摘要
利用广义投影校正技术对搜索方向进行某种修正,改进假设条件,采用一种新型的一阶修正方向并结合SQP技术,建立了求解非线性约束最优化问题(p)的一个新的SQP可行下降算法,在较温和的假设条件下证明了算法的全局收敛性.由于新算法仅需较小的存储,从而适合大规模最优化问题的计算.
In this paper, by using generalized projection rectify technique, Mutilated search direction, through improved suppose conditions, by using a new one order correcting direction and combining with the SQP skill, a new SQP feasible descent algorithm for nonlinear constrained optimitation problem （p） is presented, and under weaker conditions, we proofed the new methods still possesses global convergence, the new methods only use little storage, thus the methods are attractine for largeseale problems.
出处
《数学的实践与认识》
CSCD
北大核心
2006年第8期234-238,共5页
Mathematics in Practice and Theory
基金
国家自然科学重点基金(10231060)
河南省自然科学基金(0511013600
0611056100)
关键词
约束最优化
序列二次规划
广义投影
线搜索
全局收敛性
constrained optimization
Sequential quadratic programming
the generalized projection
llne search,global convergence
