Data driven control of underflow slurry concentration in deep cone thickener

Deep cone thickener control problem is a key point of Tailings paste fill (TSF). This paper presents a new method to extract inherent and practical parameters of the thickener, and determine control strategy based on thickening process data mining. Bypassing difficulty in deep cone thickener modeling, the proposed method could obtain practical control rules, also has good adaptability to different thickener structures and backfilling materials.