Research on intelligent optimal control strategy for the flash calcination process of magnesite in a transport bed based on deep reinforcement learning

The flash calcination process of magnesite in a transport bed involves complex mechanisms characterized by nonlinearity and coupling, making it challenging to establish an accurate mathematical model. Therefore, operators struggle to maintain the product quality indicators-loss on ignition (LOI) of light-burned magnesium (LBM) within the target range. To address these challenges, this study proposes an intelligent optimal control strategy based on the Deep Deterministic Policy Gradient (DDPG) within the framework of deep reinforcement learning. The proposed strategy achieves stable control of the LOI of LBM, optimizing product uniformity in the flash calcination process of magnesite in a transport bed.