Shih Yu Chu

Real-time active control verification via a structural simulator

Abstract Verification of active control algorithms prior to their implementation to real structures has always been a time-consuming and costly process. At the minimum, they need to be examined through analytical simulation. Laboratory testing using a scaled-down model on a shaking table or in a wind tunnel is also often necessary. This investment in experimental hardware and inherent inaccuracies in experimental set-up or modeling can be a major obstacle in structural control implementation. In this paper, an alternative real-time verification procedure via a simulator is proposed. In this procedure, the control strategy is implemented inside a dedicated PC controller with an independent Digital Signal Processor (DSP) to calculate the required control force. The DSP performs the function of data acquisition through communication with an Analog to Digital Converter (A/D) and a Digital to Analog Converter (D/A). Another PC structural simulator with a similar dedicated hardware is constructed to emulate the real-time response of a theoretic structure with the Hybrid/Active Mass Damper (HMD/AMD) in order to verify the real-time control effect. The integrated system is dedicated to verify the effectiveness of the proposed digital control algorithm and to provide also a preimplementation testing base to test the functions required for practical utilization.

Modeling and experimental verification of a variable-stiffness isolation system using a leverage mechanism

Recent studies have discovered that conventional isolation systems may incur excessive isolator displacement in a near-fault earthquake with strong long-period wave components. To overcome this problem without jeopardizing isolation efficiency, a novel semi-active isolation system called a Leverage-type Variable Stiffness Isolation System (LVSIS) is realized in this study. By utilizing a simple leverage mechanism, the isolation stiffness of the LVSIS can be easily controlled by adjusting the position of the pivot point on the leverage arm. For accurate analysis, the dynamic equation based on a mathematical model that considers the actual situation of all friction forces within the LVSIS is derived in the study. The mathematical model is then verified experimentally by using a prototype LVSIS tested dynamically on a shaking table. Furthermore, to determine the on-line pivot position of the LVSIS, this study also proposes a semi-active control law whose feedback gain is decided by utilizing a linear active control algorithm, such as the LQR or modal control. By comparing the isolation performance of its uncontrolled passive counterpart, the test results also demonstrate that the LVSIS with the proposed control law is especially effective in suppressing the excessive base displacement induced by a near-fault earthquake.

Discrete time point-to-point control of flexible structures

The primary objective of this work is to investigate linear time invariant systems undergoing rest-to-rest maneuvers in a finite time using the discrete time domain approach. Using a given sampling period, the governing equations of linear systems are first discretized into the equivalent discrete time domain representation. To decouple the resulting difference equations, the system equations are converted into the Jordan canonical form by using a similarity transformation. The decoupled Jordan canonical equations are converted to a set of algebraic input/output equations with embedded end-point conditions, by a recursive approach. The optimal or suboptimal control profiles required to achieve the desired maneuver can be easily calculated through basic manipulation. The sensitivity of the design to the uncertainties in the system parameters is reduced by introducing sensitivity equations, and the design is found to be robust to these uncertainties.

In situ test of school buildings retrofitted with external steel-framing systems

AbstractIn situ push-over experiments and analytical assessments were conducted in this study to investigate the seismic capacity of RC school buildings at the Guan-miao Elementary School in Tainan City, Taiwan. Both prototype and external steel-framing specimens were constructed and compared. Reinforcements with steel channels were adhered to the captive columns and their adjacent beams in the weak direction, to retrofit the school building through the creation of external steel-framing systems. The design and construction procedures of the beam-column connection are introduced, as well as detailed drawings of the column-base anchorage system, both of which are key to the effects of the proposed retrofitting method. The experimental results show that the steel-framing system can be firmly bonded to the retrofitted specimen, providing continuous stress transference from the beam-column joints to the steel channels. The failure mode of the retrofitted captive columns was shifted from shear failure, as seen...

Time delay effect on direct output feedback controlled mass damper systems

Investigates the stability of a single degree of freedom (SDOF) system with an optimal direct output feedback controlled mass damper. An active mass damper system can take the form of a hybrid mass damper (HMD) or a fully active mass damper (AMD) depending upon imposed design constraint resulting from space, strength and power limitations. The control effects of the HMD and the AMD are first examined. The continuous-time direct output feedback control algorithm (LQR) is used to find the required control force. In active control, the control force execution time delay can not be avoided or eliminated even with present technology, which can be critical to the performance of the control system. The influence of time delay for a SDOF model with both HMD and AMD systems using continuous-time control gains is then discussed. Explicit formulas and numerical solutions to determine the maximum delay time which causes onset of system instability is obtained.

Interaction Effects of Detuning and Time-delay on Generalized Hybrid Mass Damper Systems

The vibration mitigation performance of a conventional tuned mass damper (TMD) system is very sensitive to the fluctuation in tuning of the designed frequency to the natural frequency of the main system. In view of the stochastic characteristics of external loading and the errors of identifying system parameters, a hybrid mass damper (HMD) system with optimal selection on control parameters can enhance the designed performance with the help of a supplementary active force acted between the main system and the mass damper. However, the control performance degradation induced by the control force application time delay should be considered and investigated before practical application. Having the best versatility in mind, the generalized hybrid mass damper (GHMD) systems are reclassified and investigated in this paper to examine diversifying needs. The variations in their modal properties and tuning effects with respect to delay time are addressed based on an explicit formula derived from a delayed exponential characteristic equation (DECE). It is found that the active control force with delay-time chosen as the natural period of the system in a HMD based on an uncompensated output feedback scheme will keep the robustness on response reduction performance. This advantage degrades for non-optimal systems with delayed control force due to the detuning effects. A detailed study is addressed to demonstrate the interaction effects of detuning and time-delay in this study. In addition to modal properties, a discrete-time state-space approach is applied to verify the control performance in time domain subjected to earthquake excitations.