Optimization of hydrocyclone operation using a geno-fuzzy algorithm Original Research Article

This paper describes the use of genetic algorithms to optimize the design and performance of hydrocyclones. Hydrocyclones are simple mechanical devices that use centrifugal forces to separate materials by density or size. Here, a genetic algorithm is used to accomplish two specific tasks. First, it is used to tune a fuzzy logic-based computer model of a hydrocyclone. The extensibility of the model is tested by simulating the performance of a hydrocyclone circuit. Second, the fuzzy model is incorporated into an objective function used to drive a separate genetic algorithm seeking to optimize hydrocyclone circuit performance. Results indicate that the genetic algorithm is effective in both modeling hydrocyclones and hydrocyclone circuits, and subsequently in optimizing the performance of a hydrocyclone circuit.