Improved robust control of nonlinear stochastic systems using uncertain models

Randa Herzallah; David Lowe

Improved robust control of nonlinear stochastic systems using uncertain models

Randa Herzallah, David Lowe

Research output: Chapter in Book/Published conference output › Conference publication

Abstract

We introduce a technique for quantifying and then exploiting uncertainty in nonlinear stochastic control systems. The approach is suboptimal though robust and relies upon the approximation of the forward and inverse plant models by neural networks, which also estimate the intrinsic uncertainty. Sampling from the resulting Gaussian distributions of the inversion based neurocontroller allows us to introduce a control law which is demonstrably more robust than traditional adaptive controllers.

Original language	English
Title of host publication	Proceedings of CONTOLO 2002
Subtitle of host publication	5th Portuguese conference on automatic control
Pages	507-512
Number of pages	6
Publication status	Published - Sept 2002
Event	5th Portuguese Conference on Automatic Control - Aveiro, Portugal Duration: 5 Sept 2002 → 7 Sept 2002

Conference

Conference	5th Portuguese Conference on Automatic Control
Abbreviated title	Controlo 2002
Country/Territory	Portugal
City	Aveiro
Period	5/09/02 → 7/09/02

Keywords

uncertainity
Neural Networks
Stochastic Systems
error bar
Distribution modelling

Access to Document

Improved robust control of nonlinear stochastic systems using uncertain models

Cite this

@inproceedings{89168d9a41df4571a186ee826359021c,

title = "Improved robust control of nonlinear stochastic systems using uncertain models",

abstract = "We introduce a technique for quantifying and then exploiting uncertainty in nonlinear stochastic control systems. The approach is suboptimal though robust and relies upon the approximation of the forward and inverse plant models by neural networks, which also estimate the intrinsic uncertainty. Sampling from the resulting Gaussian distributions of the inversion based neurocontroller allows us to introduce a control law which is demonstrably more robust than traditional adaptive controllers.",

keywords = "uncertainity, Neural Networks, Stochastic Systems, error bar, Distribution modelling",

author = "Randa Herzallah and David Lowe",

year = "2002",

month = sep,

language = "English",

pages = "507--512",

booktitle = "Proceedings of CONTOLO 2002",

note = "5th Portuguese Conference on Automatic Control, Controlo 2002 ; Conference date: 05-09-2002 Through 07-09-2002",

}

TY - GEN

T1 - Improved robust control of nonlinear stochastic systems using uncertain models

AU - Herzallah, Randa

AU - Lowe, David

PY - 2002/9

Y1 - 2002/9

N2 - We introduce a technique for quantifying and then exploiting uncertainty in nonlinear stochastic control systems. The approach is suboptimal though robust and relies upon the approximation of the forward and inverse plant models by neural networks, which also estimate the intrinsic uncertainty. Sampling from the resulting Gaussian distributions of the inversion based neurocontroller allows us to introduce a control law which is demonstrably more robust than traditional adaptive controllers.

AB - We introduce a technique for quantifying and then exploiting uncertainty in nonlinear stochastic control systems. The approach is suboptimal though robust and relies upon the approximation of the forward and inverse plant models by neural networks, which also estimate the intrinsic uncertainty. Sampling from the resulting Gaussian distributions of the inversion based neurocontroller allows us to introduce a control law which is demonstrably more robust than traditional adaptive controllers.

KW - uncertainity

KW - Neural Networks

KW - Stochastic Systems

KW - error bar

KW - Distribution modelling

M3 - Conference publication

SP - 507

EP - 512

BT - Proceedings of CONTOLO 2002

T2 - 5th Portuguese Conference on Automatic Control

Y2 - 5 September 2002 through 7 September 2002

ER -

Improved robust control of nonlinear stochastic systems using uncertain models

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this