Abstract: Based on neural network technology, BP neural network discriminant model of rock blasting classification was designed by using MATLAB software.Rock blasting classification discriminant of 7 different kinds of rock in a iron mine were done by using this model.Comparative analysis was done between this discriminant results and discriminant results by fuzzy identification method, the conclusions were the same.Meanwhile, it shows BP neural network model designed by using MATLAB language has characteristics of simple and accurate, and it also has high feasibility and objectivity.

Key words: Rock mass, Neural network, Blastability classification