Sensitive Feature Selection for Industrial Flotation Process Soft Sensor Based on Multiswarm PSO With Collaborative Search
Shiwen Xie, Yongjia Yu, Yongfang Xie, Zhaohui Tang
Abstract
Concentrate grade and recovery are key p roduction indexes for industrial flotation process. To establish the soft sensor model of the concentrate grade and recovery, a lot of froth image features are extracted as the input variables. However, these image features contain some redundant and irrel evant features. To improve the efficiency without degrading the performance of the soft sensor model, a sensitive feature selection method is proposed in this paper. Sensitivity coefficient is defined to we igh the attribute significance of features to labe l, which is calculated by gray correlation analysis. Then, the criterion of sensitive feature selection based on maximum relevance minimum redundancy (mRMR) is proposed. To solve the feature selection problem, a multi swarm particle swarm optimization with collaborative search (CS PSO) is developed. Information exchange mechanism among three particle swarms in collaborative search is proposed to improve the search effect and search accuracy. Self adjusting s tructure RBFNN is employed to establish the soft s ensor model to predict the concentrate grade based on the selected froth image features. The effectiveness of the proposed method is validated by the industrial flotation process data by comparing with other methods.