Multiobjective optimization is applied in extensive areas. This thesisdiscussed optimality conditions for unconstrained and inequalityconstrained quasidifferentiable multiobjective programming involvinggeneralized invexity by using the properties of quasi-differentiablefunctions and the alternative theorem. An equivalent vector programmingproblem is constructed by a modification of the objective function, andsaddle point results are presented while introducing an kind of Lagrangefunction with respect to quasidifferentiable function. At the same time,weak duality theorem, direct duality theorem and strict converse dualitytheorem are obtained respecting to Wolfe-type and Mond-Weir-type duals.Furthermore, some relevant conclusions are drawn by analyzing theoptimality conditions and duality for quasidifferentiable multiobjectivefractional programming.
|