Statistical model and control of a ring-and-roller type grinding mill.

摘要

This project explores the use of a ring-and-roller grinding mill for fine grinding or pulverizing. Since the efficiency of this system is low, small improvements in product throughput or energy consumption can result in substantial economic improvements. To better understand this system, experiments were conducted for limestone and cement clinker. From these experiments models were formulated for product throughput, energy consumption and product size distribution. A control loop was implemented to reduce the effect of non-uniform feeding. Steady-state interactions were analyzed using the Relative Gain Array (RGA) method. Next, a multi-input and multi-output (MIMO) control system was designed using the Inverse Nyquist Array (INA) method to remove variable interactions. The robustness of this controller was explored. Finally, a self-tuning PI controller was designed using the pole placement method to improve system performance. The results of this study provide models for implementation on ring-and-roller grinding mills.