Auto-Regressive with Exogenous Input Modelling Based Model Predictive Control of Cement Grate Cooler

In the cement production process, the speed of grate coolers directly affects the residence time and cooling effect of cement clinker. Its effective control is of great significance to ensuring cement quality. But the adjustment of grate speed is a complex process, involving moderate cooling air, uniform clinker thickness, regulated output and other constraints. In this study, an Auto-Regressive with exogenous input (ARX) grate cooler prediction model is developed using a data-driven approach based on a company's cement production line, a model predictive control (MPC) approach is employed for the purpose of optimising the speed of the grate cooler based on the ARX model. The results demonstrate that the method of MPC exhibits greater stability than the method of PID, and that MPC technology has good efficacy and superiority in the adjustment of grate speed.