Real-Time Damage Detection in Building Structures Using Model-Based Adaptive Control Approach

Abstract

In the last few years, researchers have paid close attention to developing smart structures that simultaneously combine structural health monitoring and vibration control. In this regard, real-time system identification is one of the major challenges due to various time delay factors in smart structures. This study presents a new model-based online damage detection method using an adaptive control approach (MODAC) under indeterminate conditions (limitations on the number of sensors). Here, it is assumed that the input and output data of the system during the vibration under environmental excitations can be measured, and, in addition, the information about the mechanical parameters of the intact system is available. In the proposed MODAC method, unlike the common methods, which mainly use optimization algorithms and are time consuming, the variations in the structural parameters of the system are mathematically formulated and directly calculated in real time. In addition, this paper introduces a new scheme that combines the MODAC method with the semiactive control strategy to create a smart structure that compensates for possible damages to the structure. The MODAC method’s accuracy, as well as the presented scheme’s performance and efficiency, is assessed through numerical examples when various excitation records are considered. The obtained results indicate that damage can be identified instantaneously with acceptable accuracy even under indeterminate conditions and in the presence of noise. Furthermore, the proposed scheme can effectively decrease dynamic responses and compensate for the permanent damage in the smart structure during strong ground motions. In this study a new online damage detection method using a model-based adaptive control approach is proposed. It is assumed that the dynamic responses and input excitation to the system are measured and available through installed sensors. A dynamic mathematical model is updated to trace behavior of the target system in order to detect any changes in the stiffness properties of the structure. Several factors lead to a decrease in stiffness (such as a structural element failure, cracks in welded joints and concrete members, reduction in cross-section area, corrosion, reduction in modulus of elasticity, etc.). The proposed method is capable of simultaneously evaluating structures online or offline to assist in decision-making in retrofitting or repair processes. Furthermore, a semiactive control device is applied to reduce the effects of identified damages. The obtained results reveal that the proposed method can effectively detect and compensate damages caused by the stiffness reduction in the structure. Since continuous monitoring and online damage detection is very important in major infrastructure projects, the proposed method can be effectively used by the construction companies that are active in the field of both structural health monitoring and vibration control.