Recent decades have witnessed much interest in the application of active and semi-active control strategies for seismic protection of civil infrastructures. However, the reliability of these systems is still in doubt as there remains the possibility of malfunctioning of their critical components (i.e. actuators and sensors) during an earthquake. This paper focuses on the application of the sliding mode method due to the inherent robustness of its fault detection observer and faulttolerant control. The robust sliding mode observer estimates the state of the system and reconstructs the actuators' faults which are used for calculating a fault distribution matrix. Then the fault-tolerant sliding mode controller reconfigures itself by the fault distribution matrix and accommodates the fault effect on the system. Numerical simulation of a three-story structure with magneto-rheological dampers demonstrates the effectiveness of the proposed fault-tolerant control system. It was shown that the fault-tolerant control system maintains the performance of the structure at an acceptable level in the post-fault case.

SUMMARY Consistency of real‐time data transmission is an important factor in active and semi‐active control of structures. Recent advances in wireless sensors and their reliable data communication have attracted many attentions in structural control engineering. Wireless sensors have been also preferred to wire‐base types because of problems associated with installing extensive length of coaxial wires. In addition, embedded computing technologies in wireless sensors as local controller reduce size of transferring data and provide stable transmission. However, there are some issues in implementation of wireless connection like communication range and latency. This study presents the feasibility of developing a decentralized controller by using H2/LQG as robust‐optimum algorithm in cable‐stayed bridge. Because of large scale of structure, the order of model is reduced by using two different reduction methods in finite time and frequency. The controller is developed in centralized and decentralized solutions. The decentralized controller through developing a gradient‐based method is designed to efficiently solve system of closed‐loop H2 optimization problem. The structural responses in centralized and decentralized solutions are compared with each other under three different earthquake records considering time delay in data transmission. Results indicate that for large‐scale bridges with time‐delayed data communication issue, H2/LQG controller in decentralized solution gives maximum reduction of response indices in comparison with other strategies. Copyright © 2011 John Wiley & Sons, Ltd.

In recent decades, the application of semi-active control strategies has gained much attention as a way to reduce the seismic response of civil infrastructures. However, uncertainty in the modeling process of systems with possible partial or total failure during an earthquake is the main concern of engineers about the reliability of this strategy. In this regard, adaptive control algorithms are known as an effective solution to adjust control parameters with different uncertainties. In the current study, the efficiency of the simple adaptive control method (SACM) is investigated to control the seismic response of building structures in the presence of unknown structural damage and fault in the sensors. The method is evaluated in 20-story steel moment resisting frames with different arrangement of smart dampers and sensors with various damage and fault scenarios. The results show that the SACM control system can effectively reduce the maximum inter-story drift of the structure in all different assumed magnetorheological damper arrangements. Furthermore, combination of a Kalman–Bucy filter with the SACM improves robustness of the controller to the uncertainties of sensors faults and damages of structural elements.

Recent developments in semi-active control technology have led to its application in civil infrastructures as an efficient strategy to protect susceptible structures against seismic and wind induced vibration. The reliable and robust performance of semi-active systems depends on the level of uncertainties in the structural parameters as well as on the sensors’ measurement and on smart mechanical dampers. A common source of uncertainties in semi-active control devices is related to the inherent nonlinear nature of these devices, thermal variation, or their malfunctioning. This study deals with the robust H∞ control problem and aims to model different sources of uncertainty. The uncertainty of the structural model and damper force are assumed to be norm bounded random variables. By using linear fractional transformation (LFT), the uncertain part of the system is decoupled from the nominal parameters of the system. The robust H∞ controller is designed to achieve consistent performance in structures including nominal and perturbed dynamics. Additionally, to reduce the uncertainty of the damper force, an inverse model of the magnetorheological (MR) damper is developed based on an adaptive neuro-fuzzy inference system (ANFIS). The robustness of the proposed algorithm is validated by numerical simulations.

Adaptable active control strategies besides advance sensors and actuators technologies lead to higher performance of vibrational control in civil infrastructures under severe ground motions. These resilience control systems are robust against model uncertainties as well as being online recoverable from the malfunctioning of sensors and actuators. In this study, resilient control system based on sliding mode (SM) fault detection observer and SM fault tolerant control is improved for actuator fault in large‐scale systems. The SM fault detection observer is modified for eliminating the excessive chattering in estimating states and actuators’ fault, and the reconfigurable SM fault tolerant control is improved to minimizing input forces in H∞ control framework under seismic action. Design of observer and controller is performed using linear matrix inequalities. Numerical simulations on the cable‐stayed bridge benchmark demonstrate the effectiveness of the proposed fault‐tolerant system. Despite the high order of this large‐scale structure, the proposed fault detection and diagnosis method can effectively find the location and size of faults in actuators without performance degradation and computational costs. The fault‐tolerant controller maintains the performance of the structure at an acceptable level in the post‐fault case by redistribution of control signal to actuators.