Intelligent Operational Adjustment of Copper Flotation Process Based on CBR-GRNN

The mechanism of the copper flotation process is complex and highly uncertain, making it difficult to establish a precise mechanism model. For a long time, the operational adjustment of the flotation process has mainly relied on manual labor, which makes it difficult to ensure mineral quality and economic benefits for the enterprise. In order to achieve automation of the flotation process operation, this article proposes an intelligent operational adjustment method based on case-based reasoning and the generalized regression neural network (CBR-GRNN). In the offline stage, the working conditions are divided based on PCA and K-means method, and the case bases are established for each type of different working conditions. In the online stage, we first determine the category of the target case and search in the corresponding case base to obtain the most similar operational adjustment solution. Due to the fact that the most similar case may not fully match the target case, further compensation is necessary. Therefore, this paper proposes a new difference-driven case revision scheme using GRNN. The final solution for reuse is the operational adjustment value compensated by GRNN. Finally, the experimental results in an actual copper flotation process have demonstrated that the proposed method can effectively perform operational adjustment. Furthermore, the superiority of the proposed method is further demonstrated by comparing it with the other two techniques.