Using Graphics Processing Units and Compute Shaders in Real Time Multimodel Adaptive Robust Control

Cosmin-Constantin Mihai,Ciprian Lupu

doi:10.3390/electronics10202462

Cosmin-Constantin Mihai, Ciprian Lupu

Open Access

https://doi.org/10.3390/electronics10202462

Copy DOI

Journal: Electronics	Publication Date: Oct 11, 2021
Citations: 2	License type: CC BY 4.0

Affiliation: Polytechnic University of Bucharest

Abstract

Graphics processing units and video cards have seen a surge of usage in domains other than graphics computers, due to advances in hardware and software technologies; however, little uptake has been in the domain of systems engineering and real time control. This research article will demonstrate the use of video cards in multimodel adaptive robust control, using openGL and compute shaders. A software simulation will show the behavior of the adaptive robust multimodel control scheme as the target process is exposed to both parametric and structural disturbances and will show the viability of using graphics processing units in real time systems control.

Highlights

Adaptive-robust control [1,2,3] is a strategy used to control processes with nonlinear dynamics, as it guarantees reference tracking and rejection of disruption by reidentifying the process model and controller each time performance degradation is detected
This article presented the use of graphics processing units in real time adaptive robust multimodel control, which is best suited for processes that may encounter repetitive types of parametric or structural perturbances
The multimodel aspect of this implementation is based on a knowledge base of previously discovered nominal models and their robust controllers

Summary

Introduction

Adaptive-robust control [1,2,3] is a strategy used to control processes with nonlinear dynamics, as it guarantees reference tracking and rejection of disruption by reidentifying the process model and controller each time performance degradation is detected. It requires little a priori knowledge of the process and is best suited for situations where information about system dynamics is difficult to obtain or processes are subject to various types of disruptions. Instead of trying to tune algorithms or reidentify nominal models, multimodel control stores precalculated nominal models and controllers that match various working regimes

Methods

Results

Discussion

Conclusion