序列二次规划(SOP)方法求解中小规模非线性约束下的最优化问题很有效,因此,笔者研究了序列二次规划(SQP)方法及其在非线性规划中的应用.利用强次可行的SOP方法求解问题时可以避免罚函数应用的不便,笔者通过修订搜索方向克服了SOP方法中经常出现的Maratos现象,并且通过调整参数降低了计算量,从而提高了算法的收敛速度.在给定的条件下证明了算法的全局收敛性.
The successive quadratic programming method is effective in solving small-scale nonlinear programming restrain problem. In this paper, successive quadratic programming method being used in nonlinear programming was discussed. By using a strongly subfeasible SQP algorithm it can obviate the Maratos effect though revise the search direction, and avoid penalty function of inconvenient in the used. Through proper adjustment scalar reducing calculate and improving the convergence speed, the global convergence of this algorithm is proved in certain condition.