Dual Theory Research Of Generalized Geometric Programming Based On Decomposing Method

In this paper,a dual algorithm is given to solve the Generalized Geometric Programming(GGP)with positive difficulty degree.The prerequisite of strong duality property of this dual method is also given in this paper.Generalized Geometric Programming and its several equivalent form are discussed in the first part.Then we discuss Generalized Geometric Programming with zero difficulty degree and positive degree respectively.When the degree is zero,the programming already has a mature solution which decomposes the equivalent programming into several sub-programming.When the degree is positive,a Lagrange dual is formed to transfer the primal problem into the dual programming.By solving the dual problem we can get a lower bound of primal problem.With the theory of saddle point optimality we can achieve the sufficient and necessary condition of the strong duality property.Then,we give the proof of the strong duality property between the original problem and dual problem when the primal problem meets some specific conditons.At last,the condition that whether the maximum point is in the feasible region is also included.