Publications

Pro096

STOCHASTIC, DETERMINISTIC, STATISTICAL AND ARTIFICIAL INTELLIGENCE BASED MODELS TO PREDICT THE SERVICE LIFE OF RENDERED FACADES



Author(s): Silva, A., Gaspar, P. L. and Brito, J.
Paper category: Conference
Book title: XIII International Conference on Durability of Building Materials and Components - XIII DBMC
Editor(s): Marco Quattrone, Vanderley M. John
Print ISBN: none
e-ISBN: 978-2-35158-149-0
Publication year: 2015
Pages: 351-358
Total Pages: 8
Language: English


Abstract: Service life prediction has a primary role in today‟ context, allowing a more rational use of construction elements, reducing the costs associated with rehabilitation procedures. In the literature, the most common methods used to estimate the service life of buildings and its components can be classified as deterministic, probabilistic and engineering (a symbiosis of the previous two). In this study, the application of deterministic (graphical method), stochastic (logistic regression and Markov chains), statistic (multiple linear and non-linear regression techniques) and artificial intelligence based models (artificial neural networks) is proposed to predict the service life of rendered facades. Rendered facades are one of the most common
types of claddings applied in Portugal. However, the predominance of this type of coating is related with the low investment applied in this solution, which often implies an unacceptable degradation of the building heritage. A comparative study is performed to analyse the applicability of each method proposed. This analysis relates the ease of application with the effectiveness of the proposed models. The results obtained through the different methods applied in this study are coherent from an empirical point of view. Furthermore, the results obtained are consistent with other studies performed in this field of knowledge.


Online publication: 2015
Publication Type: full_text
Public price (Euros): 0.00


>> You must be connected to view the paper. You can register for free if you are not a member