Modeling Based on the Extreme Learning Machine for Raw Cement Mill Grinding Process

Abstract

Vertical grinding mill is the main grinding equipment for the new-type dry cement raw meal production, raw material grinding process in cement industries accounts for approximately 50–60 % of the total energy consumption. The dynamic characteristics of the variables in the raw material vertical mill grinding process are strongly coupled, nonlinear, and large time lag. The process of parameter adjustment requires too much human intervention, it is difficult to establish a precise mathematical model. To address these problems, we use extreme learning machine network, establish production quotas predictive model of cement raw material vertical mill grinding process, combined with the cement raw material vertical mill grinding process data obtained from a cement plant, the model is trained and tested. Experimental results show that the proposed modeling method is effective to achieve the online estimation of the key indicator parameters for the vertical mill grinding process, lying foundation for parameters optimization online of the vertical mill grinding production process, and providing reference value for the energy consumption reducing.