Design of neural networks based sliding mode control and observation: an overview

This paper presents an overview of some sliding mode control and observation schemes based on the joint use of neural networks (NNs) and the so-called integral sliding mode (ISM) algorithm. More precisely, the utilization of a NN-based approach is aimed at estimating the unknown dynamics of the process, thus providing a model for the design of the ISM strategy. Both the ISM controller and observer syntheses are discussed in this work. Moreover, assuming the presence of state and input constraints, a recent version of a control scheme based on model predictive control (MPC) and NN-based ISM is presented. The convergence performance of the proposed approaches are illustrated and assessed in simulation.