PREDICTION OF FAILURE PROBABILITY FOR SOILSTRUCTURE INTERACTION SYSTEM USING MODIFIED ANFIS BY HYBRID OF FCM-FPSO

In this study, an efficient method is introduced to predict the stability of soil-structureinteraction (SSI) system subject to earthquake loads. In the procedure of the nonlineardynamic analysis, a number of structures collapse and then lose their stability. Theprediction of failure probability is considered as stability criterion. In order to achieve thispurpose, a modified adaptive neuro fuzzy inference system (ANFIS) is proposed by a hybrid of fuzzy c-means (FCM) and fuzzy particle swarm optimization (FPSO). To train themodified ANFIS, the input–output data are classified by a hybrid algorithm consisting ofFCM-FPSO clustering. The optimum number of ANFIS fuzzy rules is determined bysubtractive algorithm (SA). Results of illustrative examples demonstrate high performanceof the modified ANFIS in comparison with the single ANFIS.