A combinatory study on MPC with subspace identification, steady-state target calculation and dynamic quadratic control

The prevailing industrial model predictive control (MPC) has two layers, i.e., the steady-state target calculation (SSTC) and the dynamic quadratic control. In this paper, it is suggested that the combination of subspace identification method (SIM) and double-layered MPC is a trend of the industrial process control. The multivariable output error state space (MOESP) method is adopted in SIM, and the multi priority rank method is designed in SSTC. This SIM-based double-layered MPC is illustrated through an example of the heavy oil fractionator model.