Experience with a sparse nonlinear programming algorithm
A method appropriate for solving large sparse nonlinear programming problems was described by \textit{J. T. Betts} and \textit{P. D. Frank} [J. Optimization Theory Appl. 82, No. 3, 519-541 (1994; Zbl 0843.90106)]. A review of the original method is presented with special attention given to a number of enhancements that have been made to the original algorithm which improve robustness and extend its utility. Particular attention is given to the method for constructing a modified Hessian approximation and the treatment of defective quadratic programming (QP) subproblems. After describing the algorithm four classes of computational results are presented. First, the behavior of the algorithm on an ill-conditioned boundary value problem illustrates the method for treating defective QP subproblems. Second, a summary of results on a set of large sparse test problems substantiates the choice of the default algorithm strategies. Third, the benefits of quadratic convergence are illustrated on an class of very large optimal control problems. Finally, the behavior of the overall method on a standard set of small dense test problems suggests the Hessian modification strategy may be beneficial for these applications as well. For the entire collection see [Zbl 0872.00046].
- Successive linearization methods for large-scale nonlinear programming problems
- A sparse nonlinear optimization algorithm
- Class of optimization algorithms for the solution of large nonlinear programming problems. I
- The ESA NLP solver WORHP
- Computational experience with a safeguarded barrier algorithm for sparse nonlinear programming
This page was built for publication: Experience with a sparse nonlinear programming algorithm
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4371067)
