Fuzzy cognitive maps enabled root cause analysis in complex projects

This paper presents a Fuzzy Cognitive Maps (FCM) enabled Root Cause Analysis (RCA) approach to assessing the TBM performance in tunnel construction. Fuzzy logic is used to capture and utilize construction experience and knowledge from domain experts, and a cause-effect model consisting of nine concepts is established for simulating the TBM performance within the FCM framework. A tunnel case in the Wuhan metro system in China is used to demonstrate the applicability of the developed approach. Results indicate that (i) C-4 (Soil Density) displays a strongest negative correlation with the concept C-T (TBM Advance Rate); while C-8 (Grouting Speed) displays a strongest positive correlation with CT; (ii) TBM performance is very sensitive to the change of operational conditions, where the values of operational parameters can be adjusted to go up (or down) in case the TBM performance negatively (or positively) reduces; and (iii) we can identify the magnitude of the adjustment scope of operational variables when the TBM operational performance suffers a reduction. The novelty of the proposed approach is that it is verified to be capable of modeling dynamics of system behaviors over time and performing many kinds of what-if scenario analysis, including predictive, diagnostic, and hybrid RCA, which turns out to be a more competitive solution that deals with uncertainty, dynamics, and interactions in the approximate reasoning process, compared to other traditional approximate methods (i.e. Fault Tree Analysis (FTA), Rule-Based Reasoning (RBR), and Case-Based Reasoning (CBR)). The proposed approach can be used as a decision support tool for ensuring the satisfactory performance of TBMs, and thus, increases the efficiency of tunnel construction projects. (C) 2017 Elsevier B.V. All rights reserved.