变测度的积分-水平集确定性算法  

Variable measure deterministic algorithm of an integral-level set method

作　　者：余泓[1] 时百胜[1] 邬冬华[2] 

机构地区：[1]苏州科技学院应用数学系,江苏苏州215009 [2]上海大学理学院,上海200436

出　　处：《高校应用数学学报（A辑）》2007年第2期194-204,共11页Applied Mathematics A Journal of Chinese Universities(Ser.A)

摘　　要：提出了一个求总极值的变测度确定性算法,对不同的箱子采用不同的测度,结合确定性数论方法选取一致分布佳点集来代替Monte-Carlo随机投点,使水平值充分地下降,更快地到达全局最小,从而提高算法的计算效率.在文中给出了算法的收敛性证明,并通过数值算例验证了它的有效性.A variable measure deterministic algorithm for solving the global optimization problem is proposed. By taking different measures for different sub-boxes and choosing a good point set of uniformity with the deterministic number theory instead of Monte-Carlo method, it can make the level value reduce enough to reach the global optimization more quickly and improve the efficiency of the algorithm. The global convergence of this algorithm and its validity in the numerical value cases are proved.

关 键 词：积分-水平集 变测度 一致分布佳点集 确定性算法 

分 类 号：O221[理学—运筹学与控制论]