Study on Nonlinear Internal Model Control Based Neural Networks: An Application to MIMO Non-Square Systems | Chapter 3 | Novel Perspectives of Engineering Research Vol. 4

 Internal Model Control (IMC) of discrete under-actuated and over-actuated non-linear systems is the subject of this book chapter. Because of their complexity, non-square systems create a number of challenges in terms of control. As a result, synthesising a non-linear internal controller is challenging. The proposed solution then combines the IMC structure with neural networks to make it easier to realise an approximate inverse of the non-linear model of the process to be controlled.

A neural network can be introduced in the internal model controller of the basic IMC structure using two methods: direct and indirect. To reflect the system's inverse dynamics, the neural network is trained using the direct method with the system's input/output data. The neural network depicts the system dynamics in the indirect method. For both overactuated and underactuated systems, the simulation results are satisfactory, demonstrating the efficiency of the suggested control technique in guaranteeing satisfactory nominal and robust performance.

Author(S) Details

Imen Saidi
Laboratory of Research in Automatic Control, University of Tuins El Manar, National Engineering School of Tunis, Tunis, Tunisia.

Islem Bejaoui
Laboratory of Research in Automatic Control, University of Tuins El Manar, National Engineering School of Tunis, Tunis, Tunisia.

Nahla Touati
Laboratory of Research in Automatic Control, University of Tuins El Manar, National Engineering School of Tunis, Tunis, Tunisia.

View Book:- https://stm.bookpi.org/NPER-V4/article/view/5103