Feature Selection for Industrial Flotation Process Performance Monitoring Based on Multi-Swarm PSO

Flotation process is an extremely important technology in the complex metal metallurgical industry. In flotation., control based on foam image is often used., and foam image features often reflect the performance of flotation process. Therefore., this paper focuses on the control of foam image in the zinc flotation process., and proposes a Feature selection method for foam image based on multi particle swarm search. A large number of image features can be preliminarily extracted from foam images to characterize foam flotation., but there are some redundant and irrelevant features that may affect the evaluation results. In this paper., the Feature selection method of minimum redundancy maximum correlation criterion (mRMR) is used to select the most favorable feature result. The simulation results show that compared with the traditional expert experience Feature selection., the proposed multi particle swarm cooperative search strategy has higher performance., and the evaluation index values calculated by it are significantly better than those obtained by expert experience., which shows that the method proposed in this paper has high practical value.