Papers by Author: Jing Ben Yin

Abstract: The sum of linear fractional functions problem has attracted the interest of researchers and practitioners for a number of years. Since these types of optimization problems are non-convex, various specialized algorithms have been proposed for globally solving these problems. However, these algorithms are only for the case that sum of linear ratios problem without coefficients, and may be difficult to be solved. In this paper, a deterministic algorithm is proposed for globally solving the sum of linear fractional functions problem with coefficients. By utilizing an equivalent problem and linear relaxation technique, the initial non-convex programming problem is reduced to a sequence of linear relaxation programming problems. The proposed algorithm is convergent to the global optimal solution by means of the subsequent solutions of a series of linear programming problems.

Abstract: Multiplicative problems are a kind of difficult global optimization problems known to be NP-hard. At the same time, these problems have some important applications in engineering, system, finance, economics, and other fields. In this paper, an optimization method is proposed to globally solve a class of multiplicative problems with coefficients. Firstly, by utilizing equivalent transformation and linearization method, a linear relaxation programming problem is established. Secondly, by using branch and bound technique, a determined algorithm is proposed for solving equivalent problem. Finally, the proposed algorithm is convergent to the global optimal solution of original problem by means of the subsequent solutions of a series of linear programming problems.

Abstract: In this technical note, we develop an approach to globally solve a class of optimization problems in system engineering based on the recent paper ([1]). Actually the problem we investigated is more general, since we extend numerators and denominators of linear ratios to generalized polynomial functions. And we give a new linear relaxation method for obtaining the lower bound of problems. Our approach is easy to be implemented, since it need not additional special program to the upper and lower bound for numerator and denominator of each generalized polynomial ratio.

Abstract: In this paper, we proposed an algorithm to globally solve a class of mathematical problems in mechanical system. Firstly, by utilizing equivalent problem and linear relaxation technique, a linear relaxation programming of original mathematical problem is established. Secondly, by using branch and bound theory, a feasible algorithm is proposed for globally solving original problem. Finally, the convergence of the proposed algorithm is proven, and numerical experiments showed that the presented algorithm is feasible.

Abstract: In this paper, we develop an algorithm to globally solve a class of mathematical models in system engineering. Firstly, by utilizing equivalent problem and linear relaxation method, a linear relaxation programming of original problem is established. Secondly, by using branch and bound technique, a determined branch and bound algorithm is proposed for globally solving original problem. Finally, the convergence of the proposed algorithm is given and numerical examples showed that the presented algorithm is feasible.