一个求解带线性约束非线性优化信赖域内点算法.pdfVIP

_______________________________________________________________________________ A Trust Region Interior-point Algorithm for Solving Linearly Constrained Nonlinear Optimization 1 2 Ou Yigui , Hou Dingpi 1. Department of Mathematics, Hainan University, Haikou, Hainan 570228, P.R.China 2. Department of Mathematics, University of Science and Technology of China, Hefei, Anhui 230026, P.R.China (Email: ouyigui@163.net ) Abstract: In this paper, a trust region algorithm based on interior point techniques is proposed for solving nonlinear optimization problems with linear equality constraints and nonnegative variables. Unlike those existing trust region interior–point methods, this pro- posed method does not require that a general quadratic subproblem with a trust region bound be solved at each iteration. Instead, a system of linear equations is solved to get a search direction, and then a linesearch of Armijo type is performed in this direction to obtain a new iteration point. From a computational point of view, this approach may in general reduce the computational complexity, and thus improve the computational efficiency. Under suitable conditions, any accumulation of the sequence generated by the method satisfies the first–order optimality condition, which is known as a KKT point of original problem involved. AMS(1991): 90C30; 65K05; 49M40 Key words: trust region method, interior-point method, linesearch of Armijo-type, KKT- point 1. Introduction Trust region methods for unco