CAJ | 학술논문

提出了一种修正的SQP算法求解带约束的极大极小问题，仅添加一个额外的变量，将带约束的极大极小问题转化为序列二次规划问题。证明了在合理的假设条件下，序列二次规划问题的极小值点就是原问题的极小值点。数值结果表明这种SQP算法是求解带约束有限极大极小问题的一种有效算法。
제출료일충수정적SQP산법구해대약속적겁대겁소문제，부첨가일개액외적변량，장대약속적겁대겁소문제전화위서렬이차규화문제。증명료재합리적가설조건하，서렬이차규화문제적겁소치점취시원문제적겁소치점。수치결과표명저충SQP산법시구해대약속유한겁대겁소문제적일충유효산법。
A SQP algorithm to tackle constrained min-max problems was introduced .Adding just one extra variable ,a constrained min-max problem was transformed into sequence quadratic programming problem .It is proved that ,under certain reasonable assumptions ,the minimize of this sequence quadratic programming problem is equivalent to the minimize of the original constrained one .The numerical results demonstrate that the SQP algorithm is an effective and promising approach for solving constrained finite min-max problems .