The aim of model-based structural identification is to identify suitable models and values for model parameters that determine structure behavior through comparing measurements with predictions. Well-known methodologies, such as traditional implementations of Bayesian model updating, have been shown to be inaccurate in cases characterized by systematic uncertainties and unknown spatial correlations. Error-domain model falsification (EDMF) is another approach to structural identification. This approach is easy to understand for practicing engineers and can provide robust parameter identification without assumptions on spatial correlations. The performance of all approaches involving sampling is affected by the number of model evaluations that is generated based on prior knowledge of parameter-value distributions. This paper focuses on a new sampling technique, called radial-basis function sampling (RBFS), and its application to EDMF, to generate a set of candidate models that represents the behavior of the structure with a certain confidence level. Radial-basis function sampling provides a good exploration of the parameter space even with a limited number of samples, which results in reduced computation times. A full-scale bridge in Singapore has been tested and a new index of sampling quality is proposed to compare this approach with other sampling techniques such as Latin hypercube sampling (LHS) and Markov-chain Monte Carlo (MCMC). Finally, a cross-validation method is used to verify the robustness of the approach and the sensitivity of sampling on prediction reliability.
The aim of structural identification is to provide accurate knowledge of the behaviour of existing structures. In most situations, finite-element models are updated using behaviour measurements and field observations. Error-domain model falsification (EDMF) is a multi-model approach that compares finite-element model predictions with sensor measurements while taking into account epistemic and stochastic uncertainties—including the systematic bias that is inherent in the assumptions behind structural models. Compared with alternative model-updating strategies such as residual minimization and traditional Bayesian methodologies, EDMF is easy-to-use for practising engineers and does not require precise knowledge of values for uncertainty correlations. However, wrong parameter identification and flawed extrapolation may result when undetected outliers occur in the dataset. Moreover, when datasets consist of a limited number of static measurements rather than continuous monitoring data, the existing signal-processing and statistics-based algorithms provide little support for outlier detection. This paper introduces a new model-population methodology for outlier detection that is based on the expected performance of the as-designed sensor network. Thus, suspicious measurements are identified even when few measurements, collected with a range of sensors, are available. The structural identification of a full-scale bridge in Exeter (UK) is used to demonstrate the applicability of the proposed methodology and to compare its performance with existing algorithms. The results show that outliers, capable of compromising EDMF accuracy, are detected. Moreover, a metric that separates the impact of powerful sensors from the effects of measurement outliers have been included in the framework. Finally, the impact of outlier occurrence on parameter identification and model extrapolation (for example, reserve capacity assessment) is evaluated.
1 November 2018 in Journal of Computing in Civil Engineering
In responsive cities, user feedback and information provided by sensors are combined to improve urban design and to support asset managers in performing decision making. Optimal management of infrastructure networks requires accurate knowledge of current asset conditions to avoid unnecessary replacement and expensive interventions when cheaper and more sustainable alternatives are available. Structural model updating is a discipline that focuses on improving behavior-model accuracy by means of measurements taken from the built environment. Error-domain model falsification (EDMF) is a simple and practice-oriented methodology that uses measurements at sensor locations to identify plausible models among an initial population generated according to engineering judgment. However, many plausible models are often identified, making result interpretations difficult for practicing engineers. In this paper, a clustering methodology based on bipartite-modularity optimization (BMO) is used to clarify identification outputs. Compared with classical clustering methods such as K-means, BMO clustering provides more accurate interpretations and better visualization of the results. Moreover, engineers can actively interact with the clustering framework to obtain the knowledge that is needed at several stages of the decision-making process.
The growth of cities means that demand for fast, reliable and safe mobility in urban environments is increasing. Transportation networks provide an essential contribution. The core of these networks is made up of infrastructure such as roads and bridges that often, have not been designed to meet current needs. Optimal management requires an accurate knowledge of how existing structures behave. This helps avoid unnecessary replacement and expensive interventions when cheaper and more sustainable alternatives are available. The aim of structural model updating is to provide accurate models of real behavior using measurements and more qualitative observations. Error domain model falsification has been proposed as a robust population methodology to identify sets of plausible models by comparing their predictions with measurements at sensor locations. However, many plausible models are often identified, making result interpretations difficult for practicing engineers. In this paper, a methodology based on a clustering measure called bipartite modularity is employed to improve understanding. Results show that bipartite modularity, when compared with classical clustering methods such as K-means, provides more accurate interpretations and better visualization of the results.