Estimation of shear strength parameters of soil using Optimized Inference Intelligence System

In recent years, machine learning techniques have been developed and used to build intelligent information systems for solving problems in various fields. In this study, we have used Optimized Inference Intelligence System namely ANFIS-PSO which is a combination of Adaptive Neural-Fuzzy Inference System (ANFIS) and Particle Swarm Optimization (PSO) for the estimation of shear strength parameters of the soils (Cohesion “C” and angle of internal friction “φ”). These parameters are required for designing the foundation of civil engineering structures. Normally, shear parameters of soil are determined either in the field or in the laboratory which require time, expertise and equipments. Therefore, in this study, we have applied a hybrid model ANFIS-PSO for quick and cost-effective estimation of shear parameters of soil based on the other six physical parameters namely clay content, natural water content, specific gravity, void ratio, liquid limit and plastic limit. In the model study, we have used data of 1252 soft soil samples collected from the different highway project sites of Vietnam. The data was randomly divided into 70:30 ratios for the model training and testing, respectively. Standard statistical measures: Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Correlation Coefficient (R) were used for the performance evaluation of the model. Results of the model study indicated that performance of the ANFIS-PSO model is very good in predicting shear parameters of the soil: cohesion (RMSE = 0.075, MAE = 0.041, and R = 0.831) and angle of internal friction (RMSE = 0.08, MAE = 0.058, and R = 0.952).