Development of Nonlinear Control Algorithms for Shaking Table Tests

Abstract

Shaking table testing method is one of the main sources of experimental means to evaluate the dynamic response of structural systems under earthquake loads. The experimental technique produces nearly realistic prototype conditions, giving important insight into critical issues such as collapse mechanisms, component failures, acceleration amplifications, residual displacements and post-earthquake capacities. Traditional tuning of shaking table relies on the use of linear controllers, which are designed to regulate linear systems. With most of the specimens being tested to highly nonlinear states (to understand the structural response under extreme loads), traditional linear controllers can no longer control the shaking table effectively. This results in experimental errors between commanded and measured shaking table movements which may produce an unintended response of the tested structure. Ultimately, it may result in a pre-mature failure of the specimen. To address this issue, a Lyapunov–based nonlinear control algorithm is utilized to develop an enhanced shaking table control system, which is based on nonlinear models accounting for the nonlinear response of the hydraulic actuator and specimens. A one-sixth-scaled model has been developed and constructed in the laboratory at the University of British Columbia, Vancouver. Advanced nonlinear system identification techniques have been developed to create the numerical model capable of recreating the nonlinearity experienced by the laboratory setup. Simulation results indicate that the developed nonlinear control algorithm can be used to achieve excellent tracking, even when the tested structure behaves nonlinearly. The example also demonstrates the ability of the nonlinear controller to compensate for disturbances in the actuating force applied to the shaking table. Thus, the proposed nonlinear shaking table control algorithm is not only a viable alternative, but also a way to significantly improve the quality of shaking table tests.