Analysis of coal distribution prediction strategy for construction based on MLP-DBN model
The distribution of structural coal is of great significance for the safety of coal mining activities.Therefore,in order to achieve accurate prediction of the distribution of structural coal,based on the characteristics of seismic attribute information of the construction of coal seams,a prediction model for the distribution of construction coal was proposed based on deep confidence networks and multi-layer perceptrons.The experimental results showed that the fitting degrees of the multi-layer perceptron deep confidence network model on noisy and non noisy datasets were 0.965 and 0.996,respectively.Compared with other models,the average coefficient of determina-tion and average explanatory variance score of the multi-layer perceptron deep confidence network model were 0.963 and 0.87,respec-tively,which were higher than other models.The average mean square error and root mean square error were 0.006 and 0.078,respec-tively,which were lower than other models.The above results indicate that the construction coal distribution prediction model based on MLP-DBN can more accurately predict the distribution of construction coal,and the fitting degree between the predicted results and the actual situation is higher,providing strong support for the advanced control of coal seam gas.
万方数据
维普
