State Space Model Predictive Control Using Partial Decoupling and Output Weighting for Improved Model/Plant Mismatch Performance

Focusing on multivariable control, this paper presents a design method of model predictive control that enjoys the benefits of both the partial decoupling and the state space design. The multivariable process is first decoupled into a set of multi-input single-output (MISO) structures and then transformed into an extended state space model (partial decoupling extended state space model, PD-ESS). Consequently, a systematic design of model predictive control is proposed. The proposed controller is tested on three typical cases for comparison with previous controllers. Results show that control performance is improved. In addition, a closed-form of transfer function representation that facilitates frequency analysis of the control system is provided for further insight into the proposed method