Augmented Lagrange primal-dual approach for generalized fractional programming problems. (English) Zbl 1276.49024
Summary: In this paper, we propose a primal-dual approach for solving the generalized fractional programming problem. The outer iteration of the algorithm is a variant of interval-type Dinkelbach algorithm, while the augmented Lagrange method is adopted for solving the inner min-max subproblems. This is indeed a very unique feature of the paper because almost all Dinkelbach-type algorithms in the literature addressed only the outer iteration, while leaving the issue of how to practically solve a sequence of min-max subproblems untouched. The augmented Lagrange method attaches a set of artificial variables as well as their corresponding Lagrange multipliers to the min-max subproblem. As a result, both the primal and the dual information is available for updating the iterate points and the min-max subproblem is then reduced to a sequence of minimization problems. Numerical experiments show that the primal-dual approach can achieve a better precision in fewer iterations.
MSC:
49M37 | Numerical methods based on nonlinear programming |
90C32 | Fractional programming |
49J35 | Existence of solutions for minimax problems |
49K35 | Optimality conditions for minimax problems |
49N15 | Duality theory (optimization) |