Abstract: The hydrocyclone separation process (HSP), as a two-input/three-output system with such complex characteristics as strong coupling, high nonlinearity and frequent variation of boundary conditions, is difficult to be expressed by an accurate mathematical model. An intelligent control method is therefore proposed for HSP by integrating the modeling with control process so as to combine the PI control, fuzzy control with RBR (rule-based reasoning). Then, the CAC(cascade averaging controller) based on the fuzzy control algorithm for the liquid level in pumping sump and ore feeding pressure of hydrocyclone, the control of abnormal liquid level based on RBR, the switching controller for slurry pump speed and the PI control for the ore feeding concentration by hydrocyclone are all available to implement. The application results in a large-scale domestic beneficiation plant revealed that the method proposed can stabilize the HSP effectively with particle sizing index evidently improved.