A natural gradient descent algorithm for the solution of discrete algebraic Lyapunov equations based on the geodesic distance

Xiaomin Duan, Huafei Sun, Linyu Peng, Xinyu Zhao

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

A new framework to calculate the numerical solution of the discrete algebraic Lyapunov equation is proposed by using the geometric structures on the Riemannian manifold. Specifically, two algorithms based on the manifold of positive definite symmetric matrices are provided. One is a gradient descent algorithm with an objective function of the classical Euclidean distance. The other is a natural gradient descent algorithm with an objective function of the geodesic distance on the curved Riemannian manifold. Furthermore, these two algorithms are compared with a traditional iteration method. Simulation examples show that the convergence speed of the natural gradient descent algorithm is the fastest one among three algorithms.

Original languageEnglish
Pages (from-to)9899-9905
Number of pages7
JournalApplied Mathematics and Computation
Volume219
Issue number19
DOIs
Publication statusPublished - 2013
Externally publishedYes

Fingerprint

Geodesic Distance
Lyapunov Equation
Descent Algorithm
Gradient Algorithm
Gradient Descent
Algebraic Equation
Riemannian Manifold
Objective function
Symmetric Positive Definite Matrix
Speed of Convergence
Geometric Structure
Iteration Method
Euclidean Distance
Numerical Solution
Calculate
Simulation

Keywords

  • Discrete Lyapunov equation
  • Geodesic distance
  • Natural gradient
  • Riemannian manifold
  • Riemannian metric

ASJC Scopus subject areas

  • Applied Mathematics
  • Computational Mathematics

Cite this

A natural gradient descent algorithm for the solution of discrete algebraic Lyapunov equations based on the geodesic distance. / Duan, Xiaomin; Sun, Huafei; Peng, Linyu; Zhao, Xinyu.

In: Applied Mathematics and Computation, Vol. 219, No. 19, 2013, p. 9899-9905.

Research output: Contribution to journalArticle

@article{119c98e77104443e9db338ee848e95d9,
title = "A natural gradient descent algorithm for the solution of discrete algebraic Lyapunov equations based on the geodesic distance",
abstract = "A new framework to calculate the numerical solution of the discrete algebraic Lyapunov equation is proposed by using the geometric structures on the Riemannian manifold. Specifically, two algorithms based on the manifold of positive definite symmetric matrices are provided. One is a gradient descent algorithm with an objective function of the classical Euclidean distance. The other is a natural gradient descent algorithm with an objective function of the geodesic distance on the curved Riemannian manifold. Furthermore, these two algorithms are compared with a traditional iteration method. Simulation examples show that the convergence speed of the natural gradient descent algorithm is the fastest one among three algorithms.",
keywords = "Discrete Lyapunov equation, Geodesic distance, Natural gradient, Riemannian manifold, Riemannian metric",
author = "Xiaomin Duan and Huafei Sun and Linyu Peng and Xinyu Zhao",
year = "2013",
doi = "10.1016/j.amc.2013.03.119",
language = "English",
volume = "219",
pages = "9899--9905",
journal = "Applied Mathematics and Computation",
issn = "0096-3003",
publisher = "Elsevier Inc.",
number = "19",

}

TY - JOUR

T1 - A natural gradient descent algorithm for the solution of discrete algebraic Lyapunov equations based on the geodesic distance

AU - Duan, Xiaomin

AU - Sun, Huafei

AU - Peng, Linyu

AU - Zhao, Xinyu

PY - 2013

Y1 - 2013

N2 - A new framework to calculate the numerical solution of the discrete algebraic Lyapunov equation is proposed by using the geometric structures on the Riemannian manifold. Specifically, two algorithms based on the manifold of positive definite symmetric matrices are provided. One is a gradient descent algorithm with an objective function of the classical Euclidean distance. The other is a natural gradient descent algorithm with an objective function of the geodesic distance on the curved Riemannian manifold. Furthermore, these two algorithms are compared with a traditional iteration method. Simulation examples show that the convergence speed of the natural gradient descent algorithm is the fastest one among three algorithms.

AB - A new framework to calculate the numerical solution of the discrete algebraic Lyapunov equation is proposed by using the geometric structures on the Riemannian manifold. Specifically, two algorithms based on the manifold of positive definite symmetric matrices are provided. One is a gradient descent algorithm with an objective function of the classical Euclidean distance. The other is a natural gradient descent algorithm with an objective function of the geodesic distance on the curved Riemannian manifold. Furthermore, these two algorithms are compared with a traditional iteration method. Simulation examples show that the convergence speed of the natural gradient descent algorithm is the fastest one among three algorithms.

KW - Discrete Lyapunov equation

KW - Geodesic distance

KW - Natural gradient

KW - Riemannian manifold

KW - Riemannian metric

UR - http://www.scopus.com/inward/record.url?scp=84877345914&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84877345914&partnerID=8YFLogxK

U2 - 10.1016/j.amc.2013.03.119

DO - 10.1016/j.amc.2013.03.119

M3 - Article

AN - SCOPUS:84877345914

VL - 219

SP - 9899

EP - 9905

JO - Applied Mathematics and Computation

JF - Applied Mathematics and Computation

SN - 0096-3003

IS - 19

ER -