Optimization Design Method And Application Of Filling Ratio Based On Neural Network And Genetic Algorithm

For a long time,metal mine tailings are almost all placed on the surface tailings piled up,due to tailings carrying excessive pollutants,directly to the mining area caused by damage to the ecological environment.With the continuous improvement of environmental requirements,tailings in the surface storage caused by pollution problems become increasingly prominent,tailings emissions and governance has become a major factor in the development of metal mines.Filling mining method can be used as far as possible to tailings backfill wells to solve the problem of insufficient capacity of ore tailings,reduce and eliminate tailings in the surface storage of environmental pollution and other advantages can be widely used.In the process of filling mining design,the design of filling slurry is the primary factor to determine the quality of filling.Under the premise of controlling the total cost of filling,the reasonable ratio of slurry can be effectively guaranteed to ensure the strength of the filling body,Process requirements.Therefore,to determine a reasonable filling slurry ratio is to ensure safe and efficient economic recovery of the important prerequisite. Select the reasonable ratio of slurry,generally through the laboratory comprehensive ratio test,measured under different matching parameters of the performance of the filling body,and then recommend to meet the mining method strength requirements of the matching parameters,but the comprehensive experimental workload and vulnerable Human operation and other factors,in the actual production and application has some limitations.In this paper,the orthogonal test scheme of filling strength is designed.The filling strength data of a large number of different proportions are obtained.The BP neural network and the existing prediction model of the filling strength are used to study the strength of the filling And the genetic algorithm is used to optimize the mix ratio of filling materials,so it is recommended to meet the requirements of different mining methods and meet the requirements of mine transportation conditions.The main contents are as follows:(1)Analyze and compare several commonly used filling strength models,choose to build a full-tail filling strength model based on BP neural network,and verify the former by checking the data samples.The results show that the prediction accuracy of the selected and established filling body strength model is high,and the strength of the filling body strength model is used to predictthe strength of the waste rock and tailings.Which can meet the accuracy requirements of the actual engineering to predict the strength of the whole tail filling,and it has a good guiding effect on the optimal design of the filling material ratio.(2)According to the theory of genetic algorithm,based on the model of filling strength prediction model,the optimal mix of filler material ratio is obtained,and the optimal solution set of filling parameters is obtained to meet the requirements of mining method.The selection method of the filling material ratio scheme with good anti-segregation and capable of realizing the self-flow transportation is recommended.For the different preparation requirements,only some parameters can be modified to obtain the optimal mix Than the parameters,greatly improving the filling material with the design efficiency and high accuracy.(3)On the basis of the above research,we use the object-oriented C#language to compile the intelligent decision-making system of filling material performance prediction and ratio optimization,which will interface with complicated computer language.The system has the characteristics of friendly interface,easy to use and simple operation Which provides a good decision support for filling material performance prediction and optimization design.It greatly improves the efficiency of filling design.