Robust Recursive Quadratic Programming Algorithm Model with Global and Superlinear Convergence Properties |
| |
Authors: | F. Facchinei |
| |
Affiliation: | (1) Dipartitmento di Informatica e Sistemistica, Università di Roma La Sapienza,, Roma, Italy |
| |
Abstract: | A new, robust recursive quadratic programming algorithm model based on a continuously differentiable merit function is introduced. The algorithm is globally and superlinearly convergent, uses automatic rules for choosing the penalty parameter, and can efficiently cope with the possible inconsistency of the quadratic search subproblem. The properties of the algorithm are studied under weak a priori assumptions; in particular, the superlinear convergence rate is established without requiring strict complementarity. The behavior of the algorithm is also investigated in the case where not all of the assumptions are met. The focus of the paper is on theoretical issues; nevertheless, the analysis carried out and the solutions proposed pave the way to new and more robust RQP codes than those presently available. |
| |
Keywords: | Recursive quadratic programming exact penalty functions nonlinear programming constrained optimization regularity conditions |
本文献已被 SpringerLink 等数据库收录! |
|