Determination of Powder Factor for an Optimized Blast Pattern Design Using Artificial Neural Network (Case Study: Gol-e-Gohar Iron Ore Mine)

Side-effects of bench blasting are including ground vibration, back break, fly rock, and having boulder, in which back break and the degree of rock fragmentation (having boulder) have the most effects on the produc tion process costs. Expecting less back break and also a good degree of fragmentation can be used as two criteria to design an optimized blast pattern with an optimum powder factor used. In this paper, artificial neural network capable of modeling the behavior of complex nonlinear processes, adopting Radial Basis Functions (RBF) network was used to carry out the modeling and optimization. Two RBF networks, one accounting for back break prediction and the other considering for the prediction of the degree of rock fragmentation were designed. The networks designed were validated against the data obtained through the field tests, carried out at Gol-e-Gohar iron ore mine of Sirjan-Iran. Different blast patterns for different geological conditions were designed and the blast pattern and relevant powder factor associated with less back break and a good degree of fragmentation was selected.

Keywords:

Powder factor , Artificial neural network , RBF , Bench blasting , Back break , Fragmentation