Aluminum Reduction Cell Three-dimensional Magnetic Field Calculation And Distribution Of Genetic Algorithm Optimization

The calculation of magnetic field plays an important role in designing an aluminum reduction cell. The distribution of magnetic field has a direct influence on the production index of aluminum electrolysis and the normal operation of the cell. In this paper the 3D magnetic field in a 180kA prebake anode aluminum reduction cell is calculated and optimized, and the calculated results are carried out visual processing.The paper discusses the influence of the cell's magnetic field on the surrounding electronic equipments and the working environment, and introduces several main methods used to calculate the cell's magnetic field. Based on the comparisons among several methods, the spline integral equation method is adopted. This method is formed with the introduction of the spline interpolation technique into the magnetization integral equation method, with which the magnetic field in TEAM Workshop 21st problem model B is calculated, the calculated results achieve a good agreement with the measured ones, the maximum error is 10.4%, the validity of the method is demonstrated.In this paper, the formulas of the conduction current producing magnetic field are deduced in detail, and the principle of applying the spline integral equation method to calculate magnetic field is represented. Based on this method, we have compiled the software with FORTRAN to calculate aluminum reduction cell's 3D magnetic field. On the condition of considering the influence of both the ferromagnetic substance and the anode column bar current or not, the magnetic field distributions in the cell are calculated and the distribution figures in every direction are plotted. From the analysis, it can be concluded that the influence of both the ferromagnetic substance and the anode column bar current should not be ignored and that the cell's magnetic field distribution is unreasonable.Based on the viewpoint of global optimization, the genetic algorithm (GA) is firstly used to optimize aluminum reduction cell's magnetic field. The basic principle and the improvements of GA are introduced, and an optimal programbased on an improved GA is compiled with MATLAB. Finally, the high-efficiency of the improved method is verified by the standard functions.According to the discussions of the factors about the cell's current efficiency, the optimal model of the aluminum reduction cell is built, which takes anode cathode distance, aluminum liquid height, inter-cell span as optimization variables, the symmetry of magnetic field distribution as optimized object. Then the cell's magnetic field is optimized by GA, and the bus-bar configuration is improved as well. It is shown by the optimized results that the maximum magnetic flux density in cell's long-axis direction decreases 43.85%, the maximum magnetic flux density in cell's short-axis direction decreases 20.47%, the average magnetic flux density in vertical direction decreases 11.9%, the four area's average magnetic flux density in vertical direction are nearly equal, and the whole magnetic field distribution is more reasonable, so the melt flow rate will decrease and the current efficiency will increase.In the paper, the Graphical User Interface (GUI) used to visualize the calculated results of the magnetic field in the cell is designed with MATLAB layout editor, with which the cell's magnetic flux density distribution in each direction can be plotted instantly.