Steady-state sequence optimization with incremental input constraints in two-layer model predictive control

Steady-state optimization is of vital importance in two-layer model predictive control for bringing better steady-state and dynamic performance. However, the global optimality of steady-state sequences provided by local steady-state optimization cannot be guaranteed. Therefore, a new steady-state sequence optimization approach is proposed in the paper, to improve the global optimality of steadystate sequences. First, the non-global optimality of local steady-state sequences is discussed using an example. Subsequently, aiming at improving the global optimality, a novel sequence optimization strategy designed for steady-state optimization is proposed. Its basic formulation is given and the lower bound of the introduced parameter is analyzed. Then, the relation and difference between the proposed steady-state sequence optimization and the existing global steady-state optimization and local steady-state optimization are discussed. Finally, the steady-state performance, dynamic performance, and computational burden of the proposed approach are studied. The proposed approach provides engineers a brand-new way to realize steady-state optimization and effectively improves the global optimality of calculated steady-state sequences. Extensive simulations verify the effectiveness and reliability of the proposed method.