A Dynamic Simulator for Evaluating Control Schemes in an Industrial Scale Rod Mill-Ball Mill Grinding Circuit

"A review of the dynamic rod mill-ball mill grinding circuit simulator developed at the University of Alberta, highlighting the modular structure and open-ended design, demonstrates the ease with which classical feedback control strategies can be designed and implemented. Various open loop simulations are presented to demonstrate the highly interactive nature of the grinding circuit and to indicate where potential control problems may be encountered. Closed loop simulations are examined, beginning with the basic SISO classical feedback control of circuit product size. Further improvement in the grinding circuit performance in the presence of various disturbances is realized through the application of a classical MIMO feedback controller. As an example of advanced (multivariable) control, static decoupling is implemented and shown to provide additional improvement in overall circuit performance.INTRODUCTIONIt has long been recognized in the processing industries that a dynamic process simulator is an indispensable tool in the design and evaluation of automatic control schemes. In mineral processing the lack of these simulators is felt to have been a major hindrance to the widespread implementation of effective process control systems. In particular dynamic simulation is a necessary precursor to any study of modern (as opposed to classical) process control methods. This paper discusses the development of a dynamic simulator that can be used to study various feedback control strategies. The simulation software provides a powerful educational tool for those individuals interested in grinding circuit control."