Abstract: To optimize the concentrate grade in the beneficiation process of non-ferrous metals， a fusion algorithm combining fuzzy logic (FL) and immune genetic algorithm (IGA) was developed. Based on this fusion algorithm， a detection model for the beneficiation of non-ferrous metals was constructed to detect and optimize the concentrate grade. Comparative test results show that the data recall rate of the IGA-FL fusion algorithm is 99.7 ％， with a computation speed of 16.7 bps. The average detection accuracy of the model based on this algorithm is 97.3 ％， with a detection time of 1.8 s. After applying the detection model based on the IGA-FL fusion algorithm， the concentrate grade of non-ferrous metal beneficiation reached 70.5 ％， indicating that this model can optimize the concentrate grade effectively.