Strategic framework for ANFIS and BIM use on risk management at natural gas pipeline project

Abstract

Risk management is a multi-criteria decision making problem that includes various factors according to literature research and expert opinions. In order to solve the relationship between these criteria and to establish effective models, researchers have presented a wide variety of methods or models in the literature. In order to manage risk management effectively and to minimize its impact on project parameters, the most appropriate methods and criteria and strategic selection are required. In this article, a case study conducted using the multi-criteria Adaptive Neuro-Fuzzy Inference System (ANFIS), which enables the grading of 40 real risk types covering the design and construction processes of projects, is examined. Combining the structures and advantages of adaptive networks with fuzzy inference methodology has demonstrated a more comprehensive and effective risk management and assessment. Root mean square error (RMSE), mean absolute percentage error (MAPE) and R2 performance indicators have shown that the artificial intelligence supported risk management approach created with the Adaptive Neuro Fuzzy Inference System gives better results. The main contribution of this study is the approach of artificial intelligence to accurately assess and grade risks with the hybrid learning method, and then assign linguistic expressions and warning texts on risk items using 3D BIM and FLS (fuzzy linguistic summarization) systems.