Cooperation between MPC Algorithms and Set-Point Optimisation Algorithms

Abstract

This chapter is concerned with cooperation between the suboptimal MPC algorithms with on-line linearisation and set-point optimisation algorithms. At first, the classical multi-layer control system structure is discussed, the main disadvantage of which is the necessity of on-line nonlinear optimisation. Three control structures with on-line linearisation for set-point optimisation are presented next: the multi-layer structure with steady-state target optimisation, the integrated structure and the structure with predictive optimiser and constraint supervisor. Implementation details are given for three classes of neural models.