An interior global method for nonlinear systems with simple bounds
- 1 August 2005
- journal article
- research article
- Published by Taylor & Francis in Optimization Methods and Software
- Vol. 20 (4-5) , 453-474
Abstract
The problem of solving a nonlinear system of equations subject to simple bounds is addressed and a new globally convergent method is developed and analyzed. The globalization process employs a trust-region strategy and possibly bends the trust-region solution by a new interior point modification of the projection onto the feasible set. The steps used are obtained by a combination of the chopped Cauchy step and the possibly modified trust-region solution. Thus, strictly feasible iterates are formed. The proposed method is shown to be globally and fast locally convergent. The practical viability of our approach is shown by a concrete implementation and numerical experience on well-known problems. The obtained results indicate that the method works well in practice.Keywords
This publication has 21 references indexed in Scilit:
- On the Convergence of a Trust-Region Method for Solving Constrained Nonlinear Equations with Degenerate SolutionsJournal of Optimization Theory and Applications, 2004
- An affine scaling trust-region approach to bound-constrained nonlinear systemsApplied Numerical Mathematics, 2003
- A web-based library for testing performance of numerical software for solving nonlinear algebraic equationsComputers & Chemical Engineering, 2002
- Strictly feasible equation-based methods for mixed complementarity problemsNumerische Mathematik, 2001
- An Active Set-Type Newton Method for Constrained Nonlinear Systems2001
- A Trust-Region Approach to Nonlinear Systems of Equalities and InequalitiesSIAM Journal on Optimization, 1999
- Engineering and Economic Applications of Complementarity ProblemsSIAM Review, 1997
- An interior point potential reduction method for constrained equationsMathematical Programming, 1996
- Mcplib: a collection of nonlinear mixed complementarity problemsOptimization Methods and Software, 1995
- Iterative linear programming strategies for constrained simulationComputers & Chemical Engineering, 1991