Grant: Collaborative Research and Development Grants

Most industrial semi-autogenous grinding (SAG) mills use heuristic control approaches such as expert systems to achieve stable operations and improved performance. Model-based control techniques have not received successful applications in industrial SAG mills due to complexity of the processes and lack of accurate models. Owing to the need to improve the performance of SAG mills beyond that of the experience-based heuristic control approach, this project proposes to develop high fidelity models for industrial SAG mills and investigate the effect of industrially important variables on their operation. Both fundamental modelling based on first principles and statistical techniques using industrial operating data will be used in model development. Model-based control will be developed and its performance will be investigated using simulations. Results of this project will generate important information for new control implementation as well as existing control improvement with minimal costs for industrial SAG mills.