摘要
本文主要思想是利用非线性互补函数将非线性规划问题的最优性条件(KKT条件)转化为一个半光滑的方程组,通过构造互补函数的光滑逼近函数,光滑化牛顿算法。然后利用MATLAB语言编写了光滑算法的程序,得到了算法的数值结果。通过对不同初始点,发现算法有较快的局部收敛速度。最后直接运用MATLAB的优化软件包求解进行了比较,充分地证明了算法的有效性。
we mainly consider a smoothing method for the solution of nonlinear program probloms. The main idea of this method is to reformulate the optimality conditions （KKT system） to a semi-smooth nonlinear system of equations by using a NCP function. Then we give a smoothing Newton method for solving the reformulated system by constructing the smoothing function of the NCP function. Then we use the MATLAB Language to compile a program,and get the numerical results for this method. By choosing the different initial-points,different smooth parameters and different definitions,we prove that this method is globally convergent and has fast local convergence. At last,we compare this method with the MATLAB method and it proves our method is valid.
出处
《微计算机信息》
2010年第19期228-229,236,共3页
Control & Automation
