机构地区: 上海大学理学院,上海200444
出 处: 《应用数学与计算数学学报》 2017年第3期341-349,共9页
摘 要: 运用双曲和三角变差积分以及罚函数技术研究和求解约束总极值问题,给出了其罚最优性条件及罚双曲和三角变差积分算法.结合Monte-Carlo技术,特别针对n=100个变量具有不连续约束总极值问题进行了数值模拟,计算结果表明所设方法是可行性的. In this paper, our main work is to study and solve the constrained global optimization problem with the hyperbolic and trigonal deviation integral and penalty technique. ~rthermore, we examine the penalty optimality condition and hyperbolic and trigonal penalty deviation integral algorithm for the constrained global optimization problem. Using the Monte-Carlo technique, we analyse specific examples for a discontinuous objective function with n = 100 variables. Computa- tional results show that the algorithm is feasible.