Abstract

SummaryNull‐space methods for solving saddle point systems of equations have long been used to transform an indefinite system into a symmetric positive definite one of smaller dimension. A number of independent works in the literature have identified that we can interpret a null‐space method as a matrix factorization. We review these findings, highlight links between them, and bring them into a unified framework. We also investigate the suitability of using null‐space factorizations to derive sparse direct methods and present numerical results for both practical and academic problems.

Highlights

  • A saddle point system is an indefinite linear system of equations of the form [ ] [ ][ ] [ ] x y =A BT B0 x y f g (1)Here, we will assume that B ∈ Rm×n(n > m) has full rank and A ∈ Rn×n is symmetric positive definite on the null space of B

  • We investigate the suitability of using null-space factorizations to derive sparse direct methods and present numerical results for both practical and academic problems

  • One approach for solving Equation 1 is to use a null-space method.[1, section 6] These methods, which we will describe in detail below, have been used in the fields of optimization, structural mechanics, fluid mechanics, and electrical engineering

Read more

Summary

INTRODUCTION

A saddle point system is an indefinite linear system of equations of the form [ ] [ ][ ] [ ]. One approach for solving Equation 1 is to use a null-space method.[1, section 6] These methods, which we will describe in detail below, have been used in the fields of optimization (where they are known as reduced Hessian methods), structural mechanics (where they are known as a “force” method or “direct elimination”), fluid mechanics (where they are known as the “dual variable” method), and electrical engineering (where they are known as “loop analysis”) This approach remains popular in the large-scale optimization literature.[2,3,4,5,6,7].

NULL-SPACE METHODS AS A FACTORIZATION
Factorizations based on the fundamental basis
A first null-space factorization
A11 A12 A21 A22
The Schilders factorization
Relationship to the Schur complement factorization
Connection with Cholesky decomposition
The antitriangular factorization
Other bases and converting between factorizations
STABILITY AND SPARSITY PROPERTIES
NUMERICAL RESULTS
An academic example
Optimization examples
Resistor networks
Fluid flow problems
CONCLUSION

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.