Abstract: In the blending process of raw slurry in alumina production,it is difficult to control the mill load state with the traditional approaches.A hybrid intelligent control approach of mill load is proposed,which is composed of the estimation model of mill load state and the adjustment model of mill load.Based on the vibration and electric current,the mill load state is estimated by the estimation model where rule-based reasoning is adopted.The total amount of fed materials is adjusted automatically by the adjustment model where case-based reasoning is adopted.Therefore,the mill load can be controlled within an appropriate range.The proposed approach is applied to the blending process of raw slurry in an alumina factory.A long-term running results show that the production capacity of mill is increased,the mill blockage is decreased,and high production efficiency and energy saving can also be achieved.