A composite control scheme consisting of modepredictive control （MPC） and disturbance observer （DOB） iproposed to solve the control performance degradationproblem of the turbidity of the treated water in the presence osignificant changes in raw water quality, water flow rate andinternal model mismatch disturbances. The MPC is employedas a feedback controller for the coagulation process with alarge time delay. The DOB is adopted to estimate the severedisturbances in the turbidity control, such as large changes inraw water quality and water flow rate. The estimated valuesare applied for feed-forward compensation to rejecdisturbances. Finally, the disturbance rejection performancesfor step disturbances and time-varying disturbances in thenominal case and model mismatch case are tested. Thesimulation results illustrate that, compared with the MPCmethod, the proposed method can significantly improve thedisturbance rejection performance in the turbidity control othe treated water, no matter if in the presence of externadisturbances or internal model mismatch disturbances.
