Multi-parameters decision-making algorithms for project level bridge maintenance

Decision-Making (DM) techniques have recently acquired considerable attention, as an essential process to determine the optimal treatment measures for bridge maintenance. In this paper, multi-parameters decision-making (DM) algorithms for project level bridge maintenance conducted by the Markov and the radial basis function (RBF) method and the Principal Component Analysis (PCA) method are investigated. A detailed study on the data classification and prediction in term of technical condition is performed by using the Markov and RBF method for the multi-parameter data, and characteristic of time series. The sub-item weights among different parameters are calculated by the PCA method, and evaluated by residual error analysis. Furthermore, an comprehensive optimal model of the maintenance DM is suggested, in which the comprehensive state index model, time-dependent target-based reliability model, degradation rate factor, and life cycle cost estimate, are built up as the evaluation criteria. The optimal process is selected and applied in an actual bridge as a case study. Results demonstrated that the comprehensive optimization model is preferable for the optimal maintenance times in bridge maintenance decision-making system (BMDMS).