Improved EDMC for the processes with high variations and/or sign changes in steady‐state gain

Abstract

EDMC represents extended dynamic matrix control, which can be applied to nonlinear process control. In this method, control inputs are determined based on a linear model that approximates the process and is updated during each sampling interval. Since nonlinear relation still exists between the prediction error and the control input, numerical (iterative) methods are used to solve the optimization problem defined in the method. For nonlinear processes with high variation and/or sign changes in their steady‐state gain, iterative methods do not converge properly to an acceptable solution for some desired outputs or external disturbances. To eliminate the problem, we augment the process with its steady‐state gain inverse (or pseudo inverse whenever required) such that the steady‐state gain for the new augmented system is constant or contains slow variations. In the case of unstable processes, the method may be applied after stabilizing the process using a proper state or output feedback. Effectiveness of the method has been examined using computer simulations of some benchmark processes. Some of the obtained results are presented in this paper.

Citation

Haeri, M. (2004), "Improved EDMC for the processes with high variations and/or sign changes in steady‐state gain", COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, Vol. 23 No. 2, pp. 361-380. https://doi.org/10.1108/03321640410510541