Prediction of subsidence risk by FMEA using artificial neural network and fuzzy inference system

Construction of metro tunnels in dense and crowded urban areas is faced with many risks, such as subsidence. The purpose of this paper was the prediction of subsidence risk by failure mode and effect analysis (FMEA) and fuzzy inference system (FIS). Fuzzy theory will be able to model uncertainties. Fuzzy FMEA provides a tool that can work in a better way with vague concepts and without sufficient information than conventional FMEA. In this paper, S and D are obtained from fuzzy rules and O is obtained from artificial neural network (ANN). FMEA is performed by developing a fuzzy risk priority number (FRPN). The FRPN for two stations in Tehran No.4 subway line is 3.1 and 5.5, respectively. To investigate the suitability of this approach, the predictions by FMEA have been compared with actual data. The results show that this method can be useful in the prediction of subsidence risk in urban tunnels.