Study On Prediction Of Fragmentation Distribution Of Bench Blasting In Beskuduke Open-pit Coal Mine

The distribution of blasting lumpiness is an important index to judge the quality of blasting effect.Reasonable blasting lumpiness can improve the efficiency of shoveling and transportation,and has an important effect on the economic benefit of the mine.Because blasting is a complex discontinuous process and is affected by many factors,it cannot be described by certain formulas and models,and the relationship between the distribution of blasting lumpiness and its influencing factors is difficult to summarize.In actual production,judgment is usually made according to experience,but it is still difficult to meet the actual production demand.Therefore,it is of great significance to deeply study the distribution of blasting lumpiness for the actual production of mines,and at the same time promote the development of blasting related theoretical research.Intelligent algorithm has developed rapidly in recent years.It has strong nonlinear processing ability,adaptive ability,generalization ability and learning ability,and has been widely used in many fields.Using intelligent algorithm to solve the complicated problem in blasting process has good effect.In this paper,the biesikuduke open-pit coal mine is taken as the engineering background.In view of the uneven distribution of blasting lumpiness,the existence of large blocks and roots,through theoretical analysis,computer programming and other means of in-depth research,mainly divided into the following aspects:(1)Theoretical analysis is made on the factors affecting blasting effect,and the influence of these factors on blasting is discussed.Through the improved AHP algorithm,the quantitative analysis of the impact of different blasting factors in the open pit coal mine of biesikuduke,the main factors and secondary factors are obtained.(2)PSO-ELM blasting fragmentation prediction model is established.Eight factors including rock tensile strength,rock compressive strength,joint frequency,hole spacing,row spacing,minimum resistance line,ultra-depth and single explosive consumption were selected as the input vector of the prediction model.The block size distribution of blasting is selected as the output vector of the prediction model: the percentage of block size less than 20 cm,the percentage of block size less than 35 cm,the percentage of block size less than 50 cm,and the percentage of block size less than 65 cm.The maximum error of the prediction is 5.57%,which is better than the traditional ELM model,and the result is close to the actual value.(3)The blasting parameters of the 1252 platform in the mining area were optimized,and the prediction and field test were carried out based on the PSO-ELM blasting block distribution prediction model.The blasting quality was improved after optimizing the blasting parameters,which effectively guided the actual production of the mine and improved the economic benefits of the biesikuduke open-pit coal mine.