Search

Neural networks (NNs) can provide a simple solution to complex structural vibration control problems. However, most past NN-based control strategies cannot guarantee an optimal policy in structural vibration control. In this study, a novel active vibration control strategy based on deep reinforcement learning is proposed, which utilizes the learning ability of NN controllers and simultaneously provides control performance comparable to traditional model-based optimal controllers. The proposed learning algorithm can determine the control policy through interaction with the environment without knowing dynamic system models. This study shows that the proposed model-free strategy can provide optimal control performance to various systems and excitations. The proposed control strategy is first verified on a single-degree-of-freedom model and subsequently extended to a multi-degree-of-freedom shear-building model. Its control performance with full-state feedback is nearly the same as that of a classical linear quadratic regulator. Moreover, the learned policy can outperform a traditional output feedback controller in a partially observed system. The robustness of the proposed control strategy against measurement noise is also tested.

SUMMARY Structural health monitoring (SHM) system provides an efficient way to the diagnosis and prognosis of critical and large-scale civil infrastructures like long-span bridges. This paper presents a long-term condition assessment approach of suspenders in a cable-suspension bridge under in-service traffic loads based on structural monitoring technique. The traffic loads identified from a monitoring system, including both highway and railway traffic loads, and the finite element model of the bridge are employed to determine the axial force response of the suspender. The stochastic axial force response in the suspender is described by a filtered Poisson process, through which the maximum value distribution of axial forces in its design reference period can be derived using the Poisson Process theory. In this paper, the long-term deterioration process of steel wires in the suspender considers simultaneously the uniform and pitting corrosion and the corrosion fatigue induced by both cyclic loading and environmental attack. Such a stochastic and coupled corrosion fatigue process of steel wires is simulated using the Monte Carlo method, and the time-variant conditions of the suspender are subsequently assessed in a probabilistic way, such as crack depth, number of broken wires, ultimate strength, etc. In particular, two load conditions—the train loads alone and the combination of train load and road traffic load—are examined within this procedure in order to investigate their respective effects on the deterioration. By employing first-order reliability method, the reliability indexes of the suspender under the traffic loads are further estimated in terms of the safety under the extreme traffic load distribution in the design reference period and the serviceability specified in the design specification. The discussions of the life-cycle reliability indexes of the suspender provide guidance to the future decision making related to maintenance and replacement of suspenders, and it may also shed light on the long-term condition assessment of other structural members. Copyright © 2010 John Wiley & Sons, Ltd.

This paper presents a simulation-based benchmark control study in which shape memory alloy (SMA)-based displacement restraining damping devices are proposed to control the seismic response of a full-scale three-dimensional seismically excited highway bridge. In this Phase II benchmark problem, the bridge is fully isolated at both the abutments and the central pier location using nonlinear hysteretic bearings with a lead core on the inside and an elastomer surrounding the lead core. The SMA restraining damping device is a passive control device employing superelastic Cu–Al–Be alloy wires as its core re-centering component, which restrains the base-isolated bridge from excessive displacement responses, especially under extreme earthquake events. In this benchmark study, a total of 20 such passive control devices are supposed to be installed at the isolation level between the deck and the isolators on bridge piers and center column at 10 locations, each location consisting of a single orthogonal pair to control the responses in both directions. The performance of the passive control devices is analyzed in terms of the performance indices in the benchmark problem definition at a variety of ambient environment temperatures at 23, 0, −25 and −50°C, respectively. The results of this simulation-based benchmark control study show that the proposed passive control device can effectively reduce the excessive displacement responses and permanent bearing deformations of the benchmark base-isolated bridge subjected to strong ground motions, and temperature seems to have little effect on the performance of the superelastic Cu–Al–Be restraining damping device in bridge response control. Copyright © 2009 John Wiley & Sons, Ltd.

Summary This study investigates a novel energy regenerative tuned mass damper (TMD) with dual functions—vibration control and energy harvesting—in a high-rise building. The energy regenerative TMD consists of a pendulum-type TMD, an electromagnetic damper, and an energy-harvesting circuit. A simple optimal design method for energy regenerative TMD is proposed, in which a fixed duty-cycle buck-boost converter is employed as the energy-harvesting circuit to optimize the energy-harvesting efficiency and damping coefficient of the TMD. This study is organized into two main tasks: (a) characterizing and modeling the energy regenerative TMD through laboratory testing of a scaled prototype and (b) evaluating the vibration control and energy-harvesting performance of the energy regenerative TMD when applied in a 76-story wind-excited benchmark building in consideration of the nonlinearities in the energy regenerative TMD. The simulations reveal that the harvested electric power averages from hundreds of watts to kilowatts level when the mean wind speed ranges 8–25 m/s. Meanwhile, the building vibration is mitigated with the control performance comparable to the optimally designed passive TMD in a wide range of wind speed. The results in this study clearly demonstrate the effectiveness of the dual-function energy regenerative TMD when applied to building structures.