An Interior-Point Algorithm for Large Scale Optimization
This paper describes an interior-point algorithm for solving large scale nonlinear programming problems. The fundamental step of the algorithm requires solution of a sparse symmetric indefinite linear system. Rowand column scaling are used to ensure that the system is well-conditioned. A globalization strategy based on a nonlinear filter is used instead of a merit function. The computational performance of the algorithm is demonstrated on a high index partial differential- algebraic equation application.
KeywordsNonlinear Program Merit Function Central Path Barrier Method Large Scale Optimization
Unable to display preview. Download preview PDF.
- 3.J. T. Betts, S. K. Eldersveld, P. D. Frank, and J. G. Lewis, An Interior-Point Nonlinear Programming Algorithm for Very Large Scale Optimization, Mathematics and Computing Technology Report M&CT-TECH-003, Mathematics and Computing Technology, The Boeing Company, PO Box 3707, Seattle, WA 98124-2207, Apr. 2000.Google Scholar
- 7.R. Fletcher and S. Leyffer, Nonlinear Programming Without a Penalty Function, University of Dundee Numerical Analysis Report NA/171, University of Dundee, Department of Mathematics, Dundee, DD1 4HN, Scotland, U.K., 1997.Google Scholar