Study On Blasting Parameters Optimization Of Open-pit Mrne Based On BP Neural Network

In the production process of Heidaigou Open-pit Coal Mine,bench blasting is one of the important production links.The effect of blasting directly affects the normal operation of the whole production system.Soft rock is the main rock in this mine.Most bench blasting of rock uses the same blasting parameters,so the blasting effect is not ideal.According to this situation,the bench blasting parameters are rationally optimized,and different blasting areas are divided according to different geological structures and rock properties,and the most reasonable blasting parameters are put forward.Based on the blasting effect,in order to reduce the cost of blasting construction,increase the efficiency of blasting construction,reduce the mining cost,and rationally optimize the bench blasting parameters,this paper classifies the rock in the mining area according to the physical and mechanical properties of the rock in the mining area by using the general coefficient method and the wave impedance of rock mass,and studies the comprehensive evaluation method of fuzzy mathematics.The weight of bench blasting effect in open pit mine is analyzed,and the weight of bench blasting effect is determined.According to the factors affecting the bench blasting effect,the previous blasting parameters and rock mechanical properties of the mine were collected,and the collected data and information were taken as the initial data of the BP neural network blasting optimization model.The BP neural network prediction model between blasting parameters and blasting effect for MATLAB language is established,and the genetic algorithm is used to optimize the established neural network to determine the initial weights and thresholds between network nodes.According to the measured mechanical properties,the input parameters of the model are ultra-deep(m),explosive unit consumption(g/m3),chassis resistance line(m),blockage length(m),hole spacing(m)and row spacing(m).Block size(<20 cm,<40 cm,<60 cm,<80 cm)and block rate(%)are selected as output parameters.The back-propagation neural network is trained by using MATLAB software.The blasting parameters are optimized.Finally,the optimized bench blasting parameters are tested and adjusted according to the actual situation.Finally,themost reasonable bench blasting parameters in each mine blasting area are determined,and the economic value saved by this method is calculated.By classifying the blastability of Heidaigou Open-pit Coal Mine,BP Neural Network is used to establish the prediction model of blasting parameters in accordance with the mine.Different parameters are obtained for different blasting areas.The model is reasonable and meets the requirements.It can effectively improve the production efficiency of the mine,reduce the mining cost and greatly increase the economic benefits of the mine.