Keywords: bridge performance, concrete compression strength, prediction, time series analysis, autoregressive integrated moving average.

The concrete compression strength of bridges under construction is one of the most essential characteristics of engineering quality. Accurately forecasting the future values of the compression strength and their trends are important information for the bridge contractors to make cost-effective decisions. To achieve this objective, one usually depends on historical data or collected data from the current condition. Such data are often obtained through a sequence of observations that are taken sequentially in time.

Essentially, forecasting concrete compression strength for particular bridge components can be decomposed into a time-series modelling problem in which one wishes to predict the values based on a series of past or current observations at the time intervals of interest. Accordingly, the use of a model that can effectively deal with time series data is essential. The regression analysis is widely used to find the best fitted mathematical model, so that a dependent variable can be predicted from independent variable(s) [1]. Nevertheless, a regression equation is developed based on the range of data; thus this approach is incapable of extrapolating well beyond the range of the data [2,3]. Besides, explanatory variables in the regression model are constant such that their interdependencies or effects do not change over time and cannot be identified. A Markovian chain model is a popular probabilistic approach for forecasting the performance of state-of-the-art infrastructure [4,5]. However, the major problem of this approach is the assumption that the probability of making a transition from one condition state to another is independent of the components [3].

A time series is a discrete set with an underlying sequential time order. Time series models provide a superior forecasting capability to discover useful decision-making information such as trends, seasonality, and random errors [6]. Autoregressive integrated moving average (ARIMA), a parametric modelling approach to time series, is a well suited for application to short-term condition forecasting [7]. Thus, this paper employs ARIMA to predict the concrete compression strength of an under-constructing bridge construction project in Taiwan. The results show the capability of the model, which can assist bridge managers to effectively forecast bridge performance.

1

L. Lapin, "Modern Engineering Statistics", Duxbury Press, New York, USA, 1997.

2

S.M. Madanant, W.H. Ibrahim Wan, "Poisson and Negative Binomial Regression Models for the Computation of Infrastructure Transition Probabilities", Journal of Transportation Engineering, ASCE, 121(3), 267-72, 1995. doi:10.1061/(ASCE)0733-947X(1995)121:3(267)

3

N.F. Pan, "Forecasting bridge deck conditions using fuzzy regression analysis", Journal of the Chinese Institute of Engineers, 30(4), 593-604, 2007.

4

Y. Jiang, M. Saito, K.C. Sinha, "Bridge Performance Prediction Model Using the Markov Chain", Transportation Research Record, 1180, 25-32, 1988.

5

W.T. Scherer, D.M. Glagola, "Markov Models for Bridge Maintenance Management", Journal of Transportation Engineering, ASCE, 120(1), 37-51, 1994. doi:10.1061/(ASCE)0733-947X(1994)120:1(37)

6

C.E.P. Box, G.M. Jenkins, G.C. Reinsel, "Time Series Analysis: Forecasting and Control", Prentice-Hall, 1994.

7

B.L. Smith, "Forecasting freeway traffic flow for intelligent transportation system applications", Doctoral dissertation, Department of Civil Engineering, University of Virginia, Charlottesville, USA, 1995.

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 £95 +P&P)