Étude théorique et numérique d’un algorithme de point intérieur de type correcteur-prédicteur pour la programmation linéaire.

Abstract

Complexité polynomiale. • Abstract: In this work, we studied a primal-dual algorithm of interior points of corrector-predictor type based on a new search direction to solve a linear problem (LP ), we have introduce algebraic transformation on the equation of centrality xz = µe. By the study of Darvay(2020), ψ(t) = t − √ t who proved that the algorithm has polynomial complexity, we have done comparative numerical tests between the theoretical choice of the displacement step during the prediction phase and the alternative choice to see the in uence of these parameters on th

Complexité polynomiale. • Abstract: In this work, we studied a primal-dual algorithm of interior points of corrector-predictor type based on a new search direction to solve a linear problem (LP ), we have introduce algebraic transformation on the equation of centrality xz = µe. By the study of Darvay(2020), ψ(t) = t − √ t who proved that the algorithm has polynomial complexity, we have done comparative numerical tests between the theoretical choice of the displacement step during the prediction phase and the alternative choice to see the in uence of these parameters on th