Fuzzy Logic: a potential control technique for mineral processing

"The application of control theory in mineral processing is more difficult than in other industries because of the complex nature of the unit processes involved and because of problems with accurate on stream measurement of process variables. Conventional control is based on proportional, integral and derivative functions of a continuously measured error term. Computer control can enhance these actions by incorporating logical routines to switch from one relationship to another. The success of this logical approach depends on our understanding of the process dynamics and the stability of the process and model under all conditions. Typically the model does not apply at all times in mineral processing because of frequent and unmeasurable changes in the ore parameters.Recent advances in the theory of fuzzy sets have led to the application of this science to process control. Fuzzy Logic can be used to develop heuristic or rule-of-thumb controllers with the strategy based on the actual operating practice of experienced plant operators. Control action has increased flexibility through the adaptation of human reasoning. Acceptance of the control system by operating personnel is significantly higher because of their participation in its formulation. The requirement of an exact model is not essential.A heuristic controller has been designed and tested with a computer simulation model of a secondary crushing plant. The results of this study are discussed and future applications of this control technique to other mineral processes are suggested. IntroductionThe application of control theory in mineral processing is more difficult than in other industries because of the complex nature of the unit processes involved. Mineral processes typically are characterized by non-linear relationships between input and output variables. Many of these variables cannot be monitored accurately on-stream and so long time lags on the measurement loop are features of the system. The processes, themselves, often have long time-varying responses while the process loads fluctuate significantly on a short-term basis.Attempts to control these processes have led to the formulation of mathematical models to describe each unit activity. Adaptive techniques have also been tried to compensate and account for process changes and vageries. These approaches are mathematically rigorous so that their complexity and precision limit their usefulness when applied to ""vague"" processes."