Prediction of Rock Tensile Strength Using Soft Computing and Statistical Methods

The tensile strength of the rocks is one effective factors in rupture structure foundations and underground spaces, stability rocky slopes, ability to drill explode rocks. This research was conducted estimate using methods such as simple regression (SR), multivariate linear (MVLR), support vector (SVR) with radial basis kernel function, multilayer feed-forward artificial neural network (MFF-ANN), Gaussian process (GPR) squared exponential (SEK) adaptive neuro-fuzzy inference system (ANFIS) based on membership function. For this purpose, petrography, engineering features limestone, sandstone, argillaceous limestone samples south Iran, were assessed. results obtained from study compared those previous research, revealing a strong correlation (R2=0.95 1.00) between our findings published works. To Brazilian (BTS), index properties including water absorption by weight, point load (PLI), porosity%, P-wave velocity (Vp), density considered inputs. Methods various criteria. SVR precision (R=0.96) higher than MFF-ANN (R=0.92), ANFIS (R=0.95), GPR (R=0.945), MVLR (R=0.89) strength. average BTS measured laboratory predicted all 5 6.62 6.71 MPa, respectively, which shows very high investigated methods. Analysis model criteria statistical analysis for developed relationships revealed that there sufficient accuracy use empirical equations.