Development of Regression Model for Predicting the Ball Loading of the Grinding Unit of a Mining Plant

The paper considers the possibility of using the apparatus of regression models for predicting the ball mill loading of the concentrator of the ore dressing plant of a mining and processing plant. This indicator has a significant impact on the operation of the unit and the quality of the obtained product. And its estimation is carried out only at stops of the unit, at the rest of the time the ball flow rate is assumed constant and additional loads are carried out discretely according to the schedule. Such an approach does not allow to take into account changes in the properties of incoming ore and drift of the unit parameters. The operator has to compensate for this disadvantage by changing the nature of the control process, which is not fully effective. The availability of the functionality that predicts the current level of ball loading will allow to continuously maintain the ball mill operation mode within the technological tolerances, which will have a positive impact on the quality characteristics of the unit operation.