Hierarchical optimization control for cascade industrial systems: A coordinated SSTO- and adaptive MPC-based approach

This paper investigates the problem of hierarchical optimization control for complex industrial systems. To address the challenges of infeasible setpoints between real-time optimization (RTO) and model predictive control (MPC) layers as well as the challenges of parametric uncertainties in the dynamic control, a hierarchical optimization control framework that integrates coordinated steady-state target optimization (SSTO) with adaptive MPC is proposed. First, a global coordination strategy based on genetic algorithm (GA) is developed to solve nonlinear SSTO problems and determine coordinated operating points. Second, an adaptive MPC scheme is designed, which incorporates an online parameter updating law together with a constrained MPC formulation for the estimated system. This design guarantees accurate tracking of the coordinated operating points while ensuring recursive feasibility and closed-loop stability. Simulation studies validate the effectiveness of the proposed framework, thereby exhibiting significant enhancements in economic performance, constraint feasibility, and robustness.