Advanced feedback control for optimal operation the mineral grinding process has been based on the model predictive control (MPC) dynamic optimization for years. Due to the MPC does not handle the disturbances directly by controller design; it cannot achieve satisfying effects in controlling the complex grinding processes in the presence of strong disturbances and large uncertainties. In this paper, an improved disturbance observer (DOB) based MPC advanced feedback control is proposed to control the multivariable grinding process operation. The improved DOB is based on the optimal achievable H2 performance and can deal with disturbance observer for the nonminimum-phase delay systems. In the presented DOB-MPC advanced feedback control, the higher-level optimizer computes the optimal operating points by maximize the profit function and passes them to the MPC level. The MPC acts as a presetting controller and is employed to generate the proper pre-setpoint for the lower-level basic feedback control system. The DOB acts as a compensator, and can enhance the process operation performance. It is used to dynamically compensate the setpoints of the basic control system according to the observed various disturbances and plant uncertainties. At last, several simulations have been performed to demonstrate the proposed control method for grinding process operation..