Improving the Reliability of a Bridge Deterioration Model for a Decision Support System Using Artificial Intelligence

Abstract

Bridge Management Systems (BMSs) have been developed since the early 1990s as a decision support system (DSS) for effective Maintenance, Repair and Rehabilitation (MR&R) activities in a large bridge network. Historical condition ratings obtained from biennial bridge inspections are major resources for predicting future bridge deteriorations through BMSs. However, available historical condition ratings are very limited in all bridge agencies. This constitutes the major barrier for achieving reliable future structural performances. To alleviate this problem, a Backward Prediction Model (BPM) technique has been developed to help generate missing historical condition ratings. Its reliability has been verified using existing condition ratings obtained from the Maryland Department of Transportation, USA. This is achieved through establishing the correlation between known condition ratings and related non-bridge factors such as climate and environmental conditions, traffic volumes and population growth. Such correlations can then be used to determine the bridge condition ratings of the missing years. With the help of these generated datasets, the currently available bridge deterioration model can be utilised to more reliably forecast future bridge conditions. In this chapter, the prediction accuracy based on 4 and 9 BPM-generated historical condition ratings as input data are also compared, using traditional bridge deterioration modeling techniques, i.e. deterministic and stochastic methods. The comparison outcomes indicate that the prediction error decreases as more historical condition ratings are available. This implies that the BPM can be utilised to generate unavailable historical data, which is crucial for bridge deterioration models to achieve more accurate prediction results. Nevertheless, there are considerable limitations in the existing bridge deterioration models. Thus, further research is essential to improve the prediction accuracy of bridge deterioration models.