A ALO-LSSVM Model for the Cell Voltage Optimization in Aluminum Electrolysis Process

Abstract

The aluminum reduction is a typical industry with high-energy consumption, and one of the ways to effectively solve the problem of high energy consumption is to reduce the cell voltage in a proper way. It is hard to express the relationship between the cell voltage and the corresponding technical conditions. In present, data mining technology highly automated analysis of enterprise data is usually used to find the optimal cell voltage and its corresponding optimal technical conditions. In order to improve the accuracy of the model, the method of density-based spatial clustering of applications with noise (DBSCAN) is adopted to remove the abnormal operating points from the raw data. Then, in order to realize the real-time measurement of cell voltage, the prediction model of cell voltage is established by least squares support vector machine (LSSVM), and optimized by Ant Lion Optimizer(ALO). Additionally, the cell voltage optimization control model is developed by analysis of the technical conditions. The cell voltage optimization control model is solved by using the ALO algorithm. The simulation results reveal that the adoption of ALO-LSSVM model significantly improves the robustness and accuracy of the cell voltage prediction, and it provides a powerful technical preparation for optimizing cell voltage. The experimental effect shows that the ALO model is stable, and the average value of cell voltage optimization is 3.8647v. It is used to guide process operation, and the theoretical DC power consumption is 14.9% lower than before optimization, achieving energy saving.