Implementation of the primal-dual interior-point algorithm for the solution of the minimum cost network flow problem
In this talk we describe an implementation of the primal-dual (PD) interior-point algorithm for the solution of large-scale minimum cost network flow problems. In this scheme the search direction is found by a Preconditioned Conjugate-Gradient (PCG) algorithm incorporating an incomplete QR decomposition technique that takes full advantage of the network structure that is associated with the minimum cost flow problem. Some other implementation details, namely, the stopping criteria for the PD and PCG algorithms and the starting point for the PD method, are also discussed. Computational experience with large-scale minimum cost network flow problems is reported and shows the validity of the techniques described in this talk.
- OSTI ID:
- 36404
- Report Number(s):
- CONF-9408161--
- Country of Publication:
- United States
- Language:
- English
Similar Records
Primal-dual approximation algorithms for the survivable network design problem
A direct nonlinear predictor-corrector primal-dual interior point algorithm for optimal power flows