Lasker P. Sinaga, Tulus, E. Herawati, Sawaluddin
MIQQP-convex is an optimization model class that combines discrete and continuous variables with objective functions or constraints in quadratic and convex forms. By relaxing the discrete variables to be continuous, this model will be generalized into a convex quadratic program. The convexity of this program guarantees that the local optimal solution will be the global optimal solution. By using the penalty function, the constrained convex quadratic program is reconstructed into an unconstrained convex quadratic program. The solution sequences will converge rapidly to the optimal solution by the combination of the penalty function and the gradient projection method in an unconstrained program. By recalling the discrete variables around the global optimal solution of the convex quadratic program, the global optimal solution approach of the MIQQP-convex is expected to be found. © 2024 Author(s).
Graduate School of Mathematics, Universitas Sumatera Utara, Medan, 20155, Indonesia; Department of Mathematics, Universitas Negeri Medan, Medan, 20221, Indonesia; Department of Mathematics, Universitas Sumatera Utara, Medan, 20155, Indonesia
