Predicting the collapsibility potential of unsaturated soils using adaptive neural fuzzy inference system and particle swarm optimization

Soil collapsibility is one of the important phenomena in unsaturated soil mechanics. this phenomenon can impose extensive financial damages on civil engineering structures due to soil subsidence. because of uncertainties in effective parameters and their measurements, no precise mathematical relation has been proposed for collapsibility potential evaluation. therefore, soft computing techniques, such as fuzzy logic, could be a suitable choice to account for different factors. adaptive neural fuzzy inference system (anfis) was used in this study. to predict the collapsibility potential, hybrid algorithm and particles swarm optimization (pso) were employed by anfis for system training. gaussian membership functions were utilized for fuzzifying the data. in addition, data classication was performed in a subtractive form in the fuzzy inference system. a total of 327 laboratory data were used in the particle swarm algorithm, 266 of which were chosen for training and 66 for testing. the obtained results showed the effects of different parameters and the rate of their changes in collapsibility potential. moreover, the comparison of different approaches of system training was done using a correlation coefficient. the superiority of the proposed method and the utilized techniques was shown by comparing the results with the ones obtained by other researches.