Nonlinear model-algorithmic control for multivariable nonminimum-phase processes (Q2776230)

A novel model-algorithmic control structure with nonminimum-phase compensation for general nonlinear multivariable discrete-time systems is presented. A systematic procedure is outlined for the selection of particular synthetic outputs, which are used to design the model-algorithmic controller. All synthetic outputs are constructed so that they are statically equivalent to the actual process outputs and depend on some free weighting parameters. These parameters are used to place the transmission zeros of the linearized system at pre-assigned locations, in order to meet the local minimum-phase requirement of a model-algorithmic control. The results of a simulation case study on the non-isothermal continuous stirred-tank reactor with a serial/parallel van der Vusse reaction show that the implementation of the proposed controller will lead to excellent setpoint tracking and regulatory behaviour.NEWLINENEWLINEFor the entire collection see [Zbl 0976.00019].