Two novel combined systems for predicting the peak shear strength using RBFNN and meta-heuristic computing paradigms

摘要: Effective prediction of the peak shear strength (PSS) is crucial importance in evaluating stability a rock slope with interlayered rocks and has both theoretical practical significance. This paper offers two novel tools for PSS based on radial basis function neural network (RBFNN) meta-heuristic computing paradigms. For this work, gray wolf optimization (GWO) ant colony (ACO) algorithms were used to select optimal parameters RBFNN. Then, these new models compared gene expression programming (GEP) model. A total 158 experimental data train test proposed using three input parameters, i.e., normal stress, compressive ratio joint walls, roughness coefficient. Finally, computational result revealed that RBFNN-GWO model, coefficient determination (R2) 0.997, produced better convergence speed higher accuracy RBFNN-ACO GEP models, R2 0.995 0.996, respectively. The model was found an efficient predictive tool can help engineers slopes design processes.