线性分式规划问题的一个输出空间二分算法

A Bisect Algorithm in Output Space for Solving Linear Fractional Programming Problem

把线性分式规划问题转化为输出空间上的非线性规划问题,然后在输出空间上利用线性搜索技术确定目标函数的上界和下界,再对所获得的上下界构成的区间利用二分技术求得原问题满足精度的解;数值结果表明所提出的算法是可行的和高效的,并且可以求解大规模问题.

This paper changes the linear fraction programming problem into a nonlinear programming problem in output space. Linearity search technology is used to determine an upper boundary and a lower boundary of objective function in the output space. To obtain the original problem＇s solution to satisfy the precision, a bisect technology is used to the sector which is constituted by upper boundary and lower boundary of objective function. It is shown with the numerical results that the given algorithm is feasible and highly effective as well as can solve large scale problems.