Belief function theory based decision support methods: application to torrent protection work effectiveness and reliability assessment

Civil engineering protection works mitigate natural risks in mountains, such as torrents. Analysing their effectiveness at several scales is an essential issue in the risk management. Based on expert knowledge, used methods have been developed under risky environment. However, decision is made under uncertainty because of 1) the lack of information and knowledge on natural phenomena and 2) the heterogeneity of available information and 3) the reliability of sources. In this paper, we propose to help decision-makers with advanced multicriteria decision making methods (MCDMs). Combining classical MCDM approaches, belief function, fuzzy sets and possibility theories, they make it possible decisions based on heterogeneous, imprecise and uncertain evaluation of criteria provided by more or less reliable sources in an uncertain context. COWA-ER (Cautious Ordered Weighted Averaging with Evidential Reasoning), Fuzzy-Cautious OWA or ER-MCDA (Evidential Reasoning for Multi Criteria Decision Analysis) are thus applied to several scales of effectiveness assessment.