No abstract

Structural health monitoring (SHM) strategies should ideally consist of continuous on-line damage detection processes, which do not need to rely on the comparison of newly acquired data with baseline references, previously defined assuming that structural systems are undamaged and unchanged during a given period of time.The present paper addresses the topic of SHM and describes an original strategy for detecting damage in an early stage without relying on the definition of data references. This strategy resorts to the combination of two statistical learning methods. Neural networks were used to estimate the structural response, and clustering methods were adopted for automatically classifying the neural networks' estimation errors. To ensure an on-line continuous process, these methods were sequentially applied in a moving windows process.The proposed original strategy was tested and validated on numerical and experimental data obtained from a cable-stayed bridge. It proved highly robust to false detections and sensitive to early damage by detecting small stiffness reductions in single stay cables as well as the detachment of neoprene pads in anchoring devices, resorting only to a small amount of inexpensive sensors. detection approaches rely on signal processing and statistical learning techniques to extract sensitive information from time-series acquired on site [8][9][10][11]. Their computational simplicity makes them cost-effective and the most suitable candidates for carrying out automated on-line damage detection [2], either based on modal information [12][13][14] or based on statistical and time-series features [2,8,15].Data-driven SHM approaches rely on two mandatory steps for conducting damage detection: response modelling and statistical classification. The first aims at separating the variations imposed by 'normal' environmental/operational actions from those caused by damage [16]. It relies on training statistical learning algorithms so that they can accurately estimate the 'normal' structural response. Any 'abnormal' variations can afterwards be highlighted by comparing the estimates with the actual responses. The most reported statistical modelling algorithms found in SHM literature consist of multi-layer perceptron (MLP) neural networks [17][18][19], support vector regressions [20], linear regressions [2], principal component analysis [21] and auto associative neural networks [22]. Regardless of the chosen algorithms, response modelling has been reported in the literature as a supervised problem, where the statistical learning algorithms are trained a priori with reference data, in which the structural systems must be assumed undamaged and unchanged [9,11,23].Statistical classification consists of discriminating SHM data as related to identical or distinct structural conditions [24,25]. This step has also been addressed under supervised approaches, where classification algorithms are trained with reference data sets (in general, the same ones used for response modelling) to define boundaries that s...

This paper aims at detecting damage in railway bridges based on traffic-induced dynamic responses. To achieve this goal, an unsupervised automatic data-driven methodology is proposed, consisting of a combination of time series analysis methods and multivariate statistical techniques. Damage-sensitive features of train-induced responses are extracted and allow taking advantage, not only of the repeatability of the loading, but also, and more importantly, of its great magnitude, thus enhancing the sensitivity to small-magnitude structural changes.The efficiency of the proposed methodology is validated in a long-span steel-concrete composite bowstring-arch railway bridge with a permanent structural monitoring system installed. An experimentally validated finite element model was used, along with experimental values of temperature, noise, and train loadings and speeds, to realistically simulate baseline and damage scenarios.The proposed methodology proved to be highly sensitive in detecting early damage, even when it consists of small stiffness reductions that do not impair the safety or use of the structure, and highly robust to false detections. The analysis and validation allowed concluding that the ability to identify early damage, imperceptible in the original signals, while avoiding observable changes induced by variations in train speed or temperature, was achieved by carefully defining the modelling and fusion sequence of the information. A single-value damage indicator, proposed as a tool for real-time structural assessment of bridges without interfering with the normal service condition, proved capable of characterizing multi-sensor data while being sensitive to identify local changes.

This article addresses the subject of data-driven structural health monitoring and proposes a real-time strategy to conduct structural assessment without the need to define a baseline period, in which the monitored structure is assumed healthy and unchanged. Independence from baseline references is achieved using unsupervised discrimination machine-learning methods, widely known as clustering algorithms, which are able to find groups in data relying only on their intrinsic features and without requiring prior knowledge as input. Real-time capability is based on the definition of symbolic data, which allows describing large amounts of information without loss of generality or structural-related information. The efficiency of the proposed methodology is illustrated using an experimental case study in which structural changes were imposed to a suspended bridge during an extensive rehabilitation programme. A single-value novelty index capable of describing multisensor data is proposed, and its effectiveness in identifying structural changes in real time, using outlier analysis, is discussed.

No abstract

No abstract

No abstract

This paper presents a progressive numerical model validation of a bowstring-arch railway bridge based on the analysis of experimental data from different structural response measurements, namely, static deformations under environmental actions, modal vibrations, and transient dynamic responses under traffic loads. This work also addresses an integrated approach that uses structural health monitoring (SHM) measurements in combination with FE modelling to understand the structural behaviour of a long-span complex bridge. An in-situ, progressively-phased SHM system has provided a diverse set of data streams ranging from static and dynamic responses to the measurement of environmental and operational traffic loads. The first phase consists of defining a detailed baseline finite element (FE) model of the bridge, envisaging the initial condition of the structure immediately after construction, and its validation using modal parameters (natural frequencies and mode shapes) derived from an ambient vibration test. Since in-service deflections generally do not exercise the non-linear regime of the response of the bridge, the second phase focuses on the analysis of static response data and temperature measurements to validate the non-linear behaviour of the structural system, particularly at the bearing devices, under slow actions. The third and final phase addresses the dynamic analysis under traffic actions, which provides greater sensitivity in the detection of non-linear behaviour due to the effects of high amplitude actions induced by regular train loading profiles. In order to guarantee the accuracy of the baseline numerical model, particularly under temperature and traffic actions, it was necessary to use contact restrictions in some specific bearing devices. This improvement is in line with the structural changes detected in some bearing devices through visual inspections. As a result, an updated numerical model capable of reproducing the modal, static, and dynamic structural responses was achieved, showing a very good agreement between experimental and numerical data. In future applications, this updated numerical model will be useful for assessing the condition of the bridge under traffic loads, namely to identify damages and support the adoption of life cycle maintenance strategies based on the integration of SHM systems with (stochastic) load and failure models.