Prediction of self-moving crusher's production based on multiple linear regression-BP neural network

The equipment running time,explosives consumption and shovel cycle time are selected to be the quantifiable argument,and the production capacity of the system is selected to be the dependent variable to establish a multiple linear regression equation,based on analyzing the factors affected the self-moving crusher system's production.The equation can predict system production.The BP neural network is established to adjust residuals of the multiple linear regression model,used with the feature of the model in nonlinear fitting.The prediction accuracy is significantly improved.The error of multiple linear regression model is 7%,and the error of the modified model is 1.42%.