NobleBlocks

Research index

Model Predictive Control Methods

Patan

3 Dec 2021

NobleIDark:/48914/w1/20260515/B3AAA084DOI10.1002/9781119882350.ch4

Paywalled

The full text is at the publisher

You can read the abstract here, or open it at the publisher.

Read at the publisher

Abstract

This chapter is devoted to some new approaches to robust model predictive control (MPC) based on artificial neural networks. It describes two robust MPC schemes: robust nonlinear MPC where the optimization process is realized in the form of nonlinear programming problem, and robust approximate MPC formulated in the form of quadratic programming. The possible sources of model uncertainty may be split into: parametric, neglected and unmodeled dynamics, and lumped. One of the possible methods to represent uncertainty is to use the so-called model error modeling. One of the possible ways to achieve robustness of MPC is to properly define output constraints. Pneumatic servo is composed of a double acting cylinder lifting an inertial mass. The modeling of the servo in open-loop control is troublesome. That is the reason for using closed-loop control for training data recording. The parameters of the robust MPC, and the constraints, are the same as those used for error model training.

Discussion

Citations

Topics

More

Suggest a correction or report