Operational Decision-Making Optimization of Aluminum Electrolysis Process Based on Health Evaluation of Aluminum Electrolytic Cell

The operational decision-making in aluminum electrolysis process can directly influence the current efficiency, energy consumption, and stability of the process. This paper proposes an Operational Decision-Making Optimization (ODMO) method based on health evaluation of aluminum electrolytic cell, which provides the refine aluminum fluoride addition, setting voltage and trapping aluminum for aluminum electrolysis process. Specifically, when the aluminum electrolytic cell is in a healthy state, we establish the ODMO based on behavior clone method. when the aluminum electrolytic cell is in an unhealthy state, we proposed to apply the offline reinforcement learning to achieve ODMO. Finally, the performance is confirmed on the real-world AEP. Experimental results demonstrate that the proposed scheme provides a satisfactory performance.