of a real earthquake using encoder and accelerometer sensors. However, some uncertainties, such as changing inertia of the table regarding installation of different structures on it, the dynamics that are not modeled, and the table sensors' noise cause decrease in precision of the control system. To solve these problems, in this paper the performance of this simulated table is improved by combining the output of sensors using an evolutionary adaptive filters EKF and UKF and two series PID controllers. In the proposed method, evolutionary adaptive filters EKF and UKF is used to decrease uncertainties and sensors' noise and also for estimation of the table parameters. PID controllers use the linear velocity estimated by the filters to stabilize, track the reference input, and remove the external disturbances of the table. The performance of the proposed method is evaluated in a simulated environment and laboratory test. The result shows the effectiveness of the proposed method.