Application of Intelligent Optimization Algorithm in Magnetic Roller

In order to improve the separation efficiency of eddy current separation in the field of scrapped vehicle recycling, a Halbach array permanent magnet roller was proposed based on simulation analysis. Firstly, 29 groups of simulation data were designed by Box-Behnken test, taking the peak magnetic induction intensity 5 mm away from the surface of the magnetic roll as the objective function and the inner diameter (Rin), outer diameter (Rex), thickness (W) of a single permanent magnet and the number of magnetic poles (k) of the magnetic roller as variables. Secondly, Secondly, these 24 sets of data and 5 sets of data were used as the training data set and testing data set of the random forest algorithm to construct the random forest regression model, and the model fit degree R2 was up to 0.9936. The bat algorithm was used to optimize the geometric parameters and the number of magnetic poles of a single permanent magnet. The optimal parameters were Rin = 95 mm, Rex = 190 mm, W = 100 mm, k = 5, and the peak magnetic induction intensity was 0.9758T. The results show that the random forest algorithm and the bat algorithm perform well in the magnetic roll problem.