The mill load control for grinding plant based on fuzzy logic

This paper considers the problem of controlling the mill load for a grinding plant The load state is essential to the powder production and quality of the mill. However, the mill load fluctuates because of many parameters associated with the material filling-rate, the size distribution of material to be ground, the coarse grain re-feeding amount, and material hardness. Additionally, the grinding process complex dynamics, nonlinearity, and uncertainty render traditional control techniques difficult to apply. Here, a fuzzy controller with real-time self-learning ability is employed to improve both the system performances and the adaptability. The control approach is implemented on a new type of ultra-fine grinding plant. Experimental results show that this control strategy maintains mill load stability and the required grain fineness distribution is obtained.