Performance and stability analysis of LP‐MPC and QP‐MPC cascade control systems

Modifier-Adaptation Methodology for Real-Time Optimization

“…Concerning the strategy for integrating constraint adaptation with an MPC constraint controller proposed in Chapter 3, the emphasis is not on MPC, but on the integration strategy from a methodological point of view. The analysis of the integrated scheme, as well as a comparison with other strategies such as LP-MPC and QP-MPC [88], are beyond the scope of this thesis. In order to estimate the experimental gradients required by modifier adaptation, only steady-state perturbation approaches are considered in this thesis.…”

“…The outer controller, which can be a linear program (LP-MPC) or a quadratic program (QP-MPC), uses a steady-state version of the prediction model used by the MPC to correct the setpoints passed from the RTO level to the MPC. The comparison of the advantages and disadvantages of the proposed integrated scheme with LP-MPC and QP-MPC (see [88]) is an interesting subject of future research. While both approaches appear to provide offset-free control performance, the two-stage MPC requires an additional LP (or QP) level.…”

A steady-state target calculation method based on “point” model for integrating processes

“…In order to maintain the process operation as close as possible to the economic optimum, Marlin et al created a direct link between regulatory control and economic optimization of the plant based on hierarchical structure, and gave set-points of the lower-level control loop adopting real-time optimization (RTO) [2]. In order to narrow the gap between the low frequency nonlinear steady-state optimization performed on the RTO layer and the relatively fast linear MPC layer, the so-called LP-MPC and QP-MPC (both have two-stage MPC structures) are frequently used in industry [3][4][5], which are cascaded structures of a steady-state optimizer and the typical MPC controller. In these structures, set-points of MPC controller are upgraded by the steady-state optimization in real time.…”

Optimization and control of a continuous polymerization reactor

“…The second layer is the target calculation (TC) algorithm that, at each time step, re-computes feasible steady-state operating points assuming that the RTO routine produces piecewise constant optimizing references for inputs and outputs. As discussed in Ying and Joseph (1999), the main purpose of the target calculation routine is to compute achievable set-points for the MPC controller. The targets Then, the solution of the TC stage is sent to the MPC controller, which is devoted to guide some of the reactor inputs and/or outputs to the desired values given by the TC stage, while keeping the other reactor controlled outputs within specified zones.…”