Regression Analysis Of Mine Ventilation Resistance Coefficient Based On PSO-SVM And Correlation Analysis

The coefficient of ventilation resistance is an important basic parameter for the safety of ventilation.The acquisition of ventilation resistance coefficient needs to be measured in each tunnel for a mine,which has a large amount of work and a long time consuming;Aiming at this problem,theoretical analysis,data analysis and simulation experiments are used to analyze the main factors affecting ventilation resistance coefficient.According to a large amount of historical data,using the mathematical modeling software,the regression model between the ventilation resistance coefficient and the influence factors is established by using the particle swarm optimization support vector machine.The regression ventilation resistance coefficient of the regression ventilation coefficient of the single influence factor modeling to the combination influence factor is studied,and the particle swarm optimization is compared.The accuracy of the regression model is analyzed,and the correlation between the ventilation resistance coefficient and the influencing factors is analyzed,and the best regression model is determined.By using the PSO-SVM regression algorithm,the regression model of friction resistance coefficient and specific support property parameters is set up for three types of roadway supported by wood support,I-steel support and anchor net support.The results show that:(1)in the regression analysis of limited samples,the regression performance of the PSO-SVM model is better.When the optimal model can reach the relative error of less than 5%,the accuracy can reach 90% and the average relative error of the model is 0.83%.(2)the correlation analysis shows that the ventilation resistance coefficient is low related to the single influence factor and has a moderate correlation with the two combination factors,which is highly correlated with the three combination factors,that is,the input variables of the optimal regression model of the support vector machine using particle swarm optimization are: Lane fault area,roadway circumference and friction resistance system.The output variable is the ventilation resistance coefficient.(3)the average relative error of the regression method of PSO-SVM regression algorithm for wood supporting laneway,I-steel supporting roadway and anchor net supporting roadway are-2.542%,0.483% and 1.605% respectively.The results show that the support vector machine algorithm has better regression performance in the nonlinear regression analysis of limited samples.The PSO-SVM regression model can more accurately regress the ventilation resistance coefficient and provide a new intelligent algorithm for the regression of ventilation resistance coefficient,which has important guiding significance for practical application.