Parallel line search in method of feasible directions
Parallel line search in method of feasible directions
复制标题
DOI：
10.1023/b:opte.0000038892.32774.0c
复制
发表时间：
2004-09-01
影响因子：
2.1
通讯作者：
St Ville, J
中科院分区：
文献类型：
--
作者：
Belegundu, AD;Damle, A;St Ville, J
本文将可行方向法中的线搜索过程并行化，并用于求解约束结构优化问题。由于目标函数在变量中是线性的，因此步长问题简化为零查找问题。也就是说，步长是沿方向向量到最近约束边界的沿着距离。零发现是在两个步骤中完成的一个“游行”沿着的方向矢量括号中的零，其次是间隔减少计划。这两个步骤都是使用MPI进行消息传递并行化的。当在工作站机群上实现时，当收敛参数为10（-6）时，复合材料压力容器的优化时间从3.5小时减少到1.5小时，当使用64个处理器时，加速比为7.0。
In this paper the line search procedure within the method of feasible directions is parallelized and used in the solution of constrained structural optimization problems. As the objective function is linear in the variables, the step size problem reduces to a zero finding problem. That is, the step size is (he distance along the direction vector to the nearest constraint boundary. Zero finding is accomplished in two steps-a 'march' along the direction vector to bracket the zero followed by an interval reduction scheme. Both these steps are parallelized using MPI for message passing. When implemented on a cluster of workstations, for a convergence parameter of 10(-6), the time for optimization of composite pressure vessel reduces from 3 1/2 hours to 1 1/2 hour when 64 processors are utilized, with a speedup of 7.0.