Double-layered model predictive control strategy with dynamic trajectory calculation

Abstract

In industrial process field, the integrations of model predictive control (MPC) and hierarchical control system are widely used. A new double-layered model predictive control strategy with dynamic trajectory calculation (DTC) in its upper layer is presented in this paper, which can be applied in the midst of a hierarchical control system. DTC includes a feasibility stage with prioritized constraints handling and an economic optimization stage in a compatible constraint set. The calculated trajectory for each output is tracked by MPC dynamically in the lower layer. The presented control algorithm with DTC inspired by the steady-state target calculation (SSTC) method from the existing research basis. It could achieve a better economic benefit and tracking effect of external targets (ET) obtained from real-time optimization (RTO). In the simulation example, the introduced and original algorithms are applied to a fluid catalytic cracking (FCC) model which is the core of refining industry. Control results and performance comparison between two methods illustrates effectiveness of the proposed control strategy.