This paper proposes the application of multiobjective model predictive control to achieve a dynamic optimization of boiler combustion system. The utopia point of the multi-objective model predictive control system is firstly calculated, and the optimal control sequence is then determined by finding the trade-off solution closed to the utopia point. By using this approach, the arbitrary weighting functions, which are generally used in the multi-objective model predictive control approach are not necessary, and moreover, the procedures of pareto front calculation and final solution decision in the conventional multi-objective optimization can be avoided. To deal with the nonlinear problem, multi-model strategy is involved to get a global model for optimization. Simulation results demonstrate the effectiveness of the proposed multi-objective model predictive control.