This paper proposed a double-layer model predictive control (DLMPC) strategy integrated with zone control. In the steady-state target calculation (SSTC) layer, the controlled output variables are pretreated based on the analysis of process principle and production data. Subsequently, the optimal input–output targets and corresponding zone parameters are obtained by solving the steady-state optimization problem and transmitted to the dynamic control layer for tracking. Meanwhile, the weighted and priority soft constraints relaxation methods are discussed when the SSTC problem is infeasible. Compared with conventional DLMPC and zone model predictive control (ZMPC), the proposed algorithm can not only achieve smooth control, but also have better robustness and economic performance. Finally, a simulation example is taken to show the effectiveness of the proposed method.