Recursive Quadratic Programming Methods Based On The Augmented Lagrangian |
Posted on:2001-05-07 | Degree:Master | Type:Thesis |
Country:China | Candidate:X G Wang | Full Text:PDF |
GTID:2120360002452382 | Subject:Operational Research and Cybernetics |
Abstract/Summary: | PDF Full Text Request |
Recursive quadratic programming is a family of tech.niques developed by Bartholomew Biggs and other authors for solving nonlinear programming problems This paper de- scil)es a new method for equality constrained optimization which obtains its search di- rections from a quadratic programming subproblem based on the well nown augmented Lagrangian function It avoids the penalty parameter tending to infinity . We employ an approximated linear search procedure based on the Fletcher exact penalty function and the use of approximated derivetives of the multiplier function that avoids the need to evaluate the second order derivatives of the problem functions . We prove that the algorithm possesses global and superLinear convergence properties . At the same time numerical results are reported and some alternative strategies are considered for extending them to deal with inequality constrained.
|
Keywords/Search Tags: | RQP methods, constrained optimization problem, exact penalty func- tion, global and superlinear convergence. |
PDF Full Text Request |
Related items |