On Some Problems Of Multiobjective Semidefinite Programming With Set-Valued Maps

In this thesis, we study multiobjective semidefinite programming with set-valued maps. First, the foundation of ordered topological linear space is constructed. Under the condition of quasi-nearly cone-subconvexlikeness, we study the scalarization, Lagrange function and unconstrained optimization, the condition of saddle point and the properties of duality of the primal programming. Finally, the connectedness of the solution set is disscussed.