Uniqueness of KKT multipliersin multiobjective optimization
نویسندگان
چکیده
K e y w o r d s M u l t i o b j e c t i v e optimization, Mangasarian-l'~romovitz type conditions, Second-order optimality conditions. 1. I N T R O D U C T I O N We consider the following constrained multiobjective program: min f(x), subject to x E X, (1) int R~ where the feasible region is described by inequalities and equalities X := {x • R": g(x) <_ O, h(x) = 0}, with f : R" ---* R t, g : R n ~ R m, and h : R n ---* R p differentiable functions; let K := {1 , . . . , g} , I : { 1 , . . . , m } , and J := { 1 , . . . , p } denote the index sets of the involved vector functions. Moreover, let I (~) := {i • I : gi(~) = 0} denote the index set of the active inequality constraints at a given £, • X. 0893-9659/04/$ see front matter (~) 2004 Elsevier Ltd. All rights reserved. doi: 10.1016/j.aml.2003.10.011 Typeset by .AA~-TEX 1286 G. BIGI AND M. CASTELLANI The notat ion minint R~_ marks vector minimum with respect to the cone int R~_ : 5: E X is called a local vector minimum point of (1) if and only if there exists a neighbourhood N of 2, such that no x E X M g satisfies f (2) f ( x ) E int R~_, that is, fk(x) > fk(x) for all k E K. It is well known (see [1]) that a necessary condition for • E X to be a local vector minimum point of (1) is that there exist 0 E ]~, A E R m, and # E ]~v not all zero such that ekVfk(~) + ~ ~Vg,(~) + ~ ~jWh(~) = 0, (2) k 6 K iEI j E J Aigi(~) = 0, i E I , (3) 0k_>0, k E K , A i k 0 , i E I . (4) The vectors (8, A,/z) satisfying the above system axe known as John multipliers; all the feasible points • for which John multipliers exist are called stationary. In the case of scalar optimization (~ -1), John multipliers are called K K T multipliers when -1. The structure of the set of KKT multipliers has been analysed by some authors (see, for instance, [2-6]). In particular, it has been proved in [2] that this set is nonempty and bounded if and only if the well-known Mangasarian-Fromovitz constraint qualification (MFCQ) holds. To achieve uniqueness, the following modification of (MFCQ) has been introduced in [3]. DEFINITION 1. The strict Mangasarian-Fromovitz constraint qualification (SMFCQ) holds at a stationary point • 6 X when there exJst John multipliers such that for the index sets I+(~) := {i e x(~): ~ > 0}, Io(~) := I(~) \ I+(~), the foflowing conditions hold: • Vgi(~), i E I+(~), Vhj(~), j 6 J, axe linearly independent; • there exists w E R n such that V g ~ ( ~ ) ~ < 0, i e Io(~) ,
منابع مشابه
Necessary Optimality Conditions for Multiobjective Bilevel Programs
The multiobjective bilevel program is a sequence of two optimization problems, with the upper-level problem being multiobjective and the constraint region of the upper level problem being determined implicitly by the solution set to the lower-level problem. In the case where the Karush-Kuhn-Tucker (KKT) condition is necessary and sufficient for global optimality of all lower-level problems near...
متن کاملNew second-order optimality conditions in multiobjective optimization problems: Differentiable case
To get positive Lagrange multipliers associated with each of the objective function, Maeda [Constraint qualification in multiobjective optimization problems: Differentiable case, J. Optimization Theory Appl., 80, 483–500 (1994)], gave some special sets and derived some generalized regularity conditions for first-order Karush–Kuhn– Tucker (KKT)-type necessary conditions of multiobjective optimiz...
متن کاملAn LP-Newton method: nonsmooth equations, KKT systems, and nonisolated solutions
We define a new Newton-type method for the solution of constrained systems of equations and analyze in detail its properties. Under suitable conditions, that do not include differentiability or local uniqueness of solutions, the method converges locally quadratically to a solution of the system of equations, thus filling an important gap in the existing theory. The new algorithm improves on kno...
متن کاملAn algorithm for approximating nondominated points of convex multiobjective optimization problems
In this paper, we present an algorithm for generating approximate nondominated points of a multiobjective optimization problem (MOP), where the constraints and the objective functions are convex. We provide outer and inner approximations of nondominated points and prove that inner approximations provide a set of approximate weakly nondominated points. The proposed algorithm can be appl...
متن کاملQuasi-Newton Methods for Nonconvex Constrained Multiobjective Optimization
Here, a quasi-Newton algorithm for constrained multiobjective optimization is proposed. Under suitable assumptions, global convergence of the algorithm is established.
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Appl. Math. Lett.
دوره 17 شماره
صفحات -
تاریخ انتشار 2004