Adaptive algorithm for estimating excavation-Induced displacements using field performance data

• Main Authors: Haijian Fan, Gangqiang Kong
• Format: Article
• Language: English
• Published: Elsevier 2020-06-01
• Series: Underground Space
• Online Access: http://www.sciencedirect.com/science/article/pii/S246796741830134X

Empirical models provide a practical way to estimate the displacements induced by excavations. However, there are uncertainties associated with the predictions of empirical models owing to: (a) the imperfect knowledge of the model and (b) the uncertainties of the input variables. The uncertainties of these models can be characterized by a bias factor which is defined as the ratio of the actual displacement to the predicted displacement. The bias factors associated with the C&O method and the KJHH model are evaluated using the Bayesian method and a database of 71 excavations in Shanghai. To improve the predictions of the maximum displacement, an adaptive algorithm is proposed using field performance data. The performance of the proposed algorithm is demonstrated by an example in which excavation-induced displacements are generated by finite element method in normally consolidated clays. The example shows that the developed algorithm can significantly improve the predictions by incorporating the field performance data. Keywords: Excavation, Displacement prediction, Bayesian updating, Model bias