Research on multisource information fusion of tunnel geological prediction based on evidence theory

Geological comprehensive prediction of tunnel is a very complicated process and the geological predicting workers have to face large amounts of geological materials and field data, which should be organized and analyzed effectively to obtain correct predicting results. Recent years have seen some comprehensive predicting models put forward (Ge et al. 2010, Xu et al. 2013, Xu et al. 2011, Yuan et al. 2011), which can integrate multiaspect information and have achieved good results in some engineering cases. There are still some shortcomings and limitations in these studies; for instance, the fuzzy evaluation method in the existing comprehensive predicting model mainly reflects the consistency but no clear description of conflict between each evaluation index. Meanwhile, as a complicated engineering system, the geological prediction of tunnel produces several uncertainties, but the current fuzzy evaluation methods are difficult to fully reflect the uncertain characteristics of the information.