Next: B. Messirdi, A. Rahmani,
Up: TOME 49 (97), no. 4
Previous: Mircea Cimpoeas: A generalization
Abstract:
In this paper the solution of nonlinear programming problems (NLP) by a
two-parameter penalty function method is considered. An algorithm that
combines penalty concepts and sequential quadratic programming techniques is
presented. The approach taken is to replace the initial problem by the more
tractable one of minimizing a non-differentiable penalty function chosen so
that the solutions of the NLP are also solutions of the penalty function
problem.
Key Words: Nonlinear Optimization, Lagrange Multipliers,
Penalty Function.
2000 Mathematics Subject Classification: Primary: 90C30,
Secondary:
90C26.
Download the paper in pdf format here.