Prediction Of Blasting Particle Size And It’s Relationship With Loading Efficiency

Blasting is one of the most important part of the work in exploitation of rock-fill dam.Blasting particle size not only affects the excavation and loading efficiency of mining materials,but also has a great influence on the loading efficiency.Therefore,It is one of the important measures for real-time blasting control that adjusting blasting design parameters to control the particle size distribution of mining materials.Based on the analysis of the influence of blasting parameters on particle size,it is proposed a neural network model of LM algorithm.In addition,in order to more better reflect the impact of different blasting parameters on blasting particle size,considering rock properties of a neural network model is proposed.At the same time,based on the mathematical model,the shovel efficiency of loaders is predicted.Through hydraulic engineering of blasting particle size analysis,the validity and practicability of the neural network model and mathematical model of shovel loading efficiency are verified by an engineering blasting test example.The research contents and achievements in this paper are as follows:（1）Blasting parameters have a great influence on the particle size after blasting,and the value of blasting particle size will directly affect the loading resistance,and then affect the loading efficiency of loaders.Therefore,firstly,it is analyzed quantitatively that the influence of blasting parameters on blasting particle size;secondly,the influence of blasting particle size on loading resistance of loader is analyzed qualitatively.This paper provides a basis for further research of blasting particle size and loading efficiency.（2）In view of the shortcomings of traditional empirical formula and BP neural network model,it is proposed that a method of predicting the particle size of blasting based on LM algorithm in this paper.At the same time,in order to more better reflect the completeness of blasting parameters,it is considered that the influence of rock parameters on blasting particle size in this paper.The neuron network models have been established which have the input layer seven parameters include the ratio of bench height and burden,the ratio of space and burden,the ratio of burden and diameter,the ratio of stemming and burden,the specific charge,block size in situ,rock elastic modulus,X₅₀ as output layer.It is analyzed and compared with the prediction error values of BP neural network,one hidden layer LM neural network model and two hidden layers LM neural network model.Through the evaluation of the model,it can be seen that the neural network model of two hidden layers LM algorithm has high accuracy.Meanwhile,the sensitivity of input layer parameters to output layer parameters is analyzed.Through the sensitivity analysis results,the reasonable selection of blasting design parameters can be optimized and guided to obtain the materials that meet the requirements of particle size and gradation.（3）Aiming at the condition of earthwork excavation engineering and the difference of the blasting particle size on the test field size value,and according to the actual demand in the spot.Two different types of loaders,2m³ and 1.6m³,are used to analyzing the shoveling efficiency.It is established that mathematical models between characteristic particle size and digging cycle time,as well as between characteristic particle size and bucket fill factor respectively.Based on the established model,the loading efficiency is predicted and estimated.Finally,It is carried out that the error analysis between the prediction value of loading efficiency and the actual loading efficiency of engineering blasting test.It provides a basis for the loading efficiency of blasting test in practical engineering,in order to guide the engineering needs.