In this paper, a new augmented Lagrangian penalty function for restricted optimization issues is studied. The twin properties of the augmented Lagrangian objective penalty feature for restrained optimization issues are proved. underneath a few conditions, the saddle point of the augmented Lagrangian objective penalty feature satisfies the primary-order Karush-Kuhn-Tucker (KKT) situation. specially, whilst the KKT circumstance holds for convex programming its saddle point exists. primarily based on the augmented Lagrangian objective penalty function, an set of rules is developed for finding a international method to an inequality constrained optimization trouble and its international convergence is also proved underneath a few situations.
Augmented Lagrangian penalty features are powerful processes to inequality restrained optimization. Their main concept is to convert a restrained optimization problem into a chain of unconstrained optimization problems which can be less complicated to remedy. Theories on and algorithms of Lagrangian penalty characteristic had been introduced in Du’s et al. works [1] . Many researchers have tried to locate opportunity augmented Lagrangian capabilities. Many literatures on augmented Lagrangian (penalty) features were posted from both theoretical and realistic factors (see [2] – [8] ), whose key issues cowl 0 gap of dual, existence of saddle point, exactness, set of rules and so forth.
All augmented Lagrangian features include elements, a Lagrangian function with a Lagrangian parameter and a penalty characteristic with a penalty parameter (see [2] – [8] ). dual and saddle point is the important thing worries of augmented Lagrangian feature. furthermore, 0 hole of Lagrangian feature’s dual is proper best for convex programming and augmented Lagrangian characteristic. consequently, augmented Lagrangian function algorithms resolve a sequence of confined optimization issues via taking differential Lagrangian parameters and penalty parameters in [2] [3] [4] [5] . Lucidi [6] and Di Pillo et al. [7] received some consequences of exact augmented Lagrangian function, however numerical consequences have been no longer given. R. S. Burachik and C. Y. Kaya gave an augmented Lagrangian scheme for a fashionable optimization problem, and established for this update primal-twin convergence the augmented penalty technique in [8] . however, in terms of computation, to apply these techniques, masses of Lagrangian parameters or penalty parameters want to be adjusted to solve a few unconstrained optimization dual troubles, which make it difficult to reap an optimization approach to the original hassle. as a result, it’s miles significant to have a look at a singular augmented Lagrangian function approach.
In current years, the penalty function technique with an objective penalty parameter has been mentioned in [9] – [16] . Burke [12] considered a more general kind. Fiacco and McCormick [13] gave a widespread introduction to sequential unconstrained minimization strategies. Mauricio and Maculan [14] mentioned a Boolean penalty technique for 0-one nonlinear programming. Meng et al. [15] studied a popular goal penalty feature technique. furthermore, Meng et al. studied homes of twin and saddle points of the augmented Lagrangian objective penalty characteristic in [16] . here, a brand new augmented Lagrangian objective penalty characteristic which differs from the one in [16] is studied. a few important effects just like the ones of the augmented Lagrangian goal penalty characteristic in [16] are acquired.
the principle conclusions of this paper include that the finest goal price of the dual problem and the top-quality goal fee of the authentic trouble is zero hole, and saddle factor is equal to the KKT circumstance of the authentic problem underneath the convexity conditions. A global set of rules and its convergence are supplied. The remainder of this paper is prepared as follows. In phase 2, an augmented Lagrangian goal penalty characteristic is described, its dual residences are proved, and an algorithm to find a worldwide solution to the unique hassle (P) with convergence is supplied. In segment three, conclusions are given.
Augmented Lagrangian penalty features are powerful processes to inequality restrained optimization. Their main concept is to convert a restrained optimization problem into a chain of unconstrained optimization problems which can be less complicated to remedy. Theories on and algorithms of Lagrangian penalty characteristic had been introduced in Du’s et al. works [1] . Many researchers have tried to locate opportunity augmented Lagrangian capabilities. Many literatures on augmented Lagrangian (penalty) features were posted from both theoretical and realistic factors (see [2] – [8] ), whose key issues cowl 0 gap of dual, existence of saddle point, exactness, set of rules and so forth.
All augmented Lagrangian features include elements, a Lagrangian function with a Lagrangian parameter and a penalty characteristic with a penalty parameter (see [2] – [8] ). dual and saddle point is the important thing worries of augmented Lagrangian feature. furthermore, 0 hole of Lagrangian feature’s dual is proper best for convex programming and augmented Lagrangian characteristic. consequently, augmented Lagrangian function algorithms resolve a sequence of confined optimization issues via taking differential Lagrangian parameters and penalty parameters in [2] [3] [4] [5] . Lucidi [6] and Di Pillo et al. [7] received some consequences of exact augmented Lagrangian function, however numerical consequences have been no longer given. R. S. Burachik and C. Y. Kaya gave an augmented Lagrangian scheme for a fashionable optimization problem, and established for this update primal-twin convergence the augmented penalty technique in [8] . however, in terms of computation, to apply these techniques, masses of Lagrangian parameters or penalty parameters want to be adjusted to solve a few unconstrained optimization dual troubles, which make it difficult to reap an optimization approach to the original hassle. as a result, it’s miles significant to have a look at a singular augmented Lagrangian function approach.
In current years, the penalty function technique with an objective penalty parameter has been mentioned in [9] – [16] . Burke [12] considered a more general kind. Fiacco and McCormick [13] gave a widespread introduction to sequential unconstrained minimization strategies. Mauricio and Maculan [14] mentioned a Boolean penalty technique for 0-one nonlinear programming. Meng et al. [15] studied a popular goal penalty feature technique. furthermore, Meng et al. studied homes of twin and saddle points of the augmented Lagrangian objective penalty characteristic in [16] . here, a brand new augmented Lagrangian objective penalty characteristic which differs from the one in [16] is studied. a few important effects just like the ones of the augmented Lagrangian goal penalty characteristic in [16] are acquired.
the principle conclusions of this paper include that the finest goal price of the dual problem and the top-quality goal fee of the authentic trouble is zero hole, and saddle factor is equal to the KKT circumstance of the authentic problem underneath the convexity conditions. A global set of rules and its convergence are supplied. The remainder of this paper is prepared as follows. In phase 2, an augmented Lagrangian goal penalty characteristic is described, its dual residences are proved, and an algorithm to find a worldwide solution to the unique hassle (P) with convergence is supplied. In segment three, conclusions are given.