Method for predicting blasting lumpiness of rock mass

Abstract

The invention discloses a method for predicting the blasting lumpiness of a rock mass, and relates to the field of prediction of the blasting lumpiness of the rock mass. The method comprises the following steps of: carrying out three-dimensional scanning on joint cracks of the rock mass to be blasted to obtain the distribution condition of the three-dimensional joint cracks of the rock mass; determining blasting design parameters; sampling the rock mass to be blasted to obtain rock mass strength parameters; determining the lumpiness size of the rock mass by applying an image recognition technology to a blasted blast heap; taking data obtained in the previous three steps as input parameters by applying a BP neural network, and correcting each connection weight and threshold value by continuously training the BP neural network to obtain an optimal rock mass blasting lumpiness prediction model; and inputting a test set into the optimal rock mass blasting lumpiness prediction model to obtain the blasting lumpiness size. According to the method, factors such as the distribution condition of the internal joint Jc and the surface joint Js of the rock mass, blasting design parameters and rock mass strength are comprehensively considered, the lumpiness size of the rock mass after blasting can be accurately predicted, and the prediction precision is high.

Description

technical field [0001] The invention relates to the technical field of predicting the fragmentation degree of rock mass blasting, in particular to a method for predicting the fragmentation degree of rock mass blasting. Background technique [0002] Existing techniques for predicting rock mass blasting fragmentation are based on PCA_BP neural network technology, GA_BP neural network technology, RBF neural network technology, Bezmatner model, Kuz-Ram model, etc. In the application of PCA_BP, GA_BP, RBF three neural network prediction technologies for blasting fragmentation, the influence of parameters such as blasting design on blasting fragmentation is considered, while the influence of parameters such as rock mass joints and strength is not considered. [0003] The Bezmatelne model is based on the probability distribution of rock mass joints and fissures being cut into natural rock blocks before blasting and the probability distribution of rock mass forming fragments after b...

AI Technical Summary

Problems solved by technology

This model establishes the relationship between blasting parameters and blasting fragmentation, but does not conduct in-depth and detailed research on the mechanism of the influence of rock mass joints and fissures on the blasting fragmentation distribution, and only uses the coefficient A 0 Reflecting the mechanical properties of rock and the development of joints and fissures, the factors considered are not comprehensive

Images