No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Application of a model- free ANN approach for SHM of the Old Lidingö Bridge
Abstract and Figures
This paper explores the decision making problem in SHM regarding the maintenance of civil
engineering structures. The aim is to assess the present condition of a bridge based exclusively on
measurements using the suggested method in this paper, such that action is taken coherently with
the information made available by the monitoring system.
Artificial Neural Networks are trained and their ability to predict structural behaviour is evaluated
in the light of a case study where acceleration measurements are acquired from a bridge located
in Stockholm, Sweden. This relatively old bridge is presently still in operation despite experiencing
obvious problems already reported in previous inspections. The prediction errors provide a
measure of the accuracy of the algorithm and are subjected to further investigation, which
comprises concepts like clustering analysis and statistical hypothesis testing. These enable to
interpret the obtained prediction errors, draw conclusions about the state of the structure and
thus support decision making regarding its maintenance.
Figures - uploaded by Cláudia Neves
Author content
All figure content in this area was uploaded by Cláudia Neves
Content may be subject to copyright.
ResearchGate has not been able to resolve any citations for this publication.
This is probably the most appropriate time for the development of robust and reliable structural damage detection systems as aging civil engineering structures, such as bridges, are being used past their life expectancy and well beyond their original design loads. At the same time, there is an increasing monetary pressure on bridge authorities to extend the life span of the structures whilst ensuring public safety. Often, when a significant damage to the structure is discovered, the deterioration has already progressed far and required repair is substantial and costly. Not seldom will the structure be demolished and a new one constructed in its place. This is both expensive and has negative impact on the environment and traffic during replacement. For the exposed reasons the demand for efficient Structural Health Monitoring techniques is currently extremely high.
This licentiate thesis presents a two stage, model-free damage detection approach based on Machine Learning. The method is applied to data on the structural condition of a railway bridge gathered in a numerical experiment using a three dimensional finite element model of a single track bridge. The initial step in this study consists in collecting the dynamic response of the structure that is simulated during the passage of a train, considering the bridge in healthy state and in two different damaged states. The first stage of the proposed algorithm consists in the design and unsupervised training of Artificial Neural Networks that, provided with input data composed of measured accelerations in previous instants, are capable of predicting future output acceleration. In the second stage and following the selection of the best performing trained networks, the acceleration prediction errors are used to fit a Gaussian Process that enables to perform a statistical analysis of the distribution of errors. Subsequently, the concept of Damage Index is introduced and the probabilities associated with the errors of issuing a false alarm and missing existing damage are studied. Following the former steps it is possible to generate a Receiver Operating Characteristic curve, based on which the threshold of the detection system can be adjusted in order to achieve optimal performance as a function of the trade off between errors. Lastly, supported by the Receiver Operating Characteristic curve and the Bayes’ Theorem, a simplified method for the calculation of the expected cost of the strategy is proposed and exemplified with two cost scenarios.
Data mining is a process which finds useful patterns from large amount of data. The paper discusses few of the data mining techniques, algorithms and some of the organizations which have adapted data mining technology to improve their businesses and found excellent results.
Understanding the rheological properties of fresh concrete, the hydration phenomenon of cement responsible for structuration, the relationship between the characteristics of the porous solid obtained and its mechanical performances or resistance to the aggressive penetration requires a complex knowledge of the physicochemistry of reactive porous materials. The development of simple formulation rules therefore requires the assimilation of this knowledge and a good command of the properties of these materials. The purpose of this book is to provide the mix designer with useful knowledge on granular materials and porous materials, which will enable the innovative design of concrete. Topics covered include the characterization of granular materials, the concepts of porosity and specific surface area, and the transport properties (diffusion and permeation) of concrete. Some of these topics are already covered in other general books dedicated to granular or porous materials. The objective here is to bring them together in one book by adapting them for use by concrete specialists. Applications in the form of exercises are offered at the end of each chapter to enable readers to assimilate the theoretical knowledge and to apply such knowledge to concrete problems encountered in civil engineering.
Novelty detection is the task of classifying test data that differ in some respect from the data that are available during training. This may be seen as “one-class classification”, in which a model is constructed to describe “normal” training data. The novelty detection approach is typically used when the quantity of available “abnormal” data is insufficient to construct explicit models for non-normal classes. Application includes inference in datasets from critical systems, where the quantity of available normal data is very large, such that “normality” may be accurately modelled. In this review we aim to provide an updated and structured investigation of novelty detection research papers that have appeared in the machine learning literature during the last decade.
Sammanfattning av 2015 års inspektioner - Gamla Lidingöbron
- A Ahne
A. Ahne, "Sammanfattning av 2015 års
inspektioner -Gamla Lidingöbron. Rapport 2015-12-03" [Summary of the 2015 inspections -Old
Lidingö Bridge. Report 2015-12-03], Grontmij
Anläggningsunderhåll Stockholm, 2015.
Gamla Lidingöbron: Fördjupad analys av mätta accelerationer på fundament 6 och 8
- R Karoumi
R.
Karoumi,
"Gamla
Lidingöbron:
Fördjupad analys av mätta accelerationer på
fundament 6 och 8" [Old Lidingö Bridge:
8], KTH
Byggvetenskap, Stockholm, 2012.
Neural Networks in C++
- A Blum
A. Blum, "Neural Networks in C++", N.Y.:
Willey, 1992.
Data Mining Techniques
- M Berry
- G Linoff
Gamla Lidingöbron. Töjningsmätningar - Preliminär rapport
- J Leander
Gamla Lidingöbron. Rapport 2006-01-31
- T Lundmark