序列二次规划方法(SQP)是解决非线性规划问题最有效的算法之一,但是当QP子问题不可行时算法可能会失败.而且线搜索中的罚参数的选择通常比较困难.在文献[1]中,SQP方法得到了修正,使得QP子问题可行.在本文中,我们利用滤子技术避免了罚函数的使用同时提出了带线搜索的滤子方法,最终保证了SQP方法总是可行的,而且得到了方法的全局收敛性.
The current sequential quadratic programming (SQP) type algorithm may fail if the QP suhprohlem is infeasible. Moreover the choice of penalty function in line search is difficult. In [1], the author modified the SQP method for the former problem. In this paper,on the base of paper [1],a line search filter search technique is presented to conquer the second problem. Under some conditions,the global convergence of the algorithm in this paper has shown at the end.