Optimization of blasting cost in lime stone mines using PSO metaheuristic algorithm

The prediction and optimization of blasting cost to achieve optimal fragmentation is significant, considering the control of the adverse consequences of the blast. In this study, by collecting blasting data in six limestone mines in Iran, a model was developed to predict the cost of blast using nonlinear multivariate regression. Compared to linear regression model, this model has a higher correlation coefficient (0.913) and root mean square error (1089) and in comparison, with the linear model, the nonlinear model shows a better match with the actual cost of the blast. Based on sensitivity analysis, spacing and number of holes had the highest and the least effect on the cost model of the blast, respectively. In addition, in this study along with achieving the blast cost function, the restrictive functions of the blast include fragmentation, fly rock and back break was modeled using nonlinear multivariate regression method, and these functions as inputs in the metaheuristic algorithm of Particles Swarm Optimization (PSO) were used to optimize the cost of blast. Using this method of spacing, the number and length of holes as design parameters of blast are 3.6 meters, 462 loops, and 13 meters respectively and fragmentation fly rock and back break as blasting limitations are 44 cm, 84.5 meters and 3.6 meters and blasting cost was 6235 Rials per ton, which results in a 12.9 percent reduction in the cost of blast and optimal control of the adverse consequences of the blast.