Support Vector Machine And Particle Swarm Optimization Algorithm To Predict Backfill Strength

With the development of artificial intelligence technology,computer technology as an auxiliary application has achieved good results in various industries,and the mining industry also needs to enjoy the results.Most of the domestic mines have begun to turn from the open pit to underground.The most important thing is the use of backfill mining.Important part of the backfill mining method is the pipeline transportation of filler slurry and the compressive strength of the backfill.Therefore,it is of great significance to use computer technology to assist filling mining to explore the compressive strength of cemented fillings and the resistance of slurry pipeline transportation.Using computer technology to serve mines has become a new development direction.PSO-SVR is used to predict the strength of mine fillings and the Lite Flownet is used to estimate the movement of filling slurry pipeline transportation.The main research contents are as follows2 Aspects:1)Compare the prediction models in 5 and Realize the visualization of the filling body slurry movement based on the Lite Flownet.Through the visual model of filler slurry pipeline transportation,the influence of pipe diameter,slurry concentration and initial pipeline transportation speed on the pipeline transportation speed of tailings slurry is quantitatively analyzed.Combined with theoretical analysis,the law affecting the change of transportation speed is explained: As the diameter increases,the pipeline transportation resistance increases,and the transportation speed is lower;with the increase of the slurry concentration,the pipeline transportation resistance decreases and the transportation speed increases;the initial speed during pipeline transportation should not be too large.2)Established a model based on PSO-SVR to predict filling intensity.Through the use of physical analysis and chemical analysis,the main influencing factors of the cement tailings ratio,solid content and curing age of the filling body are selected as the input factors of the model.After PSO optimizes the parameters of SVR,high correlation coefficient training is realized The sets are(R~2(linear fit value)= 0.996)and test set(R~2 = 0.993),and lower MSE(mean square error)values training set is 0.000393,test set is 0.00072613),indicating that PSO-SVR is applicable,To predict the compressive strength of the filling body.Figure 34;Table 6;Reference 52...