On the Newton interior-point method for nonlinear programming problems (Q1973482): Difference between revisions

The author generalizes the global Newton interior-point method introduced by (*) \textit{A. S. El-Bakry, R. A. Tapia, T. Tsuchiya}, and \textit{Y. Zhang} [J. Optimization Theory Appl. 89, 507-541 (1996; Zbl 0851.90115)] and establishes a global convergence theory for it, under the same assumptions as those stated in (*). The idea of his generalization consists in interpreting the solution of a perturbed Newton system as a descent direction for a given merit function. Numerical experimentation on discrete optimal control problems are included.