Optimization of the performance of flotation circuits using a genetic algorithm oriented by process-based rules

Abstract

Simultaneous optimization of the performance and simplification of flotation circuits using experimental approaches is almost impossible. The logical way is to use a mathematical model to describe the process as accurately as possible along with an appropriate optimization method. In this research the flotation process was modeled using first order kinetics approach with discrete-distributed rate constants. The contribution of non-selective particles transfer mechanism to the concentrate because of hydraulic entrainment was also considered. Then, genetic algorithm was applied for both optimization of the circuit performance and simplification of the circuit. Since there are only few appropriate solutions in the search space which sounds logical regarding the flotation process, the search routine was oriented to satisfy four process-based rules. One of these rules stated as concentrate and tailing streams must not recycle between two stages. The gene mutation which changed chromosomes in an intelligent way based on flotation process not only reduced the search time but also provided simpler circuits. The algorithm was applied for two optimization examples with the objective of achieving a desired concentrate grade within a specific total cells volume. The comparison of the results with the published data indicated that the proposed oriented genetic algorithm decreased the calculation time by 1/60 for a two-stage flotation system and provided a simpler circuit with a similar performance. The algorithm was then used to optimize a flotation circuit with four stages processing 100 t/h feed containing three species with high, medium and low rate constants. The best separation efficiency for this circuit with the concentrate grade of 65% was found to be 82%.