Abstract: The simulation system of grinding and classification process can realize an on-line detection of product size under various operating conditions, providing guidance to the optimization of operating parameters. The ball mill is an important equipment unit in the grinding-classification circuit and therefore, the mathematic model of ball mill is the basis of this process simulation system. The paper uses the theory of the grinding kinetics and material balance to analyze the relationship between the practical operating parameters and mill power consumption, introduces the selection function to establish the dynamic mathematical model of ball mill, establishes the mill power consumption prediction model based on gray-box neural network, constructs the simulation model by Matlab's Simulink module and makes the simulation analysis of the model. The model efficacy and disturbance tests have verified the effectiveness and rationality of the model.

Key words: Ball mill, Dynamic model, Gray-box neural network model, Simulation