| Coal industry is an important basic energy pillar industry in China,and the mine scale and production capacity keep rising steadily year by year.Heidaigou open coal mine,the first mine in China that has successfully adopted the technology of throwing blasting and pulling bucket shovel to dump and stack,one of the largest open coal mines with the largest output of coal,is pursuing the maximum of production capacity and efficiency.Throwing blasting technology,as an important production segment,needs to be continuously optimized and upgraded.Considering that the shape of the muck-pile is the key factor to evaluate the throwing blasting technology,this paper focuses on the prediction of the shape of the muck-pile of throwing blasting.First.Throwing blasting involves complex parameters,so it is difficult to distinguish the main influencing parameters in different scene condition.And there is a lack of specific analysis and design of the key parameters that affect the shape of muck-pile.Therefore,it is necessary to study the key influencing parameters or evaluation indexes of the current working area,such as working face elements,blasting design parameters and muck-pile shape data,and demonstrate the safety parameter values.At the same time,it is necessary to evaluate the influence of the location of the blasting on the shape of the muck-pile.Secondly.On the one hand,a large number of statistical samples of field experiments are needed to support the analysis and study of the morphological characteristics of muck-piles.On the other hand,the samples need to be classified and screened to ensure the accuracy of prediction.However,due to the large amount of data,there are many problems in the current research work,such as the lack of systematic sample selection methods.Therefore,the standardized analysis,treatment and regression methods of blasting shape are put forward to screen the shape characteristics of working face and blasting shape of a large number of samples for many times.And the current regression method of local weighting of sample set is put forward.On the premise of retaining the local detail characteristics of blasting curve,a typical blasting curve with universal interpretation significance is obtained,which is blasting curve The study of reactor configuration provides a fast analysis method.Thirdly.In the study of the shape control of the muck-pile,only the overall effective throwing rate is often used to evaluate the effect of the blasting program in the actual project.There is no further mining for the relationship between the better muck-pile shape and its corresponding design parameters,and the feedback and optimization of the parameters in the program from the shape data of the muck-pile are ignored.Therefore,research and design a three-layer neural network prediction model with adaptive erreor reduction and weight adjustment.Taking the shape parameters of the typical curve and the design parameters corresponding to the curve as input and output nodes,the mapping relationship of sample data is established through Matlab software.The key construction parameters of the optimization neural network model are studied to reduce the calculation error of the design parameters value,and ensure the optimization effect of the new blasting design scheme.The results show that 77.8% of the samples can be explained by the typical explosion curves of different step heights after screening and grouping,and it is suitable for the pre design of the construction scheme of the inverted platform.At the same time,using the network model established by each parameter,the prediction error is within 4.53%,and using the typical explosion curve to predict the blasting parameters,the comparison error between the output parameter value and similar historical data is within 6.56%.The research are applied in Heidaigou open coal mine,which improves the shape of muck-pile and effect of throwing blasting,and then improves the production efficiency of dragline. |
PDF Full Text Request