Prediction of blasting mean fragment size using support vector regression combined with five optimization algorithms

The main purpose of blasting operation is to produce desired and optimum mean size rock fragments. Smaller or fine fragments cause the loss ore during loading transportation, whereas large coarser need be further processed, which enhances production cost. Therefore, accurate prediction fragmentation crucial in operations. Mean fragment (MFS) a index that measures goodness designs. Over past decades, various models have been proposed evaluate predict fragmentation. Among these models, artificial intelligence (AI)-based are becoming more popular due their outstanding results for multi-influential factors. In this study, support vector regression (SVR) techniques adopted as basic tools, five types optimization algorithms, i.e. grid search (GS), grey wolf (GWO), particle swarm (PSO), genetic algorithm (GA) salp (SSA), implemented improve performance optimize hyper-parameters. model involves 19 influential factors constitute comprehensive MFS evaluation system based on AI techniques. all GWO-v-SVR-based shows best predicting operation. Three mathematical indices, square error (MSE), coefficient determination (R2) variance accounted (VAF), utilized evaluating different models. R2, MSE VAF values training set 0.8355, 0.00138 80.98, respectively, 0.8353, 0.00348 82.41, respectively testing set. Finally, sensitivity analysis performed understand influence input parameters MFS. It most sensitive factor uniaxial compressive strength.