Advanced search
Start date
Betweenand


Finding all linearly dependent rows in large-scale linear programming

Full text
Author(s):
Daniele Costa Silva
Total Authors: 1
Document type: Master's Dissertation
Press: Campinas, SP.
Institution: Universidade Estadual de Campinas (UNICAMP). Instituto de Matemática, Estatística e Computação Científica
Defense date:
Examining board members:
Aurelio Ribeiro Leite de Oliveira; Frederico Ferreira Campos Filho; Francisco de Assis Magalhães Gomes Neto
Advisor: Aurelio Ribeiro Leite de Oliveira
Abstract

The presence of dependent rows in the constraint matrix is frequent in real large-scale problems. If the method of solution adopted is the simplex method, there are efficient procedures easy to implement that circumvent this problem. The same applies when interior point methods are adopted and the resulting linear systems are solved for directed methods. However, there are large-scale problems whose only possible solution is to solve linear systems by iterative methods. In this situation, the dependent rows create a singular matrix and the iterative method does not converge. The only viable alternative is to find and remove these rows before applying the method. This dissertation proposes an efficient implementation of a procedure for detection dependent rows, include in a PCx modification that solves linear systems by iterative methods (AU)

FAPESP's process: 07/06188-0 - Redundant rows detection in large-scale linear programming problems
Grantee:Daniele Costa Silva
Support Opportunities: Scholarships in Brazil - Master