Keywords: highway bridge, performance, prediction, fuzzy regression model.

Due to the crucial importance of bridges in ensuring the proper service of highway networks, maintenance of highway bridges plays a principal role in this effort. Generally, current bridge management systems evaluate bridge safety and reliability using inspection consequences that disregard overall system effects such as component reliability importance. Moreover, the performance of each component obtained through visual inspection usually is of an inconsistent measure of safety because the existence of defects may not be visible so that it does not truthfully reflect the importance of degradation on safety [1]. The desired performance of the bridge is the principal issue of bridge safety, however many of the bridges have been discovered that they have partially or seriously deteriorated or were destroyed before reaching the end of their service lives. Therefore assessment and prediction of bridge performance has become a primarily important task for the bridge maintenance manager.

The repercussions of deterioration or damage of the bridge during its service life present their own unique concerns and safety challenges. It is first necessary to carefully investigate all the possible causes contributing to bridge deterioration and to accurately examine the interrelationships among these possible causes. Deck slab, pavement, girder, pier, cap beam, abutment and foundation are the major components of a concrete bridge. Components of a bridge are subject to similar environmental conditions. Consequently, their deterioration is usually correlated. This dependency affects the overall performance of a highway bridge and thus it needs to be considered.

Potential factors could cause damage, deterioration or failure of the bridge during its lifecycle include the type of bridge, traffic loads, the age of bridge, design strength, improper construction, improper maintenance, adverse environment (weather, wind, earthquake, flood, etc), and human errors, etc. These variables are often difficult for a bridge engineer to measure because these uncertainty factors are often associated with vague data or information. Many of them can not be quantified objectively, but rather, they can be estimated subjectively and imprecisely through the engineer's experience and knowledge. To cope with vague or fuzzy data, fuzzy set theory is a commonly used technique.

Conventional regression analysis is one of the most commonly used statistical tools by engineers to build a prediction equation from collected data for the entire population. Essentially, ordinary regression analysis is used to deal with crisp observed data rather than fuzzy observed data. Regression analysis based on fuzzy data in coping with fuzziness is called fuzzy regression analysis. A major difference between fuzzy regression and conventional regression is that the deviations between the observed values and the estimated values are assumed to depend on the vagueness of the parameters in fuzzy regression models rather than on its measurement errors or randomness in ordinary regression techniques [2,3]. Since observations and variables involved in the cause-and-effect relation of the bridge performance are often vague and imprecise such as "the condition of the deck slab is bad" or "the crack numbers of the deck slab are around three", etc; thus an ordinary regression model is incapable of capturing this information and analyzing such problems.

This paper presents a fuzzy regression model with the aim of assessing and forecasting the performance of a bridge slab. A case study based on inspection data of highway bridges in Taiwan is performed to examine the variables contributing to the performance of the bridge slab using this model. The results demonstrate the effectiveness and capability of this model, which can be used to assist bridge managers to better predict bridge performance.

1

D.M. Frangopol, J.S. Kong, and E.S. Gharaibeh, "Reliability-based life-cycle management of highway bridge", Journal of Computing in Civil Engineering, ASCE, 15(1), pp. 147-57, 2001. doi:10.1061/(ASCE)0887-3801(2001)15:1(27)

2

N.F. Pan, F.C. Hadipriono and J.W. Duane, "Assessment of Causes of Building Collapse using Fuzzy Regression Analysis", in Proceedings of the Fourth International Conference on Engineering Computational Technology, B.H.V. Topping and C.A. Mota Soares, (Editors), Civil-Comp Press, Stirling, United Kingdom, paper 38, 2004. doi:10.4203/ccp.80.38

3

H. Tanaka, S. Uejima, and K. Asai, "Linear regression analysis with fuzzy models", IEEE Systems Trans. Systems Man Cybernet, SMC-2(6), pp. 903-07, 1982.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)