摘要
使用广义既约梯度法(GRG法)解非线性规划问题并进行了灵敏度分析。局部最优解满足二阶充分条件,问题满足非退化假设的前提下,给出带参数问题的局部最优解关于参数的连续性质。在适当的条件下建立了灵敏度信息的估计,并在后面部分中提供了使用GRG算法解非线性规划,产生灵敏度信息的计算过程。
A sensitivity analysis for nonlinear programming using generalized reduced gradient method (GRG) is made . The continuity of local solutions for the problems with parameters is established when the solutions satisfy the second order sufficient conditions and the problems meet the non - degeneracy assumptions . Under proper conditions , the estimation of sensitivity information is obtained . Finally , a computational procedure of sensitivity analysis in solving a NLP problem is given .
出处
《北京理工大学学报》
EI
CAS
CSCD
1989年第4期88-96,共9页
Transactions of Beijing Institute of Technology
关键词
非线性规划
GRG法
灵敏度
nonlinear programming , sensitivity / generalized reduced gradient method .
